Technologies For Managing Allocation Of Accelerator Resources

Abstract

Technologies for dynamically managing the allocation of accelerator resources include an orchestrator server. The orchestrator server is to assign a workload to a managed node for execution, determine a predicted demand for one or more accelerator resources to accelerate the execution of one or more jobs within the workload, provision, prior to the predicted demand, one or more accelerator resources to accelerate the one or more jobs, and allocate the one or more provisioned accelerator resources to the managed node to accelerate the execution of the one or more jobs. Other embodiments are also described and claimed.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] The present application claims the benefit of U.S. Provisional Patent Application No. 62/365,969, filed Jul. 22, 2016, U.S. Provisional Patent Application No. 62/376,859, filed Aug. 18, 2016, and U.S. Provisional Patent Application No. 62/427,268, filed Nov. 29, 2016.

BACKGROUND

[0002] In a typical cloud-based computing environment (e.g., a data center), multiple compute nodes may execute workloads (e.g., processes, applications, services, etc.) on behalf of customers. One or more of the workloads may include sets of functions (e.g., jobs), that could be accelerated using accelerator resources such as field programmable gate arrays (FPGAs), dedicated graphics processors, or other specialized devices for accelerating specific types of jobs. In typical data centers, all or a subset of the compute nodes may be physically equipped (e.g., on the same board as the central processing unit) with one or more accelerator resources. However, in such data centers, the accelerator resources may go unused or may be used only a subset of the time that the workloads are being executed, as many jobs assigned to the compute nodes may not include jobs that are amenable to acceleration. Furthermore, even in data centers in which each compute node is assembled from resources distributed across the data center when a workload is assigned to the compute node, information regarding whether the assigned workload may benefit from acceleration may be unavailable. As such, the compute node may be assembled without the accelerator resources that could be beneficial to the execution of the workload, or may be assembled with one or more accelerator resources that are underutilized (e.g., idle more than a threshold amount of time) during the execution of the workload. As such, the allocation of accelerator resources in typical data centers is problematic and can often result in inefficient use of resources and, as result, unnecessary costs for the operator of the data center.

BRIEF DESCRIPTION OF THE DRAWINGS

[0003] The concepts described herein are illustrated by way of example and not by way of limitation in the accompanying figures. For simplicity and clarity of illustration, elements illustrated in the figures are not necessarily drawn to scale. Where considered appropriate, reference labels have been repeated among the figures to indicate corresponding or analogous elements.

[0004] FIG. 1 is a diagram of a conceptual overview of a data center in which one or more techniques described herein may be implemented according to various embodiments;

[0005] FIG. 2 is a diagram of an example embodiment of a logical configuration of a rack of the data center of FIG. 1;

[0006] FIG. 3 is a diagram of an example embodiment of another data center in which one or more techniques described herein may be implemented according to various embodiments;

[0007] FIG. 4 is a diagram of another example embodiment of a data center in which one or more techniques described herein may be implemented according to various embodiments;

[0008] FIG. 5 is a diagram of a connectivity scheme representative of link-layer connectivity that may be established among various sleds of the data centers of FIGS. 1, 3, and 4;

[0009] FIG. 6 is a diagram of a rack architecture that may be representative of an architecture of any particular one of the racks depicted in FIGS. 1-4 according to some embodiments;

[0010] FIG. 7 is a diagram of an example embodiment of a sled that may be used with the rack architecture of FIG. 6;

[0011] FIG. 8 is a diagram of an example embodiment of a rack architecture to provide support for sleds featuring expansion capabilities;

[0012] FIG. 9 is a diagram of an example embodiment of a rack implemented according to the rack architecture of FIG. 8;

[0013] FIG. 10 is a diagram of an example embodiment of a sled designed for use in conjunction with the rack of FIG. 9;

[0014] FIG. 11 is a diagram of an example embodiment of a data center in which one or more techniques described herein may be implemented according to various embodiments;

[0015] FIG. 12 is a simplified block diagram of at least one embodiment of a system for managing the allocation of accelerator resources to managed nodes;

[0016] FIG. 13 is a simplified block diagram of at least one embodiment of an orchestrator server of the system of FIG. 12;

[0017] FIG. 14 is a simplified block diagram of at least one embodiment of an environment that may be established by the orchestrator server of FIGS. 12 and 13; and

[0018] FIGS. 15-17 are a simplified flow diagram of at least one embodiment of a method for managing the allocation of accelerator resources among managed nodes as the managed nodes execute workloads, that may be performed by the orchestrator server of FIGS. 12-14.

DETAILED DESCRIPTION OF THE DRAWINGS

[0019] While the concepts of the present disclosure are susceptible to various modifications and alternative forms, specific embodiments thereof have been shown by way of example in the drawings and will be described herein in detail. It should be understood, however, that there is no intent to limit the concepts of the present disclosure to the particular forms disclosed, but on the contrary, the intention is to cover all modifications, equivalents, and alternatives consistent with the present disclosure and the appended claims.

[0020] References in the specification to "one embodiment," "an embodiment," "an illustrative embodiment," etc., indicate that the embodiment described may include a particular feature, structure, or characteristic, but every embodiment may or may not necessarily include that particular feature, structure, or characteristic. Moreover, such phrases are not necessarily referring to the same embodiment. Further, when a particular feature, structure, or characteristic is described in connection with an embodiment, it is submitted that it is within the knowledge of one skilled in the art to effect such feature, structure, or characteristic in connection with other embodiments whether or not explicitly described. Additionally, it should be appreciated that items included in a list in the form of "at least one A, B, and C" can mean (A); (B); (C); (A and B); (A and C); (B and C); or (A, B, and C). Similarly, items listed in the form of "at least one of A, B, or C" can mean (A); (B); (C); (A and B); (A and C); (B and C); or (A, B, and C).

[0021] The disclosed embodiments may be implemented, in some cases, in hardware, firmware, software, or any combination thereof. The disclosed embodiments may also be implemented as instructions carried by or stored on a transitory or non-transitory machine-readable (e.g., computer-readable) storage medium, which may be read and executed by one or more processors. A machine-readable storage medium may be embodied as any storage device, mechanism, or other physical structure for storing or transmitting information in a form readable by a machine (e.g., a volatile or non-volatile memory, a media disc, or other media device).

[0022] In the drawings, some structural or method features may be shown in specific arrangements and/or orderings. However, it should be appreciated that such specific arrangements and/or orderings may not be required. Rather, in some embodiments, such features may be arranged in a different manner and/or order than shown in the illustrative figures. Additionally, the inclusion of a structural or method feature in a particular figure is not meant to imply that such feature is required in all embodiments and, in some embodiments, may not be included or may be combined with other features.

[0023] FIG. 1 illustrates a conceptual overview of a data center 100 that may generally be representative of a data center or other type of computing network in/for which one or more techniques described herein may be implemented according to various embodiments. As shown in FIG. 1, data center 100 may generally contain a plurality of racks, each of which may house computing equipment comprising a respective set of physical resources. In the particular non-limiting example depicted in FIG. 1, data center 100 contains four racks 102A to 102D, which house computing equipment comprising respective sets of physical resources (PCRs) 105A to 105D. According to this example, a collective set of physical resources 106 of data center 100 includes the various sets of physical resources 105A to 105D that are distributed among racks 102A to 102D. Physical resources 106 may include resources of multiple types, such as--for example--processors, co-processors, accelerators, field programmable gate arrays (FPGAs), memory, and storage. The embodiments are not limited to these examples.

[0024] The illustrative data center 100 differs from typical data centers in many ways. For example, in the illustrative embodiment, the circuit boards ("sleds") on which components such as CPUs, memory, and other components are placed are designed for increased thermal performance In particular, in the illustrative embodiment, the sleds are shallower than typical boards. In other words, the sleds are shorter from the front to the back, where cooling fans are located. This decreases the length of the path that air must to travel across the components on the board. Further, the components on the sled are spaced further apart than in typical circuit boards, and the components are arranged to reduce or eliminate shadowing (i.e., one component in the air flow path of another component). In the illustrative embodiment, processing components such as the processors are located on a top side of a sled while near memory, such as DIMMs, are located on a bottom side of the sled. As a result of the enhanced airflow provided by this design, the components may operate at higher frequencies and power levels than in typical systems, thereby increasing performance. Furthermore, the sleds are configured to blindly mate with power and data communication cables in each rack 102A, 102B, 102C, 102D, enhancing their ability to be quickly removed, upgraded, reinstalled, and/or replaced. Similarly, individual components located on the sleds, such as processors, accelerators, memory, and data storage drives, are configured to be easily upgraded due to their increased spacing from each other. In the illustrative embodiment, the components additionally include hardware attestation features to prove their authenticity.

[0025] Furthermore, in the illustrative embodiment, the data center 100 utilizes a single network architecture ("fabric") that supports multiple other network architectures including Ethernet and Omni-Path. The sleds, in the illustrative embodiment, are coupled to switches via optical fibers, which provide higher bandwidth and lower latency than typical twisted pair cabling (e.g., Category 5, Category 5e, Category 6, etc.). Due to the high bandwidth, low latency interconnections and network architecture, the data center 100 may, in use, pool resources, such as memory, accelerators (e.g., graphics accelerators, FPGAs, ASICs, etc.), and data storage drives that are physically disaggregated, and provide them to compute resources (e.g., processors) on an as needed basis, enabling the compute resources to access the pooled resources as if they were local. The illustrative data center 100 additionally receives usage information for the various resources, predicts resource usage for different types of workloads based on past resource usage, and dynamically reallocates the resources based on this information.

[0026] The racks 102A, 102B, 102C, 102D of the data center 100 may include physical design features that facilitate the automation of a variety of types of maintenance tasks. For example, data center 100 may be implemented using racks that are designed to be robotically-accessed, and to accept and house robotically-manipulatable resource sleds. Furthermore, in the illustrative embodiment, the racks 102A, 102B, 102C, 102D include integrated power sources that receive a greater voltage than is typical for power sources. The increased voltage enables the power sources to provide additional power to the components on each sled, enabling the components to operate at higher than typical frequencies.

[0027] FIG. 2 illustrates an exemplary logical configuration of a rack 202 of the data center 100. As shown in FIG. 2, rack 202 may generally house a plurality of sleds, each of which may comprise a respective set of physical resources. In the particular non-limiting example depicted in FIG. 2, rack 202 houses sleds 204-1 to 204-4 comprising respective sets of physical resources 205-1 to 205-4, each of which constitutes a portion of the collective set of physical resources 206 comprised in rack 202. With respect to FIG. 1, if rack 202 is representative of--for example--rack 102A, then physical resources 206 may correspond to the physical resources 105A comprised in rack 102A. In the context of this example, physical resources 105A may thus be made up of the respective sets of physical resources, including physical storage resources 205-1, physical accelerator resources 205-2, physical memory resources 205-3, and physical compute resources 205-5 comprised in the sleds 204-1 to 204-4 of rack 202. The embodiments are not limited to this example. Each sled may contain a pool of each of the various types of physical resources (e.g., compute, memory, accelerator, storage). By having robotically accessible and robotically manipulatable sleds comprising disaggregated resources, each type of resource can be upgraded independently of each other and at their own optimized refresh rate.

[0028] FIG. 3 illustrates an example of a data center 300 that may generally be representative of one in/for which one or more techniques described herein may be implemented according to various embodiments. In the particular non-limiting example depicted in FIG. 3, data center 300 comprises racks 302-1 to 302-32. In various embodiments, the racks of data center 300 may be arranged in such fashion as to define and/or accommodate various access pathways. For example, as shown in FIG. 3, the racks of data center 300 may be arranged in such fashion as to define and/or accommodate access pathways 311A, 311B, 311C, and 311D. In some embodiments, the presence of such access pathways may generally enable automated maintenance equipment, such as robotic maintenance equipment, to physically access the computing equipment housed in the various racks of data center 300 and perform automated maintenance tasks (e.g., replace a failed sled, upgrade a sled). In various embodiments, the dimensions of access pathways 311A, 311B, 311C, and 311D, the dimensions of racks 302-1 to 302-32, and/or one or more other aspects of the physical layout of data center 300 may be selected to facilitate such automated operations. The embodiments are not limited in this context.

[0029] FIG. 4 illustrates an example of a data center 400 that may generally be representative of one in/for which one or more techniques described herein may be implemented according to various embodiments. As shown in FIG. 4, data center 400 may feature an optical fabric 412. Optical fabric 412 may generally comprise a combination of optical signaling media (such as optical cabling) and optical switching infrastructure via which any particular sled in data center 400 can send signals to (and receive signals from) each of the other sleds in data center 400. The signaling connectivity that optical fabric 412 provides to any given sled may include connectivity both to other sleds in a same rack and sleds in other racks. In the particular non-limiting example depicted in FIG. 4, data center 400 includes four racks 402A to 402D. Racks 402A to 402D house respective pairs of sleds 404A-1 and 404A-2, 404B-1 and 404B-2, 404C-1 and 404C-2, and 404D-1 and 404D-2. Thus, in this example, data center 400 comprises a total of eight sleds. Via optical fabric 412, each such sled may possess signaling connectivity with each of the seven other sleds in data center 400. For example, via optical fabric 412, sled 404A-1 in rack 402A may possess signaling connectivity with sled 404A-2 in rack 402A, as well as the six other sleds 404B-1, 404B-2, 404C-1, 404C-2, 404D-1, and 404D-2 that are distributed among the other racks 402B, 402C, and 402D of data center 400. The embodiments are not limited to this example.

[0030] FIG. 5 illustrates an overview of a connectivity scheme 500 that may generally be representative of link-layer connectivity that may be established in some embodiments among the various sleds of a data center, such as any of example data centers 100, 300, and 400 of FIGS. 1, 3, and 4. Connectivity scheme 500 may be implemented using an optical fabric that features a dual-mode optical switching infrastructure 514. Dual-mode optical switching infrastructure 514 may generally comprise a switching infrastructure that is capable of receiving communications according to multiple link-layer protocols via a same unified set of optical signaling media, and properly switching such communications. In various embodiments, dual-mode optical switching infrastructure 514 may be implemented using one or more dual-mode optical switches 515. In various embodiments, dual-mode optical switches 515 may generally comprise high-radix switches. In some embodiments, dual-mode optical switches 515 may comprise multi-ply switches, such as four-ply switches. In various embodiments, dual-mode optical switches 515 may feature integrated silicon photonics that enable them to switch communications with significantly reduced latency in comparison to conventional switching devices. In some embodiments, dual-mode optical switches 515 may constitute leaf switches 530 in a leaf-spine architecture additionally including one or more dual-mode optical spine switches 520.

[0031] In various embodiments, dual-mode optical switches may be capable of receiving both Ethernet protocol communications carrying Internet Protocol (IP packets) and communications according to a second, high-performance computing (HPC) link-layer protocol (e.g., Intel's Omni-Path Architecture's, Infiniband) via optical signaling media of an optical fabric. As reflected in FIG. 5, with respect to any particular pair of sleds 504A and 504B possessing optical signaling connectivity to the optical fabric, connectivity scheme 500 may thus provide support for link-layer connectivity via both Ethernet links and HPC links. Thus, both Ethernet and HPC communications can be supported by a single high-bandwidth, low-latency switch fabric. The embodiments are not limited to this example.

[0032] FIG. 6 illustrates a general overview of a rack architecture 600 that may be representative of an architecture of any particular one of the racks depicted in FIGS. 1 to 4 according to some embodiments. As reflected in FIG. 6, rack architecture 600 may generally feature a plurality of sled spaces into which sleds may be inserted, each of which may be robotically-accessible via a rack access region 601. In the particular non-limiting example depicted in FIG. 6, rack architecture 600 features five sled spaces 603-1 to 603-5. Sled spaces 603-1 to 603-5 feature respective multi-purpose connector modules (MPCMs) 616-1 to 616-5.

[0033] FIG. 7 illustrates an example of a sled 704 that may be representative of a sled of such a type. As shown in FIG. 7, sled 704 may comprise a set of physical resources 705, as well as an MPCM 716 designed to couple with a counterpart MPCM when sled 704 is inserted into a sled space such as any of sled spaces 603-1 to 603-5 of FIG. 6. Sled 704 may also feature an expansion connector 717. Expansion connector 717 may generally comprise a socket, slot, or other type of connection element that is capable of accepting one or more types of expansion modules, such as an expansion sled 718. By coupling with a counterpart connector on expansion sled 718, expansion connector 717 may provide physical resources 705 with access to supplemental computing resources 705B residing on expansion sled 718. The embodiments are not limited in this context.

[0034] FIG. 8 illustrates an example of a rack architecture 800 that may be representative of a rack architecture that may be implemented in order to provide support for sleds featuring expansion capabilities, such as sled 704 of FIG. 7. In the particular non-limiting example depicted in FIG. 8, rack architecture 800 includes seven sled spaces 803-1 to 803-7, which feature respective MPCMs 816-1 to 816-7. Sled spaces 803-1 to 803-7 include respective primary regions 803-1A to 803-7A and respective expansion regions 803-1B to 803-7B. With respect to each such sled space, when the corresponding MPCM is coupled with a counterpart MPCM of an inserted sled, the primary region may generally constitute a region of the sled space that physically accommodates the inserted sled. The expansion region may generally constitute a region of the sled space that can physically accommodate an expansion module, such as expansion sled 718 of FIG. 7, in the event that the inserted sled is configured with such a module.

[0035] FIG. 9 illustrates an example of a rack 902 that may be representative of a rack implemented according to rack architecture 800 of FIG. 8 according to some embodiments. In the particular non-limiting example depicted in FIG. 9, rack 902 features seven sled spaces 903-1 to 903-7, which include respective primary regions 903-1A to 903-7A and respective expansion regions 903-1B to 903-7B. In various embodiments, temperature control in rack 902 may be implemented using an air cooling system. For example, as reflected in FIG. 9, rack 902 may feature a plurality of fans 919 that are generally arranged to provide air cooling within the various sled spaces 903-1 to 903-7. In some embodiments, the height of the sled space is greater than the conventional "1U" server height. In such embodiments, fans 919 may generally comprise relatively slow, large diameter cooling fans as compared to fans used in conventional rack configurations. Running larger diameter cooling fans at lower speeds may increase fan lifetime relative to smaller diameter cooling fans running at higher speeds while still providing the same amount of cooling. The sleds are physically shallower than conventional rack dimensions. Further, components are arranged on each sled to reduce thermal shadowing (i.e., not arranged serially in the direction of air flow). As a result, the wider, shallower sleds allow for an increase in device performance because the devices can be operated at a higher thermal envelope (e.g., 250 W) due to improved cooling (i.e., no thermal shadowing, more space between devices, more room for larger heat sinks, etc.).

[0036] MPCMs 916-1 to 916-7 may be configured to provide inserted sleds with access to power sourced by respective power modules 920-1 to 920-7, each of which may draw power from an external power source 921. In various embodiments, external power source 921 may deliver alternating current (AC) power to rack 902, and power modules 920-1 to 920-7 may be configured to convert such AC power to direct current (DC) power to be sourced to inserted sleds. In some embodiments, for example, power modules 920-1 to 920-7 may be configured to convert 277-volt AC power into 12-volt DC power for provision to inserted sleds via respective MPCMs 916-1 to 916-7. The embodiments are not limited to this example.

[0037] MPCMs 916-1 to 916-7 may also be arranged to provide inserted sleds with optical signaling connectivity to a dual-mode optical switching infrastructure 914, which may be the same as--or similar to--dual-mode optical switching infrastructure 514 of FIG. 5. In various embodiments, optical connectors contained in MPCMs 916-1 to 916-7 may be designed to couple with counterpart optical connectors contained in MPCMs of inserted sleds to provide such sleds with optical signaling connectivity to dual-mode optical switching infrastructure 914 via respective lengths of optical cabling 922-1 to 922-7. In some embodiments, each such length of optical cabling may extend from its corresponding MPCM to an optical interconnect loom 923 that is external to the sled spaces of rack 902. In various embodiments, optical interconnect loom 923 may be arranged to pass through a support post or other type of load-bearing element of rack 902. The embodiments are not limited in this context. Because inserted sleds connect to an optical switching infrastructure via MPCMs, the resources typically spent in manually configuring the rack cabling to accommodate a newly inserted sled can be saved.

[0038] FIG. 10 illustrates an example of a sled 1004 that may be representative of a sled designed for use in conjunction with rack 902 of FIG. 9 according to some embodiments. Sled 1004 may feature an MPCM 1016 that comprises an optical connector 1016A and a power connector 1016B, and that is designed to couple with a counterpart MPCM of a sled space in conjunction with insertion of MPCM 1016 into that sled space. Coupling MPCM 1016 with such a counterpart MPCM may cause power connector 1016 to couple with a power connector comprised in the counterpart MPCM. This may generally enable physical resources 1005 of sled 1004 to source power from an external source, via power connector 1016 and power transmission media 1024 that conductively couples power connector 1016 to physical resources 1005.

[0039] Sled 1004 may also include dual-mode optical network interface circuitry 1026. Dual-mode optical network interface circuitry 1026 may generally comprise circuitry that is capable of communicating over optical signaling media according to each of multiple link-layer protocols supported by dual-mode optical switching infrastructure 914 of FIG. 9. In some embodiments, dual-mode optical network interface circuitry 1026 may be capable both of Ethernet protocol communications and of communications according to a second, high-performance protocol. In various embodiments, dual-mode optical network interface circuitry 1026 may include one or more optical transceiver modules 1027, each of which may be capable of transmitting and receiving optical signals over each of one or more optical channels. The embodiments are not limited in this context.

[0040] Coupling MPCM 1016 with a counterpart MPCM of a sled space in a given rack may cause optical connector 1016A to couple with an optical connector comprised in the counterpart MPCM. This may generally establish optical connectivity between optical cabling of the sled and dual-mode optical network interface circuitry 1026, via each of a set of optical channels 1025. Dual-mode optical network interface circuitry 1026 may communicate with the physical resources 1005 of sled 1004 via electrical signaling media 1028. In addition to the dimensions of the sleds and arrangement of components on the sleds to provide improved cooling and enable operation at a relatively higher thermal envelope (e.g., 250 W), as described above with reference to FIG. 9, in some embodiments, a sled may include one or more additional features to facilitate air cooling, such as a heatpipe and/or heat sinks arranged to dissipate heat generated by physical resources 1005. It is worthy of note that although the example sled 1004 depicted in FIG. 10 does not feature an expansion connector, any given sled that features the design elements of sled 1004 may also feature an expansion connector according to some embodiments. The embodiments are not limited in this context.

[0041] FIG. 11 illustrates an example of a data center 1100 that may generally be representative of one in/for which one or more techniques described herein may be implemented according to various embodiments. As reflected in FIG. 11, a physical infrastructure management framework 1150A may be implemented to facilitate management of a physical infrastructure 1100A of data center 1100. In various embodiments, one function of physical infrastructure management framework 1150A may be to manage automated maintenance functions within data center 1100, such as the use of robotic maintenance equipment to service computing equipment within physical infrastructure 1100A. In some embodiments, physical infrastructure 1100A may feature an advanced telemetry system that performs telemetry reporting that is sufficiently robust to support remote automated management of physical infrastructure 1100A. In various embodiments, telemetry information provided by such an advanced telemetry system may support features such as failure prediction/prevention capabilities and capacity planning capabilities. In some embodiments, physical infrastructure management framework 1150A may also be configured to manage authentication of physical infrastructure components using hardware attestation techniques. For example, robots may verify the authenticity of components before installation by analyzing information collected from a radio frequency identification (RFID) tag associated with each component to be installed. The embodiments are not limited in this context.

[0042] As shown in FIG. 11, the physical infrastructure 1100A of data center 1100 may comprise an optical fabric 1112, which may include a dual-mode optical switching infrastructure 1114. Optical fabric 1112 and dual-mode optical switching infrastructure 1114 may be the same as--or similar to--optical fabric 412 of FIG. 4 and dual-mode optical switching infrastructure 514 of FIG. 5, respectively, and may provide high-bandwidth, low-latency, multi-protocol connectivity among sleds of data center 1100. As discussed above, with reference to FIG. 1, in various embodiments, the availability of such connectivity may make it feasible to disaggregate and dynamically pool resources such as accelerators, memory, and storage. In some embodiments, for example, one or more pooled accelerator sleds 1130 may be included among the physical infrastructure 1100A of data center 1100, each of which may comprise a pool of accelerator resources--such as co-processors and/or FPGAs, for example--that is globally accessible to other sleds via optical fabric 1112 and dual-mode optical switching infrastructure 1114.

[0043] In another example, in various embodiments, one or more pooled storage sleds 1132 may be included among the physical infrastructure 1100A of data center 1100, each of which may comprise a pool of storage resources that is available globally accessible to other sleds via optical fabric 1112 and dual-mode optical switching infrastructure 1114. In some embodiments, such pooled storage sleds 1132 may comprise pools of solid-state storage devices such as solid-state drives (SSDs). In various embodiments, one or more high-performance processing sleds 1134 may be included among the physical infrastructure 1100A of data center 1100. In some embodiments, high-performance processing sleds 1134 may comprise pools of high-performance processors, as well as cooling features that enhance air cooling to yield a higher thermal envelope of up to 250 W or more. In various embodiments, any given high-performance processing sled 1134 may feature an expansion connector 1117 that can accept a far memory expansion sled, such that the far memory that is locally available to that high-performance processing sled 1134 is disaggregated from the processors and near memory comprised on that sled. In some embodiments, such a high-performance processing sled 1134 may be configured with far memory using an expansion sled that comprises low-latency SSD storage. The optical infrastructure allows for compute resources on one sled to utilize remote accelerator/FPGA, memory, and/or SSD resources that are disaggregated on a sled located on the same rack or any other rack in the data center. The remote resources can be located one switch jump away or two-switch jumps away in the spine-leaf network architecture described above with reference to FIG. 5. The embodiments are not limited in this context.

[0044] In various embodiments, one or more layers of abstraction may be applied to the physical resources of physical infrastructure 1100A in order to define a virtual infrastructure, such as a software-defined infrastructure 1100B. In some embodiments, virtual computing resources 1136 of software-defined infrastructure 1100B may be allocated to support the provision of cloud services 1140. In various embodiments, particular sets of virtual computing resources 1136 may be grouped for provision to cloud services 1140 in the form of SDI services 1138. Examples of cloud services 1140 may include--without limitation--software as a service (SaaS) services 1142, platform as a service (PaaS) services 1144, and infrastructure as a service (IaaS) services 1146.

[0045] In some embodiments, management of software-defined infrastructure 1100B may be conducted using a virtual infrastructure management framework 1150B. In various embodiments, virtual infrastructure management framework 1150B may be designed to implement workload fingerprinting techniques and/or machine-learning techniques in conjunction with managing allocation of virtual computing resources 1136 and/or SDI services 1138 to cloud services 1140. In some embodiments, virtual infrastructure management framework 1150B may use/consult telemetry data in conjunction with performing such resource allocation. In various embodiments, an application/service management framework 1150C may be implemented in order to provide QoS management capabilities for cloud services 1140. The embodiments are not limited in this context.

[0046] As shown in FIG. 12, an illustrative system 1210 for managing the allocation of accelerator resources (e.g., physical accelerator resources 205-2) among a set of managed nodes 1260 includes an orchestrator server 1240 in communication with the set of managed nodes 1260. Each managed node 1260 may be embodied as an assembly of resources (e.g., physical resources 206), such as compute resources (e.g., physical compute resources 205-4), storage resources (e.g., physical storage resources 205-1), accelerator resources (e.g., physical accelerator resources 205-2), or other resources (e.g., physical memory resources 205-3) from the same or different sleds (e.g., the sleds 204-1, 204-2, 204-3, 204-4, etc.) or racks (e.g., one or more of racks 302-1 through 302-32). Each managed node 1260 may be established, defined, or "spun up" by the orchestrator server 1240 at the time a workload is to be assigned to the managed node 1260 or at any other time, and may exist regardless of whether any workloads are presently assigned to the managed node 1260. The system 1210 may be implemented in accordance with the data centers 100, 300, 400, 1100 described above with reference to FIGS. 1, 3, 4, and 11. In the illustrative embodiment, the set of managed nodes 1260 includes managed nodes 1250, 1252, and 1254. While three managed nodes 1260 are shown in the set, it should be understood that in other embodiments, the set may include a different number of managed nodes 1260 (e.g., tens of thousands). The system 1210 may be located in a data center and provide storage and compute services (e.g., cloud services) to a client device 1220 that is in communication with the system 1210 through a network 1230. The orchestrator server 1240 may support a cloud operating environment, such as OpenStack, and the managed nodes 1250 may execute one or more applications or processes (i.e., workloads), such as in virtual machines or containers, on behalf of a user of the client device 1220.

[0047] As discussed in more detail herein, the orchestrator server 1240, in operation, is configured to assign workloads to managed nodes 1260, receive telemetry data indicative of performance and conditions from the managed nodes 1260 as the workloads are performed, identify jobs within the workloads to be accelerated with one or more accelerator resources 205-2, provision (e.g., configure) the accelerator resources 205-2 to accelerate the identified jobs, and allocate the provisioned accelerator resources 205-2 to the managed nodes 1260 to accelerate the identified jobs. In the illustrative embodiment, the accelerator resources 205-2 include field programmable gate arrays (FPGAs) and the orchestrator server provisions the FPGAs by sending bitstreams indicative of desired configurations of the FPGAs to accelerate particular jobs. The orchestrator server 1240, in the illustrative embodiment, determines when the demand for acceleration for a particular job is likely to occur, based on evaluating the telemetry data and identifying patterns in the execution of the jobs, and sends the bitstreams to the FPGAs ahead of time, to provision the FPGAs in time to accelerate the jobs when the acceleration demand occurs. Additionally, the orchestrator server may receive resource allocation objective data indicative of one or more objectives to be achieved during the execution of the workloads. In the illustrative embodiment, the objectives pertain to power consumption, life expectancy, heat production, and/or performance of the resources allocated to the managed nodes 1260. As the workloads are executed, the orchestrator server 1240 may selectively allocate or deallocate the accelerator resources 205-2 to achieve the resource allocation objectives. In the illustrative embodiment, the achievement of an objective may be measured, equal to, or otherwise defined as the degree to which a measured value from one or more managed nodes 1260 satisfies a target value associated with the objective. For example, in the illustrative embodiment, increasing the achievement may be performed by decreasing the error (e.g., difference) between the measured value (e.g., a time taken to complete a workload or an operation in a workload) and the target value (e.g., a target time to complete the workload or operation in the workload). Conversely, decreasing the achievement may be performed by increasing the error (e.g., difference) between the measured value and the target value.

[0048] Referring now to FIG. 13, the orchestrator server 1240 may be embodied as any type of compute device capable of performing the functions described herein, including issuing a request to have cloud services performed, receiving results of the cloud services, assigning workloads to compute devices, analyzing telemetry data indicative of performance and conditions (e.g., resource utilization, one or more temperatures, fan speeds, etc.) as the workloads are executed, and managing the allocation of resources, including accelerator resources 205-2, across the managed nodes 1260 as the workloads are executed. For example, the orchestrator server 1240 may be embodied as a computer, a distributed computing system, one or more sleds (e.g., the sleds 204-1, 204-2, 204-3, 204-4, etc.), a server (e.g., stand-alone, rack-mounted, blade, etc.), a multiprocessor system, a network appliance (e.g., physical or virtual), a desktop computer, a workstation, a laptop computer, a notebook computer, a processor-based system, or a network appliance. As shown in FIG. 13, the illustrative orchestrator server 1240 includes a central processing unit (CPU) 1302, a main memory 1304, an input/output (I/O) subsystem 1306, communication circuitry 1308, and one or more data storage devices 1312. Of course, in other embodiments, the orchestrator server 1240 may include other or additional components, such as those commonly found in a computer (e.g., display, peripheral devices, etc.). Additionally, in some embodiments, one or more of the illustrative components may be incorporated in, or otherwise form a portion of, another component. For example, in some embodiments, the main memory 1304, or portions thereof, may be incorporated in the CPU 1302.

[0049] The CPU 1302 may be embodied as any type of processor capable of performing the functions described herein. The CPU 1302 may be embodied as a single or multi-core processor(s), a microcontroller, or other processor or processing/controlling circuit. In some embodiments, the CPU 1302 may be embodied as, include, or be coupled to a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), reconfigurable hardware or hardware circuitry, or other specialized hardware to facilitate performance of the functions described herein. Similarly, the main memory 1304 may be embodied as any type of volatile (e.g., dynamic random access memory (DRAM), etc.) or non-volatile memory or data storage capable of performing the functions described herein. In some embodiments, all or a portion of the main memory 1304 may be integrated into the CPU 1302. In operation, the main memory 1304 may store various software and data used during operation such as telemetry data, resource allocation objective data, workload labels, workload classifications, job data, resource allocation data, operating systems, applications, programs, libraries, and drivers.

[0050] The I/O subsystem 1306 may be embodied as circuitry and/or components to facilitate input/output operations with the CPU 1302, the main memory 1304, and other components of the orchestrator server 1240. For example, the I/O subsystem 1306 may be embodied as, or otherwise include, memory controller hubs, input/output control hubs, integrated sensor hubs, firmware devices, communication links (e.g., point-to-point links, bus links, wires, cables, light guides, printed circuit board traces, etc.), and/or other components and subsystems to facilitate the input/output operations. In some embodiments, the I/O subsystem 1306 may form a portion of a system-on-a-chip (SoC) and be incorporated, along with one or more of the CPU 1302, the main memory 1304, and other components of the orchestrator server 1240, on a single integrated circuit chip.

[0051] The communication circuitry 1308 may be embodied as any communication circuit, device, or collection thereof, capable of enabling communications over the network 1230 between the orchestrator server 1240 and another compute device (e.g., the client device 1220, and/or the managed nodes 1260). The communication circuitry 1308 may be configured to use any one or more communication technology (e.g., wired or wireless communications) and associated protocols (e.g., Ethernet, Bluetooth.RTM., Wi-Fi.RTM., WiMAX, etc.) to effect such communication.

[0052] The illustrative communication circuitry 1308 includes a network interface controller (NIC) 1310, which may also be referred to as a host fabric interface (HFI). The NIC 1310 may be embodied as one or more add-in-boards, daughtercards, network interface cards, controller chips, chipsets, or other devices that may be used by the orchestrator server 1240 to connect with another compute device (e.g., the client device 1220 and/or the managed nodes 1260). In some embodiments, the NIC 1310 may be embodied as part of a system-on-a-chip (SoC) that includes one or more processors, or included on a multichip package that also contains one or more processors. In some embodiments, the NIC 1310 may include a local processor (not shown) and/or a local memory (not shown) that are both local to the NIC 1310. In such embodiments, the local processor of the NIC 1310 may be capable of performing one or more of the functions of the CPU 1302 described herein. Additionally or alternatively, in such embodiments, the local memory of the NIC 1310 may be integrated into one or more components of the orchestrator server 1240 at the board level, socket level, chip level, and/or other levels.

[0053] The one or more illustrative data storage devices 1312, may be embodied as any type of devices configured for short-term or long-term storage of data such as, for example, memory devices and circuits, memory cards, hard disk drives, solid-state drives, or other data storage devices. Each data storage device 1312 may include a system partition that stores data and firmware code for the data storage device 1312. Each data storage device 1312 may also include an operating system partition that stores data files and executables for an operating system.

[0054] Additionally or alternatively, the orchestrator server 1240 may include one or more peripheral devices 1314. Such peripheral devices 1314 may include any type of peripheral device commonly found in a compute device such as a display, speakers, a mouse, a keyboard, and/or other input/output devices, interface devices, and/or other peripheral devices.

[0055] The client device 1220 and the managed nodes 1260 may have components similar to those described in FIG. 13. The description of those components of the orchestrator server 1240 is equally applicable to the description of components of the client device 1220 and the managed nodes 1260 and is not repeated herein for clarity of the description. Further, it should be appreciated that any of the client device 1220 and the managed nodes 1260 may include other components, sub-components, and devices commonly found in a computing device, which are not discussed above in reference to the orchestrator server 1240 and not discussed herein for clarity of the description. As discussed above, each managed node 1260 may include resources distributed across multiple sleds and in such embodiments, the CPU 1302, memory 1304, and/or communication circuitry 1308 may include portions thereof located on the same sled or different sled.

[0056] As described above, the client device 1220, the orchestrator server 1240, and the managed nodes 1260 are illustratively in communication via the network 1230, which may be embodied as any type of wired or wireless communication network, including global networks (e.g., the Internet), local area networks (LANs) or wide area networks (WANs), cellular networks (e.g., Global System for Mobile Communications (GSM), 3G, Long Term Evolution (LTE), Worldwide Interoperability for Microwave Access (WiMAX), etc.), digital subscriber line (DSL) networks, cable networks (e.g., coaxial networks, fiber networks, etc.), or any combination thereof.

[0057] Referring now to FIG. 14, in the illustrative embodiment, the orchestrator server 1240 may establish an environment 1400 during operation. The illustrative environment 1400 includes a network communicator 1420, a telemetry monitor 1430, and a resource manager 1440. Each of the components of the environment 1400 may be embodied as hardware, firmware, software, or a combination thereof. As such, in some embodiments, one or more of the components of the environment 1400 may be embodied as circuitry or a collection of electrical devices (e.g., network communicator circuitry 1420, telemetry monitor circuitry 1430, resource manager circuitry 1440, etc.). It should be appreciated that, in such embodiments, one or more of the network communicator circuitry 1420, telemetry monitor circuitry 1430, or resource manager circuitry 1440 may form a portion of one or more of the CPU 1302, the main memory 1304, the I/O subsystem 1306, and/or other components of the orchestrator server 1240. In the illustrative embodiment, the environment 1400 includes telemetry data 1402 which may be embodied as data indicative of the performance and conditions (e.g., resource utilization, operating frequencies, power usage, one or more temperatures, fan speeds, etc.) of resources allocated to each managed node 1260 and individual jobs (e.g., set of functions) of the workloads that are performed as the managed nodes 1260 execute the workloads assigned to them. Additionally, the illustrative environment 1400 includes resource allocation objective data 1404 indicative of user-defined thresholds or goals ("objectives") to be satisfied during the execution of the workloads. In the illustrative embodiment, the objectives pertain to power consumption, life expectancy, heat production, and performance of the resources allocated to the managed nodes 1260. Further, the illustrative environment 1400 includes workload labels 1406 which may be embodied as any identifiers (e.g., process numbers, executable file names, alphanumeric tags, etc.) that uniquely identify each workload executed by the managed nodes 1260.

[0058] Additionally, the illustrative environment 1400 includes workload classifications 1408 which may be embodied as any data indicative of the general resource utilization tendencies of each workload (e.g., processor intensive, memory intensive, network bandwidth intensive, etc.). Further, the illustrative environment 1400 includes job data 1410 indicative of jobs (e.g., sets of functions) within each workload that may be accelerated. In the illustrative embodiment, the job data 1410 is embodied as a queue of jobs to be processed, an indication of the types of functions within the job (e.g., compression, encryption, matrix operations, etc.), information about the format and size of input data used by the job (e.g., number of bytes, whether the input data is formatted as a matrix or otherwise, an encoding scheme for the input data, etc.), a globally unique identifier (GUID) associated with each job, counters indicative of how many times a particular job has been in the queue within a predefined time frame for each workload and across all workloads executed in the data center 1100, the average amount of time each job resides in the queue, and/or other characteristics of the jobs. Additionally, the illustrative embodiment 1400 includes resource allocation data 1412 indicative of the resources, including accelerator resources 205-2, within the data center 1100 that have been allocated to each managed node 1260 at any given time.

[0059] In the illustrative environment 1400, the network communicator 1420, which may be embodied as hardware, firmware, software, virtualized hardware, emulated architecture, and/or a combination thereof as discussed above, is configured to facilitate inbound and outbound network communications (e.g., network traffic, network packets, network flows, etc.) to and from the orchestrator server 1240, respectively. To do so, the network communicator 1420 is configured to receive and process data packets from one system or computing device (e.g., the client device 1220) and to prepare and send data packets to another computing device or system (e.g., the managed nodes 1260). Accordingly, in some embodiments, at least a portion of the functionality of the network communicator 1420 may be performed by the communication circuitry 1308, and, in the illustrative embodiment, by the NIC 1310.

[0060] The telemetry monitor 1430, which may be embodied as hardware, firmware, software, virtualized hardware, emulated architecture, and/or a combination thereof as discussed above, is configured to collect the telemetry data 1402 from the managed nodes 1260 as the managed nodes 1260 execute the workloads assigned to them. The telemetry monitor 1430 may actively poll each of the managed nodes 1260 for updated telemetry data 1402 on an ongoing basis or may passively receive telemetry data 1402 from the managed nodes 1260, such as by listening on a particular network port for updated telemetry data 1402. The telemetry monitor 1430 may further parse and categorize the telemetry data 1402, such as by separating the telemetry data 1402 into an individual file or data set for each managed node 1260.

[0061] The resource manager 1440, which may be embodied as hardware, firmware, software, virtualized hardware, emulated architecture, and/or a combination thereof, is configured to assign workloads to managed nodes, identify jobs within the workloads to accelerate, predict when acceleration demand will occur within the workloads, provision (e.g., configure) accelerator resources 205-2 in advance of the predicted acceleration demand, and adjust the allocation of accelerator resources 205-2 to and from the managed nodes 1260 on an ongoing basis to improve the efficiency of workload execution and/or satisfy other resource allocation objectives (e.g. from the resource allocation objective data 1404).

[0062] To do so, the resource manager 1440 includes a workload labeler 1442, a workload classifier 1444, a workload behavior predictor 1446, an acceleration manager 1448, and a multi-objective analyzer 1450. The workload labeler 1442, in the illustrative embodiment, is configured to assign a workload label 1406 to each workload presently performed or scheduled to be performed by the managed nodes 1260. The workload labeler 1442 may generate the workload label 1406 as a function of an executable name of the workload, a hash of all or a portion of the code of the workload, or based on any other method to uniquely identify each workload. The workload classifier 1444, in the illustrative embodiment, is configured to categorize each labeled workload based on the average resource utilization of each workload (e.g., generally utilizes 65% of processor capacity, generally utilizes 40% of memory capacity, etc.).

[0063] The workload behavior predictor 1446, in the illustrative embodiment, is configured to analyze the telemetry data 1402 to identify different phases of resource utilization within the telemetry data 1402 for each workload. Each resource utilization phase may be embodied as a period of time in which the resource utilization of one or more resources allocated to a managed node 1260 satisfies a predefined threshold. For example, a utilization of at least 85% of the allocated processor capacity may be indicative of a high processor utilization phase, and a utilization of at least 85% of the allocated memory capacity may be indicative of a high memory utilization phase. In the illustrative embodiment, the workload behavior predictor 1446 is further to identify patterns in the resource utilization phases of the workloads (e.g., a high processor utilization phase, followed by a high memory utilization phase, followed by a phase of low resource utilization, which is then followed by the high processor utilization phase again). The workload behavior predictor 1446 may be configured to utilize the identifications of the resource utilization phase patterns, determine a present resource utilization phase of a given workload, predict the next resource utilization phase based on the patterns, and determine an amount of remaining time until the workload transitions to the next resource utilization phase.

[0064] The acceleration manager 1448, in the illustrative embodiment, is configured to identify, generate, from the telemetry data 1402, the job data 1410, identify jobs within the workloads to be accelerated, based on their types, residency time in the job queue, how often the jobs are executed, and other factors, coordinate selecting and provisioning accelerator resources 205-2, such as FPGAs, available within the data center 1100, and manage the timing of the allocation and/or deallocation of the accelerator resources 205-2 to coincide with predicting times when the jobs to be accelerated are likely to be initiated (e.g., called) by the workloads.

[0065] The multi-objective analyzer 1450, in the illustrative embodiment, is configured to whether an efficiency objective and/or other resource allocation objective data 1404 is being met during the execution of workloads, and, determine adjustments to the allocation of resources among the managed nodes 1260 to enable the one more objectives to be satisfied. As such, with regard to the allocation of accelerator resources 205-2, the multi-objective analyzer 1450 coordinates with the acceleration manager 1448 to determine which accelerator resources 205-2 to allocate to which managed nodes 1260 and at what time. In the illustrative embodiment, the multi-objective analyzer 1450 may include a model of the data center 1100 that simulates the expected effects, including power consumption, heat generation, changes to compute capacity, and other factors, in response to various adjustments to the allocations of resources among the managed nodes 1260 and/or the settings of components (e.g., increasing or decreasing clock speeds, enabling or disabling support for extended instruction sets, etc.) within the resources. To do so, in the illustrative embodiment, the multi-objective analyzer 1450 includes a resource allocator 1452 and a resource settings adjuster 1454. The resource allocator 1452, in the illustrative embodiment, is configured to issue instructions to the managed nodes 1260 to allocate or deallocate resources as determined by the multi-objective analyzer 1450 and the acceleration manager 1448, and to update the resource allocation data 1412 to indicate the present state of allocation of the resources among the managed nodes 1260. Similarly, the resource settings adjuster 1454, in the illustrative embodiment, is configured issue instructions to one or more of the managed nodes 1260 to adjust settings of resources allocated to the managed nodes 1260, such as by adjusting a firmware setting to increase or decrease a clock speed of a processor, increasing or decreasing power utilization settings, and/or other settings that affect the operation of the resources.

[0066] It should be appreciated that each of the workload labeler 1442, the workload classifier 1444, the workload behavior predictor 1446, the acceleration manager 1448, the multi-objective analyzer 1450, the resource allocator 1452, and the resource settings adjuster 1454 may be separately embodied as hardware, firmware, software, virtualized hardware, emulated architecture, and/or a combination thereof. For example, the workload labeler 1442 may be embodied as a hardware component, while the workload classifier 1444, the workload behavior predictor 1446, the acceleration manager 1448, the multi-objective analyzer 1450, the resource allocator 1452, and the resource settings adjuster 1454 are embodied as virtualized hardware components or as some other combination of hardware, firmware, software, virtualized hardware, emulated architecture, and/or a combination thereof.

[0067] Referring now to FIG. 15, in use, the orchestrator server 1240 may execute a method 1500 for managing the allocation of accelerator resources 205-2 among the managed nodes 1260 as the managed nodes 1260 execute workloads. The method 1500 begins with block 1502, in which the orchestrator server 1240 determines whether to manage the allocation of resources among the managed nodes 1260. In the illustrative embodiment, the orchestrator server 1240 determines to manage the allocation of resources if the orchestrator server 1240 is powered on, in communication with the managed nodes 1260, and has received at least one request from the client device 1220 to provide cloud services (i.e., to perform one or more workloads). In other embodiments, the orchestrator server 1240 may determine whether to manage the allocation of resources based on other factors. Regardless, in response to a determination to manage the allocation of resources, in the illustrative embodiment, the method 1500 advances to block 1504 in which the orchestrator server 1240 may obtain resource allocation objective data (e.g., the resource allocation objective data 1404). In doing so, the orchestrator server 1240 may obtain the resource allocation objective data 1404 from a user (e.g., an administrator) through a graphical user interface (not shown), from a configuration file, or from another source. The orchestrator server 1240, in the illustrative embodiment, may obtain performance objective data, indicative of a target speed at which workloads are to be executed (e.g., a target time period in which to complete execution of a workload, a target number of operations per second, etc.), as indicated in block 1506. In receiving the resource allocation objective data 1404, the orchestrator server 1240 may receive power consumption objective data indicative of a target power usage or threshold amount of power usage of the resource allocated to each managed node 1260 as they execute the workloads, as indicated in block 1508. Additionally or alternatively, the orchestrator server 1240 may receive reliability objective data indicative of a target life cycle of one or more resources (e.g., a target life cycle of a data storage device, a target life cycle of a cooling fan, etc.), as indicated in block 1510. As indicated in block 1512, the orchestrator server 1240 may also receive thermal objective data indicative of one or more target temperatures of one or more resources (e.g., one or more CPUs 1302, etc.).

[0068] In block 1514, in the illustrative embodiment, the orchestrator server 1240 allocates resources to the managed nodes 1260. Initially, the orchestrator server 1240 has not received any telemetry data 1402 to inform a decision as to which resources to allocate to the various managed nodes 1260. As such, as indicated in block 1516, the orchestrator server 1240 may initially allocate no accelerator resources 205-2 to any of the managed nodes 1260. Alternatively, as indicated in block 1518, the orchestrator server 1240 may assign accelerator resources 205-2 among the managed nodes 1260 according to a default scheme (e.g., dividing the accelerator resources 205-2 among the managed nodes 1260 evenly, allocating a predefined number of accelerator resources 205-2 to each managed node 1260 as the managed nodes 1260 are defined until no more available accelerator resources 205-2 are available, etc.). In doing so, the orchestrator server 1240 may defer allocating any FPGAs to the managed nodes 1260 until after the workloads have been assigned and the FPGAs have been provisioned (e.g., configured) to perform one or more jobs to be accelerated, as described in more detail herein. In block 1520, the orchestrator server 1240 assigns workloads to the managed nodes 1260 for execution and, as indicated in block 1522, begins receiving the telemetry data 1402 as the workloads are executed by the managed nodes 1260. Subsequently, the method 1500 advances to block 1524 of FIG. 16 in which the orchestrator server 1240 determines, from the telemetry data 1402, predicted demand for acceleration (e.g., for which accelerator resources 205-2 should be allocated) as explained in more detail herein.

[0069] Referring now to FIG. 16, in determining, from the telemetry data 1402, the predicted demand for acceleration, the orchestrator server 1240 may identify jobs within the assigned workloads for acceleration, as indicated in block 1526. As indicated in block 1528, in the illustrative embodiment, the orchestrator server 1240 may analyze a job queue (e.g., the job data 1410) to identify jobs within the assigned workloads for acceleration. In doing so, the orchestrator server 1240 may determine an average amount of time each job resides in the queue (e.g., before being completed), as indicated in block 1530. As indicated in block 1532, the orchestrator serer 1240 may apply a smoothing algorithm such as an exponential smoothing algorithm, to one or more times indicated by the job queue to determine the average amount of time each job resides in the job queue. As indicated in block 1534, the orchestrator server 1240 may determine local counts and global counts of jobs executed, and compare the local and global counts to one or more threshold count values. For example, the orchestrator server 1240 may maintain a count of how many times each job has been performed for each workload (e.g., a local count) as well as a count of how many times each job, regardless of the particular workload or managed node 1260 associated with it, has been performed. If either of the counts satisfies (e.g., is equal to or exceeds) a predefined threshold value, the orchestrator server 1240 may identify the corresponding job as one that should be accelerated (e.g., executed with one or more accelerator resources 205-2).

[0070] Still referring to FIG. 16, in determining the predicted demand for acceleration, the orchestrator server 1240 may additionally identify characteristics of the jobs being executed, as indicated in block 1536. In doing so, the orchestrator server 1240 may determine whether each job is amenable to acceleration (e.g., whether an accelerator resource could execute the job faster or more efficiently than a general purpose processor 1302). In doing so, as indicated in block 1540, the orchestrator server 1240 may determine the type of each job, such as by analyzing and classifying the types of functions as indicative of certain types of operations (e.g., compression operations, encryption operations, etc.). As indicated in block 1542, the orchestrator server 1240 may determine characteristics of the input data used by the jobs, such as whether the input data is formatted as a matrix of values or in another format, the size (e.g., in bytes) of the input data, and/or characteristics of the input data. As described above, the analysis may be performed on the job data 1410 which, in the illustrative embodiment, is generated from the telemetry data 1402 reported by the managed nodes 1260. As such, in the illustrative embodiment, the managed nodes 1260 may be configured to provide information indicative of the types of functions within each job and the input data characteristics for each job. As indicated in block 1544, in the process of identifying the characteristics of the jobs, the orchestrator server 1240 may assign a globally unique identifier (e.g., a number, tag, alphanumeric sequence, or other identifier that is unique) to each job. The globally unique identifier may be generated from an identifier for each job reported from each managed node 1260 in the managed node's 1260 corresponding telemetry data 1402, such as by appending a hash of the workload label and a unique identifier of the managed node 1260 to the identifier of the corresponding job indicated in the telemetry data 1402 from the managed node 1260. In block 1546, the orchestrator server 1240 may determine a predicted time of the demand for acceleration (e.g., when the demand will likely occur). As indicated in block 1548, the orchestrator server 1240 may determine the predicted time of the demand by analyzing a pattern of the job executions for each workload (e.g., job A resides in the job queue for 10 seconds, followed by job B, which resides in the job queue for 15 seconds, followed again by job A). Afterwards, the method 1500 advances to block 1550 of FIG. 17 in which the orchestrator server 1240 provisions, prior to the predicted demand for acceleration, one or more accelerator resources 205-2 to accelerate the jobs within the workloads.

[0071] Referring now to FIG. 17, in provisioning the accelerator resources 205-2, the orchestrator server 1240, in the illustrative embodiment, selects one or more field programmable gate arrays to provision, as indicated in block 1552. As indicated in block 1554, the orchestrator server 1240, in the illustrative embodiment, prefers FPGAs that are already configured (e.g., provisioned) to perform a given job that is to be accelerated in the future. By preferring (e.g., selecting over other FPGAs) FPGAs that are already provisioned to perform the job to be accelerated, the orchestrator server 1240 may save time that would otherwise be consumed to provide a bitstream indicative of the desired configuration to the FPGA and wait for the FPGA to configure its field programmable gates according to the desired configuration. The orchestrator server 1240 may store, in the resource allocation data 1412, information indicative of which FPGAs have been provisioned to perform which jobs. In block 1556, the orchestrator server 1240, in the illustrative embodiment, determines the number of FPGAs to provision, such as by counting the number of jobs that have been identified for acceleration, determining the number of available FPGAs, and determining to use one available FPGA for each job or up to the number of available FPGAs, if the number of available FPGAs is less than the number of jobs to accelerate. In block 1558, the orchestrator server 1240 may select FPGAs on sleds (e.g., accelerator sled 1130) that are different form the sleds on which the workloads are executed by general purpose processors (e.g., compute sled 204-4). As indicated in block 1560, the orchestrator server 1240 may select FPGAs as a function of a target heat generation, a target power consumption, and/or a target economic cost. For example, some FPGAs may be more efficient in terms of heat generation and/or power consumption than other FPGAs, because they are composed of smaller or otherwise more efficient components. As such, the cost of cooling and powering less efficient FPGAs may be greater than cooling and powering other FPGAs. In block 1562, the orchestrator server 1240 may determine a configuration time for each FPGA (e.g., the amount of time that will elapse to configure the FPGA to perform a job). Initially, the orchestrator server 1240 may not have access to data indicative of the amount of time required to provision a particular FPGA and may instead use a default estimated time (e.g., two minutes). If and when the orchestrator server 1240 does provision the FPGA, the orchestrator server 1240 may measure the actual amount of time that elapses to provision the FPGA and refer to that measured time in later determinations.

[0072] In block 1564, the orchestrator server 1240 provides (e.g., sends) a bitstream indicative of a desired configuration of each FPGA to each FPGA to be provisioned. The bitstream may include a portion specific to the architecture of the particular FPGA (e.g., to initialize the FPGA for configuration) and another portion indicative of the desired configuration of the gates within the FPGA to perform the corresponding job to be accelerated. In providing the bitstreams, in the illustrative embodiment and as indicated in block 1566, the orchestrator server 1240 provides the bitstreams in advance of the predicted time (e.g., the time predicted in block 1546 of FIG. 16) that the job to be accelerated is scheduled to be executed (e.g., in advance of the time of the predicted demand) by the determined configuration time for the corresponding FPGA. For example, if the configuration time for the FPGA is two minutes, in the illustrative embodiment, the orchestrator server 1240 sends the bitstream to the FPGA at least two minutes before the corresponding job is to be executed (e.g., two minutes before the job enters the job queue).

[0073] Afterwards, the method 1500 advances to block 1568 in which the orchestrator server 1240 allocates the accelerator resources 205-2 to the managed nodes 1260 to accelerate execution of the workloads (e.g., the workload jobs that were identified for acceleration in block 1526 of FIG. 16). In block 1570, the orchestrator server 1240, in the illustrative embodiment, allocates the provisioned FPGAs from block 1550 to the managed nodes 1260 associated with the jobs identified for acceleration. The orchestrator server 1240 may do so by providing each managed node 1260 with address information for the corresponding FPGAs to enable the managed nodes 1260 to communicate with the FPGAs. As indicated in block 1572, the orchestrator server 1240 may allocated other accelerator resources 205-2 (e.g., graphics accelerators, etc.) to the managed nodes 1260, such as if one or more jobs are not suitable for acceleration by an FPGA, as determined in blocks 1538 through 1542 of FIG. 16, or if the set of available FPGAs in the data center 1100 has been depleted. In block 1574, the orchestrator server 1240 may deallocate one more accelerator resources 205-2 from one or more managed nodes 1260 (e.g., if the corresponding jobs have completed), thereby replenishing the set of available accelerator resources 205-2. As indicated in block 1576, in allocating and/or deallocating the accelerator resources 205-2, the orchestrator server 1240, in the illustrative embodiment, does so to satisfy the one or more resource allocation objectives (e.g., objectives in the resource allocation objective data 1404). For example, if accelerating a particular job would increase performance beyond a target resource allocation objective (e.g. a number of operations per second) and would cause heat generation in excess of a target temperature in one or more areas of the data center 1100, the orchestrator server 1240 may determine not to allocate an accelerator resource 205-2 to that job. In some embodiment, the orchestrator server 1240 may determine whether to ultimately allocate an accelerator resource 205-2 to accelerate a particular job in view of the resource allocation objectives prior to the provisioning operations in block 1550. Subsequently, the method 1500 returns to block 1522 of FIG. 15 in which the orchestrator server 1240 continues collecting telemetry data 1402 as the workloads are executed.

EXAMPLES

[0074] Illustrative examples of the technologies disclosed herein are provided below. An embodiment of the technologies may include any one or more, and any combination of, the examples described below.

[0075] Example 1 includes an orchestrator server to dynamically manage the allocation of accelerator resources, the orchestrator server comprising one or more processors; one or more memory devices having stored therein a plurality of instructions that, when executed by the one or more processors, cause the orchestrator server to assign a workload to a managed node for execution; determine a predicted demand for one or more accelerator resources to accelerate the execution of one or more jobs within the workload; provision, prior to the predicted demand, one or more accelerator resources to accelerate the one or more jobs; and allocate the one or more provisioned accelerator resources to the managed node to accelerate the execution of the one or more jobs.

[0076] Example 2 includes the subject matter of Example 1, and wherein to determine the predicted demand comprises to determine a demand for one or more field programmable gate arrays (FPGAs).

[0077] Example 3 includes the subject matter of any of Examples 1 and 2, and wherein to provision the one or more accelerator resources comprises to provide, to the one or more FPGAs, a bit stream indicative of a configuration of each FPGA to accelerate execution of the one or more jobs.

[0078] Example 4 includes the subject matter of any of Examples 1-3, and wherein to determine the predicted demand comprises to determine the number of accelerator resources to allocate to satisfy the predicted demand.

[0079] Example 5 includes the subject matter of any of Examples 1-4, and wherein to provision the one or more accelerator resources comprises to provision one or more accelerator resources located on one or more sleds that are different than a sled on which the workload is presently executed.

[0080] Example 6 includes the subject matter of any of Examples 1-5, and wherein the plurality of instructions, when executed, further cause the orchestrator server to determine a configuration time period to provision each of the one or more accelerator resources; and determine a predicted time of the predicted demand; and wherein to provision the one or more accelerator resources comprises to begin configuration of the one or more accelerator resources for accelerated execution of the one or more jobs at a time that is earlier than the predicted time by at least the configuration time period.

[0081] Example 7 includes the subject matter of any of Examples 1-6, and wherein the plurality of instructions, when executed, further cause the orchestrator server to identify one or more jobs within the workload to be accelerated with one or more field programmable gate arrays (FPGAs); associate each identified job with a globally unique identifier indicative of one or more of a specific interface of the job or a definition of the job.

[0082] Example 8 includes the subject matter of any of Examples 1-7, and wherein to associate each identified job with a globally unique identifier comprises to associate each identified job with a globally unique identifier indicative of one or more of a size of an input or a format of an input to the job.

[0083] Example 9 includes the subject matter of any of Examples 1-8, and wherein the managed node is one of a plurality of managed nodes and the workload is one of a plurality of workloads executed by the managed nodes and the plurality of instructions, when executed, further cause the orchestrator server to determine, for each workload, a local count indicative of a number of times a job is executed in each workload; determine a global count indicative of a number of times a job is executed by all of the managed nodes; determine whether one or more of the local count or the global count satisfies a threshold count value; and identify, in response to a determination that one or more of the local count or the global count satisfies the threshold count value, the associated job as a job to be accelerated.

[0084] Example 10 includes the subject matter of any of Examples 1-9, and wherein the plurality of instructions, when executed, further cause the orchestrator server to identify, from a plurality of accelerator resources, the one or more accelerator resources to accelerate the one or more jobs.

[0085] Example 11 includes the subject matter of any of Examples 1-10, and wherein to identify the one or more accelerator resources comprises to determine whether one or more of the accelerator resources is already configured to perform one or more of the jobs; and select, in response to a determination that one or more the accelerator resources is already configured to perform one or more of the jobs, the one or more already-configured accelerator resources for acceleration of the one or more jobs.

[0086] Example 12 includes the subject matter of any of Examples 1-11, and wherein to identify the one or more accelerator resources comprises to select the one or more accelerator resources as a function of one or more of a target heat generation, a target power usage, or a target economic cost of utilization of the one or more accelerator resources.

[0087] Example 13 includes the subject matter of any of Examples 1-12, and wherein the managed node is one of a plurality of managed nodes and the workload is one of a plurality of workloads executed by the managed nodes, and wherein to determine the demand comprises to establish a job queue indicative of all jobs for all of the workloads to be performed; determine an average time period in which each job resides in the job queue; and determine the demand for each job as a function of the average time period for each job.

[0088] Example 14 includes the subject matter of any of Examples 1-13, and wherein to determine the demand for each job further comprises to apply an exponential averaging algorithm to the time period in which each job resides in the job queue.

[0089] Example 15 includes a method for dynamically managing the allocation of accelerator resources, the method comprising assigning, by an orchestrator server, a workload to a managed node for execution; determining, by the orchestrator server, a predicted demand for one or more accelerator resources to accelerate the execution of one or more jobs within the workload; provisioning, by the orchestrator server and prior to the predicted demand, one or more accelerator resources to accelerate the one or more jobs; and allocating, by the orchestrator server, the one or more provisioned accelerator resources to the managed node to accelerate the execution of the one or more jobs.

[0090] Example 16 includes the subject matter of Example 15, and wherein determining the predicted demand comprises determining a demand for one or more field programmable gate arrays (FPGAs).

[0091] Example 17 includes the subject matter of any of Examples 15 and 16, and wherein provisioning the one or more accelerator resources comprises providing, to the one or more FPGAs, a bit stream indicative of a configuration of each FPGA to accelerate execution of the one or more jobs.

[0092] Example 18 includes the subject matter of any of Examples 15-17, and wherein determining the predicted demand comprises determining the number of accelerator resources to allocate to satisfy the predicted demand.

[0093] Example 19 includes the subject matter of any of Examples 15-18, and wherein provisioning the one or more accelerator resources comprises provisioning one or more accelerator resources located on one or more sleds that are different than a sled on which the workload is presently executed.

[0094] Example 20 includes the subject matter of any of Examples 15-19, and further including determining, by the orchestrator server, a configuration time period to provision each of the one or more accelerator resources; and determining, by the orchestrator server, a predicted time of the predicted demand; and wherein provisioning the one or more accelerator resources comprises beginning configuration of the one or more accelerator resources for accelerated execution of the one or more jobs at a time that is earlier than the predicted time by at least the configuration time period.

[0095] Example 21 includes the subject matter of any of Examples 15-20, and further including identifying, by the orchestrator server, one or more jobs within the workload to be accelerated with one or more field programmable gate arrays (FPGAs); and associating, by the orchestrator server, each identified job with a globally unique identifier indicative of one or more of a specific interface of the job or a definition of the job.

[0096] Example 22 includes the subject matter of any of Examples 15-21, and wherein associating each identified job with a globally unique identifier comprises associating each identified job with a globally unique identifier indicative of one or more of a size of an input or a format of an input to the job.

[0097] Example 23 includes the subject matter of any of Examples 15-22, and wherein the managed node is one of a plurality of managed nodes and the workload is one of a plurality of workloads executed by the managed nodes, the method further comprising determining, by the orchestrator server and for each workload, a local count indicative of a number of times a job is executed in each workload; determining, by the orchestrator server, a global count indicative of a number of times a job is executed by all of the managed nodes; determining, by the orchestrator server, whether one or more of the local count or the global count satisfies a threshold count value; and identifying, by the orchestrator server and in response to a determination that one or more of the local count or the global count satisfies the threshold count value, the associated job as a job to be accelerated.

[0098] Example 24 includes the subject matter of any of Examples 15-23, and further including identifying, by the orchestrator server and from a plurality of accelerator resources, the one or more accelerator resources to accelerate the one or more jobs.

[0099] Example 25 includes the subject matter of any of Examples 15-24, and wherein identifying the one or more accelerator resources comprises determining whether one or more of the accelerator resources is already configured to perform one or more of the jobs, the method further comprising selecting, by the orchestrator server in response to a determination that one or more the accelerator resources is already configured to perform one or more of the jobs, the one or more already-configured accelerator resources for acceleration of the one or more jobs.

[0100] Example 26 includes the subject matter of any of Examples 15-25, and wherein identifying the one or more accelerator resources comprises selecting the one or more accelerator resources as a function of one or more of a target heat generation, a target power usage, or a target economic cost of utilization of the one or more accelerator resources.

[0101] Example 27 includes the subject matter of any of Examples 15-26, and wherein the managed node is one of a plurality of managed nodes and the workload is one of a plurality of workloads executed by the managed nodes, and wherein determining the demand comprises establishing a job queue indicative of all jobs for all of the workloads to be performed; determining an average time period in which each job resides in the job queue; and determining the demand for each job as a function of the average time period for each job.

[0102] Example 28 includes the subject matter of any of Examples 15-27, and wherein determining the demand for each job further comprises applying an exponential averaging algorithm to the time period in which each job resides in the job queue.

[0103] Example 29 includes an orchestrator server comprising means for performing the method of any of Examples 15-28.

[0104] Example 30 includes an orchestrator server to dynamically manage the allocation of accelerator resources, the orchestrator server comprising one or more processors; one or more memory devices having stored therein a plurality of instructions that, when executed by the one or more processors, cause the orchestrator server to perform the method of any of Examples 15-28.

[0105] Example 31 includes one or more machine-readable storage media comprising a plurality of instructions stored thereon that, in response to being executed, cause an orchestrator server to perform the method of any of Examples 15-28.

[0106] Example 32 includes an orchestrator server to dynamically manage the allocation of accelerator resources, the orchestrator server comprising resource manager circuitry to assign a workload to a managed node for execution, determine a predicted demand for one or more accelerator resources to accelerate the execution of one or more jobs within the workload, provision, prior to the predicted demand, one or more accelerator resources to accelerate the one or more jobs, and allocate the one or more provisioned accelerator resources to the managed node to accelerate the execution of the one or more jobs.

[0107] Example 33 includes the subject matter of Example 32, and wherein to determine the predicted demand comprises to determine a demand for one or more field programmable gate arrays (FPGAs).

[0108] Example 34 includes the subject matter of any of Examples 32 and 33, and wherein to provision the one or more accelerator resources comprises to provide, to the one or more FPGAs, a bit stream indicative of a configuration of each FPGA to accelerate execution of the one or more jobs.

[0109] Example 35 includes the subject matter of any of Examples 32-34, and wherein to determine the predicted demand comprises to determine the number of accelerator resources to allocate to satisfy the predicted demand.

[0110] Example 36 includes the subject matter of any of Examples 32-35, and wherein to provision the one or more accelerator resources comprises to provision one or more accelerator resources located on one or more sleds that are different than a sled on which the workload is presently executed.

[0111] Example 37 includes the subject matter of any of Examples 32-36, and wherein the resource manager circuitry is further to determine a configuration time period to provision each of the one or more accelerator resources; and determine a predicted time of the predicted demand; and wherein to provision the one or more accelerator resources comprises to begin configuration of the one or more accelerator resources for accelerated execution of the one or more jobs at a time that is earlier than the predicted time by at least the configuration time period.

[0112] Example 38 includes the subject matter of any of Examples 32-37, and wherein resource manager circuitry is further to identify one or more jobs within the workload to be accelerated with one or more field programmable gate arrays (FPGAs); associate each identified job with a globally unique identifier indicative of one or more of a specific interface of the job or a definition of the job.

[0113] Example 39 includes the subject matter of any of Examples 32-38, and wherein to associate each identified job with a globally unique identifier comprises to associate each identified job with a globally unique identifier indicative of one or more of a size of an input or a format of an input to the job.

[0114] Example 40 includes the subject matter of any of Examples 32-39, and wherein the managed node is one of a plurality of managed nodes and the workload is one of a plurality of workloads executed by the managed nodes and the resource manager circuitry is further to determine, for each workload, a local count indicative of a number of times a job is executed in each workload; determine a global count indicative of a number of times a job is executed by all of the managed nodes; determine whether one or more of the local count or the global count satisfies a threshold count value; and identify, in response to a determination that one or more of the local count or the global count satisfies the threshold count value, the associated job as a job to be accelerated.

[0115] Example 41 includes the subject matter of any of Examples 32-40, and wherein the resource manager circuitry is further to identify, from a plurality of accelerator resources, the one or more accelerator resources to accelerate the one or more jobs.

[0116] Example 42 includes the subject matter of any of Examples 32-41, and wherein to identify the one or more accelerator resources comprises to determine whether one or more of the accelerator resources is already configured to perform one or more of the jobs; and select, in response to a determination that one or more the accelerator resources is already configured to perform one or more of the jobs, the one or more already-configured accelerator resources for acceleration of the one or more jobs.

[0117] Example 43 includes the subject matter of any of Examples 32-42, and wherein to identify the one or more accelerator resources comprises to select the one or more accelerator resources as a function of one or more of a target heat generation, a target power usage, or a target economic cost of utilization of the one or more accelerator resources.

[0118] Example 44 includes the subject matter of any of Examples 32-43, and wherein the managed node is one of a plurality of managed nodes and the workload is one of a plurality of workloads executed by the managed nodes, and wherein to determine the demand comprises to establish a job queue indicative of all jobs for all of the workloads to be performed; determine an average time period in which each job resides in the job queue; and determine the demand for each job as a function of the average time period for each job.

[0119] Example 45 includes the subject matter of any of Examples 32-44, and wherein to determine the demand for each job further comprises to apply an exponential averaging algorithm to the time period in which each job resides in the job queue.

[0120] Example 46 includes an orchestrator server to dynamically manage the allocation of accelerator resources, the orchestrator server comprising circuitry for assigning a workload to a managed node for execution; means for determining a predicted demand for one or more accelerator resources to accelerate the execution of one or more jobs within the workload; circuitry for provisioning, by the orchestrator server and prior to the predicted demand, one or more accelerator resources to accelerate the one or more jobs; and circuitry for allocating the one or more provisioned accelerator resources to the managed node to accelerate the execution of the one or more jobs.

[0121] Example 47 includes the subject matter of Example 46, and wherein the means for determining the predicted demand comprises means for determining a demand for one or more field programmable gate arrays (FPGAs).

[0122] Example 48 includes the subject matter of any of Examples 46 and 47, and wherein the circuitry for provisioning the one or more accelerator resources comprises circuitry for providing, to the one or more FPGAs, a bit stream indicative of a configuration of each FPGA to accelerate execution of the one or more jobs.

[0123] Example 49 includes the subject matter of any of Examples 46-48, and wherein the means for determining the predicted demand comprises means for determining the number of accelerator resources to allocate to satisfy the predicted demand.

[0124] Example 50 includes the subject matter of any of Examples 46-49, and wherein the circuitry for provisioning the one or more accelerator resources comprises circuitry for provisioning one or more accelerator resources located on one or more sleds that are different than a sled on which the workload is presently executed.

[0125] Example 51 includes the subject matter of any of Examples 46-50, and further including circuitry for determining a configuration time period to provision each of the one or more accelerator resources; and means for determining a predicted time of the predicted demand; and wherein the circuitry for provisioning the one or more accelerator resources comprises circuitry for beginning configuration of the one or more accelerator resources for accelerated execution of the one or more jobs at a time that is earlier than the predicted time by at least the configuration time period.

[0126] Example 52 includes the subject matter of any of Examples 46-51, and further including means for identifying one or more jobs within the workload to be accelerated with one or more field programmable gate arrays (FPGAs); and circuitry for associating each identified job with a globally unique identifier indicative of one or more of a specific interface of the job or a definition of the job.

[0127] Example 53 includes the subject matter of any of Examples 46-52, and wherein the circuitry for associating each identified job with a globally unique identifier comprises circuitry for associating each identified job with a globally unique identifier indicative of one or more of a size of an input or a format of an input to the job.

[0128] Example 54 includes the subject matter of any of Examples 46-53, and wherein the managed node is one of a plurality of managed nodes and the workload is one of a plurality of workloads executed by the managed nodes, the orchestrator server further comprising circuitry for determining, for each workload, a local count indicative of a number of times a job is executed in each workload; circuitry for determining a global count indicative of a number of times a job is executed by all of the managed nodes; circuitry for determining whether one or more of the local count or the global count satisfies a threshold count value; and circuitry for identifying, in response to a determination that one or more of the local count or the global count satisfies the threshold count value, the associated job as a job to be accelerated.

[0129] Example 55 includes the subject matter of any of Examples 46-54, and further including circuitry for identifying, from a plurality of accelerator resources, the one or more accelerator resources to accelerate the one or more jobs.

[0130] Example 56 includes the subject matter of any of Examples 46-55, and wherein the circuitry for identifying the one or more accelerator resources comprises circuitry for determining whether one or more of the accelerator resources is already configured to perform one or more of the jobs, the orchestrator server further comprising circuitry for selecting, in response to a determination that one or more the accelerator resources is already configured to perform one or more of the jobs, the one or more already-configured accelerator resources for acceleration of the one or more jobs.

[0131] Example 57 includes the subject matter of any of Examples 46-56, and wherein the circuitry for identifying the one or more accelerator resources comprises circuitry for selecting the one or more accelerator resources as a function of one or more of a target heat generation, a target power usage, or a target economic cost of utilization of the one or more accelerator resources.

[0132] Example 58 includes the subject matter of any of Examples 46-57, and wherein the managed node is one of a plurality of managed nodes and the workload is one of a plurality of workloads executed by the managed nodes, and wherein the means for determining the demand comprises circuitry for establishing a job queue indicative of all jobs for all of the workloads to be performed; circuitry for determining an average time period in which each job resides in the job queue; and circuitry for determining the demand for each job as a function of the average time period for each job.

[0133] Example 59 includes the subject matter of any of Examples 46-58, and wherein the means for determining the demand for each job further comprises circuitry for applying an exponential averaging algorithm to the time period in which each job resides in the job queue.

Claims

1. An orchestrator server to dynamically manage the allocation of accelerator resources, the orchestrator server comprising: one or more processors; one or more memory devices having stored therein a plurality of instructions that, when executed by the one or more processors, cause the orchestrator server to: assign a workload to a managed node for execution; determine a predicted demand for one or more accelerator resources to accelerate the execution of one or more jobs within the workload; provision, prior to the predicted demand, one or more accelerator resources to accelerate the one or more jobs; and allocate the one or more provisioned accelerator resources to the managed node to accelerate the execution of the one or more jobs.

2. The orchestrator server of claim 1, wherein to determine the predicted demand comprises to determine a demand for one or more field programmable gate arrays (FPGAs).

3. The orchestrator server of claim 2, wherein to provision the one or more accelerator resources comprises to provide, to the one or more FPGAs, a bit stream indicative of a configuration of each FPGA to accelerate execution of the one or more jobs.

4. The orchestrator server of claim 1, wherein to determine the predicted demand comprises to determine the number of accelerator resources to allocate to satisfy the predicted demand.

5. The orchestrator server of claim 1, wherein to provision the one or more accelerator resources comprises to provision one or more accelerator resources located on one or more sleds that are different than a sled on which the workload is presently executed.

6. The orchestrator server of claim 1, wherein the plurality of instructions, when executed, further cause the orchestrator server to: determine a configuration time period to provision each of the one or more accelerator resources; and determine a predicted time of the predicted demand; and wherein to provision the one or more accelerator resources comprises to begin configuration of the one or more accelerator resources for accelerated execution of the one or more jobs at a time that is earlier than the predicted time by at least the configuration time period.

7. The orchestrator server of claim 1, wherein the plurality of instructions, when executed, further cause the orchestrator server to: identify one or more jobs within the workload to be accelerated with one or more field programmable gate arrays (FPGAs); associate each identified job with a globally unique identifier indicative of one or more of a specific interface of the job or a definition of the job.

8. The orchestrator server of claim 7, wherein to associate each identified job with a globally unique identifier comprises to associate each identified job with a globally unique identifier indicative of one or more of a size of an input or a format of an input to the job.

9. The orchestrator server of claim 1, wherein the managed node is one of a plurality of managed nodes and the workload is one of a plurality of workloads executed by the managed nodes and the plurality of instructions, when executed, further cause the orchestrator server to: determine, for each workload, a local count indicative of a number of times a job is executed in each workload; determine a global count indicative of a number of times a job is executed by all of the managed nodes; determine whether one or more of the local count or the global count satisfies a threshold count value; and identify, in response to a determination that one or more of the local count or the global count satisfies the threshold count value, the associated job as a job to be accelerated.

10. The orchestrator server of claim 9, wherein the plurality of instructions, when executed, further cause the orchestrator server to identify, from a plurality of accelerator resources, the one or more accelerator resources to accelerate the one or more jobs.

11. The orchestrator server of claim 10, wherein to identify the one or more accelerator resources comprises to determine whether one or more of the accelerator resources is already configured to perform one or more of the jobs; and select, in response to a determination that one or more the accelerator resources is already configured to perform one or more of the jobs, the one or more already-configured accelerator resources for acceleration of the one or more jobs.

12. The orchestrator server of claim 10, wherein to identify the one or more accelerator resources comprises to select the one or more accelerator resources as a function of one or more of a target heat generation, a target power usage, or a target economic cost of utilization of the one or more accelerator resources.

13. The orchestrator server of claim 1, wherein the managed node is one of a plurality of managed nodes and the workload is one of a plurality of workloads executed by the managed nodes, and wherein to determine the demand comprises to: establish a job queue indicative of all jobs for all of the workloads to be performed; determine an average time period in which each job resides in the job queue; and determine the demand for each job as a function of the average time period for each job.

14. The orchestrator server of claim 13, wherein to determine the demand for each job further comprises to apply an exponential averaging algorithm to the time period in which each job resides in the job queue.

15. One or more machine-readable storage media comprising a plurality of instructions stored thereon that, in response to being executed, cause an orchestrator server to: assign a workload to a managed node for execution; determine a predicted demand for one or more accelerator resources to accelerate the execution of one or more jobs within the workload; provision, prior to the predicted demand, one or more accelerator resources to accelerate the one or more jobs; and allocate the one or more provisioned accelerator resources to the managed node to accelerate the execution of the one or more jobs.

16. The one or more machine-readable storage media of claim 15, wherein to determine the predicted demand comprises to determine a demand for one or more field programmable gate arrays (FPGAs).

17. The one or more machine-readable storage media of claim 16, wherein to provision the one or more accelerator resources comprises to provide, to the one or more FPGAs, a bit stream indicative of a configuration of each FPGA to accelerate execution of the one or more jobs.

18. The one or more machine-readable storage media of claim 15, wherein to determine the predicted demand comprises to determine the number of accelerator resources to allocate to satisfy the predicted demand.

19. The one or more machine-readable storage media of claim 15, wherein to provision the one or more accelerator resources comprises to provision one or more accelerator resources located on one or more sleds that are different than a sled on which the workload is presently executed.

20. The one or more machine-readable storage media of claim 15, wherein the plurality of instructions, when executed, further cause the orchestrator server to: determine a configuration time period to provision each of the one or more accelerator resources; and determine a predicted time of the predicted demand; and wherein to provision the one or more accelerator resources comprises to begin configuration of the one or more accelerator resources for accelerated execution of the one or more jobs at a time that is earlier than the predicted time by at least the configuration time period.

21. The one or more machine-readable storage media of claim 15, wherein the plurality of instructions, when executed, further cause the orchestrator server to: identify one or more jobs within the workload to be accelerated with one or more field programmable gate arrays (FPGAs); associate each identified job with a globally unique identifier indicative of one or more of a specific interface of the job or a definition of the job.

22. The one or more machine-readable storage media of claim 21, wherein to associate each identified job with a globally unique identifier comprises to associate each identified job with a globally unique identifier indicative of one or more of a size of an input or a format of an input to the job.

23. The one or more machine-readable storage media of claim 15, wherein the managed node is one of a plurality of managed nodes and the workload is one of a plurality of workloads executed by the managed nodes and the plurality of instructions, when executed, further cause the orchestrator server to: determine, for each workload, a local count indicative of a number of times a job is executed in each workload; determine a global count indicative of a number of times a job is executed by all of the managed nodes; determine whether one or more of the local count or the global count satisfies a threshold count value; and identify, in response to a determination that one or more of the local count or the global count satisfies the threshold count value, the associated job as a job to be accelerated.

24. The one or more machine-readable storage media of claim 23, wherein the plurality of instructions, when executed, further cause the orchestrator server to identify, from a plurality of accelerator resources, the one or more accelerator resources to accelerate the one or more jobs.

25. An orchestrator server to dynamically manage the allocation of accelerator resources, the orchestrator server comprising: circuitry for assigning a workload to a managed node for execution; means for determining a predicted demand for one or more accelerator resources to accelerate the execution of one or more jobs within the workload; circuitry for provisioning, by the orchestrator server and prior to the predicted demand, one or more accelerator resources to accelerate the one or more jobs; and circuitry for allocating the one or more provisioned accelerator resources to the managed node to accelerate the execution of the one or more jobs.

26. A method for dynamically managing the allocation of accelerator resources, the method comprising: assigning, by an orchestrator server, a workload to a managed node for execution; determining, by the orchestrator server, a predicted demand for one or more accelerator resources to accelerate the execution of one or more jobs within the workload; provisioning, by the orchestrator server and prior to the predicted demand, one or more accelerator resources to accelerate the one or more jobs; and allocating, by the orchestrator server, the one or more provisioned accelerator resources to the managed node to accelerate the execution of the one or more jobs.

27. The method of claim 26, wherein determining the predicted demand comprises determining a demand for one or more field programmable gate arrays (FPGAs).

28. The method of claim 27, wherein provisioning the one or more accelerator resources comprises providing, to the one or more FPGAs, a bit stream indicative of a configuration of each FPGA to accelerate execution of the one or more jobs.