U.S. patent application number 15/331250 was filed with the patent office on 2017-05-04 for rack level pre-installed interconnect for enabling cableless server/storage/networking deployment.
This patent application is currently assigned to lntel Corporation. The applicant listed for this patent is lntel Corporation. Invention is credited to Matthew J. Adiletta, Aaron Gorius, Amit Y. Kumar, Myles Wilde, Hugh Wilkinson.
Application Number | 20170126330 15/331250 |
Document ID | / |
Family ID | 54191827 |
Filed Date | 2017-05-04 |
United States Patent
Application |
20170126330 |
Kind Code |
A1 |
Adiletta; Matthew J. ; et
al. |
May 4, 2017 |
RACK LEVEL PRE-INSTALLED INTERCONNECT FOR ENABLING CABLELESS
SERVER/STORAGE/NETWORKING DEPLOYMENT
Abstract
Apparatus and methods for rack level pre-installed interconnect
for enabling cableless server, storage, and networking deployment.
Plastic cable waveguides are configured to couple millimeter-wave
radio frequency (RF) signals between two or more Extremely High
Frequency (EHF) transceiver chips, thus supporting millimeter-wave
wireless communication links enabling components in the separate
chassis to communicate without requiring wire or optical cables
between the chassis. Various configurations are disclosed,
including multiple configurations for server chassis, storage
chassis and arrays, and network/switch chassis. A plurality of
plastic cable waveguide may be coupled to applicable
support/mounting members, which in turn are mounted to a rack
and/or top-of-rack switches. This enables the plastic cable
waveguides to be pre-installed at the rack level, and further
enables racks to be installed and replaced without requiring
further cabling for the supported communication links. The
communication links support link bandwidths of up to 6 gigabits per
second, and may be aggregated to facilitate multi-lane links.
Inventors: |
Adiletta; Matthew J.;
(Bolton, MA) ; Gorius; Aaron; (Upton, MA) ;
Wilde; Myles; (Charlestown, MA) ; Wilkinson;
Hugh; (Newton, MA) ; Kumar; Amit Y.;
(Marlborough, MA) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
lntel Corporation |
Santa Clara |
CA |
US |
|
|
Assignee: |
lntel Corporation
Santa Clara
CA
|
Family ID: |
54191827 |
Appl. No.: |
15/331250 |
Filed: |
October 21, 2016 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
14227497 |
Mar 27, 2014 |
9496592 |
|
|
15331250 |
|
|
|
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
H04Q 2011/0052 20130101;
H04B 10/803 20130101; H04L 49/40 20130101; H01P 5/00 20130101; H04B
5/00 20130101; H04B 1/40 20130101; H04Q 11/0066 20130101; H01P 5/02
20130101; H04B 3/52 20130101; H01P 5/028 20130101; H01P 3/10
20130101; H04B 10/40 20130101; H04B 10/90 20130101 |
International
Class: |
H04B 10/80 20060101
H04B010/80; H04B 10/40 20060101 H04B010/40; H04Q 11/00 20060101
H04Q011/00 |
Claims
1. A method, comprising: operatively coupling a first extremely
high frequency (EHF) transceiver chip to a first component in a
first chassis; operatively coupling a second EHF transceiver chip
to a second component in a second chassis; coupling a plastic cable
waveguide to at least one of the first chassis, the second chassis,
or a rack in which the first and second chassis are installed, the
plastic cable waveguide configured to couple millimeter-wave radio
frequency (RF) signals output from the first EHF transceiver chip
into the plastic cable waveguide and to communicatively couple the
millimeter-wave RF signals to the second EHF transceiver chip; and
facilitating communication between the first and second components
by transmitting a millimeter-wave RF signal from the first EHF
transceiver chip to the second EHF transceiver chip via the plastic
cable waveguide.
2. The method of claim 1, wherein the first and second chassis are
installed in the same rack.
3. The method of claim 1, wherein the first and second chassis are
installed in separate racks.
4. The method of claim 1, wherein the first and second EHF
transceiver chips use a 60 GHz carrier frequency.
5. The method of claim 1, wherein the communication between the
first and second components has a bandwidth of 6 gigabits per
second.
6. The method of claim 1, wherein the millimeter-wave RF signal is
transmitted via the plastic cable waveguide by transmitting a
millimeter-wave RF signal from an antenna of the first EHF
transceiver chip toward a first end of the plastic cable waveguide,
wherein the first end comprises the first millimeter-wave radio
frequency (RF) coupling means and is configured to couple the
millimeter-wave RF signal into the plastic cable waveguide.
7. The method of claim 1, wherein the plastic cable waveguide
includes a dielectric manifold that is configured to coupled
millimeter-wave RF signals between the plastic cable waveguide and
an EHF transceiver chip.
8. The method of claim 1, further comprising facilitating a
bi-directional communication link between the first and second
component EHF transceiver chips via the plastic cable
waveguide.
9. The method of claim 1, wherein the plastic cable waveguide
includes a plurality of legs along a portion of its length, the
method further comprises: coupling millimeter-wave RF signals into
each of the plurality of legs transmitted from a respective EHF
transceiver chip disposed proximate to that leg; and coupling
millimeter-wave RF signals out of each of the plurality of legs
toward the respective EHF transceiver chip disposed proximate to
that leg.
10. An apparatus comprising: a first chassis, including a first
component contained therein and having a first extremely high
frequency (EHF) transceiver chip operatively coupled in
communication with the first component; a second chassis, including
a second component contained therein having a second EHF
transceiver chip operatively coupled in communication therewith;
and a first plastic waveguide, operatively coupled to the first and
second chassis, having a first end proximate to the first EHF
transceiver chip and a second end proximate to the second EHF
transceiver chip, wherein when the first plastic waveguide is
configured to facilitate a bi-directional millimeter-wave
communication link between the first and second EFH transceiver
chips when the first and second components are operating.
11. The apparatus of claim 10, further wherein the first EHF
transceiver chip is located within a chassis frame of the first
chassis, and the chassis frame includes a hole proximate to the
first EHF transceiver chip that is configured to enable
millimeter-wave RF signals transmitted from and received by the
first EHF transceiver chip to be passed through the hole.
12. The apparatus of claim 10, wherein the first component
comprises a network interface component or network adaptor.
13. The apparatus of claim 10, wherein the first component
comprises a server blade or server module to which the first EHF
transceiver chip is coupled.
14. The apparatus of claim 10, wherein the first component
comprises a backplane to which the first EHF transceiver chip is
mounted.
15. The apparatus of claim 10, wherein the first and second EHF
transceiver chips use a 60 GHz carrier frequency.
16. An apparatus, comprising: a chassis frame including a metal top
plate in which a plurality of holes are formed; a backplane,
mounted to the chassis frame proximate to the metal top plate,
having a plurality of extremely high frequency (EHF) transceiver
chips mounted thereto, wherein the plurality of EHF transceiver
chips are aligned with the plurality of holes formed in the metal
top plate; and at least one plastic waveguide having a plurality of
legs, each leg disposed proximate to a respective EHF transceiver
chip, wherein upon operation of the apparatus, millimeter-wave
radio frequency (RF) signals transmitted from each EHF transceiver
chip is coupled into the plastic waveguide via the leg that is
disposed proximate to the EHF transceiver chip.
17. The apparatus of claim 16, wherein the backplane further
comprises switching circuitry that is communicatively coupled to
the plurality of EHF transceiver chips and a plurality of network
connectors commutatively coupled to the switching circuitry.
18. The apparatus of claim 16, wherein at least one of the
plurality of the legs extends through a respective hole in the
metal top plate.
19. The apparatus of claim 16, wherein the plurality of EHF
transceiver chips are configured in a plurality of rows, and the
apparatus further comprises a respective plastic waveguide having a
plurality of legs for each row, wherein each leg of the respective
plastic waveguide is disposed proximate to a respective EHF
transceiver chip in the row.
Description
RELATED APPLICATIONS
[0001] The present application is a continuation of U.S. patent
application Ser. No. 14/227,497, filed on Mar. 27, 2014, entitled
"RACK LEVEL PRE-INSTALLED INTERCONNECT FOR ENABLING CABLELESS
SERVER/STORAGE/NETWORKING DEPLOYMENT", which is hereby incorporated
herein by reference in its entirety and for all purposes.
BACKGROUND INFORMATION
[0002] Ever since the introduction of the microprocessor, computer
systems have been getting faster and faster. In approximate
accordance with Moore's law (based on Intel.RTM. Corporation
co-founder Gordon Moore's 1965 publication predicting the number of
transistors on integrated circuits to double every two years), the
speed increase has shot upward at a fairly even rate for nearly
three decades. At the same time, the size of both memory and
non-volatile storage has also steadily increased, such that many of
today's personal computers are more powerful than supercomputers
from just 10-15 years ago. In addition, the speed of network
communications has likewise seen astronomical increases.
[0003] Increases in processor speeds, memory, storage, and network
bandwidth technologies have resulted in the build-out and
deployment of networks with ever increasing capacities. More
recently, the introduction of cloud-based services, such as those
provided by Amazon (e.g., Amazon Elastic Compute Cloud (EC2) and
Simple Storage Service (S3)) and Microsoft (e.g., Azure and Office
365) has resulted in additional network build-out for public
network infrastructure, in addition to the deployment of massive
data centers to support these services that employ private network
infrastructure.
[0004] Cloud-based services are typically facilitated by a large
number of interconnected high-speed servers, with host facilities
commonly referred to as server "farms" or data centers. These
server farms and data centers typically comprise a large-to-massive
array of rack and/or blade servers housed in specially-designed
facilities. Many of the larger cloud-based services are hosted via
multiple data centers that are distributed across a geographical
area, or even globally. For example, Microsoft Azure has multiple
very large data centers in each of the United States, Europe, and
Asia. Amazon employs co-located and separate data centers for
hosting its EC2 and AWS services, including over a dozen AWS data
centers in the US alone.
[0005] In order for the various server blades and modules to
communicate with one another and to data storage, an extensive
amount of cabling is used. Installing the cabling is very
time-consuming and prone to error. In addition, the cost of the
cables and connectors themselves are significant. For example, a
3-foot SAS (Serial attached SCSI) cable may cost $45 alone.
Multiply this by thousands of cables and installations, and the
costs add up quickly, as does the likelihood of cabling errors.
BRIEF DESCRIPTION OF THE DRAWINGS
[0006] The foregoing aspects and many of the attendant advantages
of this invention will become more readily appreciated as the same
becomes better understood by reference to the following detailed
description, when taken in conjunction with the accompanying
drawings, wherein like reference numerals refer to like parts
throughout the various views unless otherwise specified:
[0007] FIG. 1 illustrates a radio frequency antenna output emitted
from a transmitter EHF transceiver chip and being received by a
receiver EHF transceiver chip;
[0008] FIG. 2 is a block diagram of one embodiment of an EHF
transceiver chip;
[0009] FIGS. 3a-3d illustrate launch orientations between pairs of
EHF transceiver chips, wherein FIG. 3a depicts a vertical launch,
FIG. 3b depicts an offset vertical launch, FIG. 3c depicts a side
launch, and FIG. 3d depicts a diagonal launch;
[0010] FIGS. 3e-3i, 3i-a and 3i-b illustrate various communication
link configurations between EHF transceiver chips having their
signals coupled via plastic cable waveguides, wherein FIG. 3e
depicts two EHF transceiver chips in the same orientation, FIGS.
3f-3h depict pairs of EHF transceiver chips oriented 90.degree.
apart, and FIGS. 3i, 3i-a, and 3i-b depict three EHF transceiver
chips in the same orientation, wherein FIG. 3i depicts no sheet
between the ends of the plastic cable waveguide and the EHF
transceiver chips, FIG. 3i-a includes a sheet with holes through
which the signals are passed, and FIG. 3i-b shows the legs of the
plastic cable waveguide passing through holes in a sheet;
[0011] FIGS. 4a and 4b illustrate a millimeter-wave wireless link
between respective EHF transceiver chips in a blade server chassis
above a storage array chassis under which signals are passed
through holes in three metal layers, according to one
embodiment;
[0012] FIG. 4c illustrates a millimeter-wave wireless link between
respective EHF transceiver chips in a blade server chassis above a
storage array chassis under which signals are passed through holes
in one plastic layer and two metal layers, according to one
embodiment;
[0013] FIG. 4d illustrates a millimeter-wave wireless link between
respective EHF transceiver chips in a storage array chassis above a
blade server chassis under which signals are passed through holes
in three metal layers, according to one embodiment;
[0014] FIG. 4e illustrates a millimeter-wave wireless link between
respective EHF transceiver chips in a storage array chassis above a
blade server chassis under which signals are passed through holes
in two metal layers, according to one embodiment;
[0015] FIG. 4f illustrates a millimeter-wave wireless link between
respective EHF transceiver chips in a storage array chassis above a
blade server chassis under which signals are passed through a hole
in one metal layer, according to one embodiment;
[0016] FIGS. 5a and 5b illustrate a millimeter-wave wireless link
between respective EHF transceiver chips in a network/switch
chassis above a blade server chassis under which signals are passed
through holes in three metal layers, according to one
embodiment;
[0017] FIG. 5c illustrates a millimeter-wave wireless link between
respective EHF transceiver chips in a network/switch chassis above
a blade server chassis under which signals are passed through one
plastic layer and holes in two metal layers, according to one
embodiment;
[0018] FIG. 5d illustrates a millimeter-wave wireless link between
respective EHF transceiver chips in a network/switch chassis above
a blade server chassis under which signals are passed through holes
in two metal layers, according to one embodiment;
[0019] FIG. 5e illustrates a millimeter-wave wireless link between
respective EHF transceiver chips in a network/switch chassis above
a blade server chassis with an open top under which signals are
passed through holes in two metal layers, according to one
embodiment;
[0020] FIG. 5f illustrates a millimeter-wave wireless link between
respective EHF transceiver chips in a network/switch chassis above
a blade server chassis under which signals are passed through a
hole in one metal layer, according to one embodiment;
[0021] FIG. 6 is a graphic diagram depicting an electromagnetic
field strength of signals emitted from a transmitting EHF
transceiver chip and passing through holes in two metal layers
using a vertical launch configuration;
[0022] FIG. 7 is a graphic diagram depicting an electromagnetic
field strength of signals emitted from a transmitting EHF
transceiver chip and passing through holes in three metal layers
using a diagonal launch configuration;
[0023] FIGS. 8a and 8b illustrate a configuration under which an
array of EHF transceiver chips in a server chassis are wirelessly
linked with an array of EHF transceiver chips in a storage chassis
below the server chassis, according to one embodiment;
[0024] FIGS. 9a and 9b illustrated a modified version of the
configuration of FIGS. 8a and 9b further adding four fabric
backplanes with EHF transceiver chips on both sides in the server
chassis, according to one embodiment;
[0025] FIGS. 10a and 10b illustrate a configuration under which
components in a middle server chassis are enabled to wirelessly
communicate with components in storage chassis above and below the
server chassis, according to one embodiment;
[0026] FIGS. 11a and 11b illustrate a configuration under which a
6U server chassis is disposed below a network/switch chassis and
above a storage array, according to one embodiment;
[0027] FIGS. 12a and 12b respective show topside and underside
isometric perspective views of a storage array employing an upper
backplane including an array of EHF transceiver chips, according to
one embodiment;
[0028] FIG. 12c shows a topside isometric perspective view of a
storage array employing EHF transceiver chips mounted to vertical
boards to which storage drives are coupled, according to one
embodiment;
[0029] FIG. 12d illustrates a backplane configured for use in a
storage array including an array of SATA connectors on its topside
and an array of EHF transceiver chips on its underside;
[0030] FIGS. 13a and 13b illustrate a backplane configured for use
in a network/switch chassis, according to one embodiment;
[0031] FIG. 14a shows a network switch chassis implementing the
backplane of FIGS. 13a and 13b;
[0032] FIG. 14b shows a network switch chassis implementing two
backplanes of FIGS. 13a and 13b under which the upper backplane is
inverted;
[0033] FIG. 15 illustrates a server module including a pair of EHF
transceiver chips mounted to its main PCB board, according to one
embodiment;
[0034] FIG. 16 is a schematic diagram illustrating a technique for
combining multiple millimeter-wave wireless links in parallel to
increase link bandwidth, according to one embodiment;
[0035] FIG. 17a shows a server rack employing a rack level
pre-installed interconnect employing a plurality of plastic cable
waveguides that are operatively coupled to the rack and/or server
and switch chassis and configured to coupled millimeter-wave RF
signals between EHF transceiver chips in separate chassis;
[0036] FIG. 17b shows further details of the rack level
pre-installed interconnect structure, according to one
embodiment;
[0037] FIG. 17c shows a frontal perspective view of the rack with
multiple server chassis removed to illustrate details of the
provisions in the server chassis for facilitating signal coupling
between EHF transceiver chips inside of the chassis and plastic
cable waveguides on the outside of the chassis;
[0038] FIG. 17d shows an isometric perspective view of a server
chassis configured to facilitate signal coupling between EHF
transceiver chips inside of the chassis and plastic cable
waveguides on the outside of the chassis;
[0039] FIG. 17e shows four racks of servers with top of rack
switches that are communicatively coupled through use of EHF
transceiver chips and pre-installed interconnects comprising
plastic cable waveguides;
[0040] FIG. 17f shows a frontal perspective view illustrating the
routing of plastic cable waveguides in a manner that facilitates
coupling of millimeter-wave RF signals between servers in one rack
and top of the rack switches in the adjacent rack;
[0041] FIGS. 18a and 18b show further details of a top-of-rack
switch and pre-installed interconnect configuration illustrated in
FIGS. 17a-17c, 17e, and 17f, according to one embodiment;
[0042] FIGS. 19a and 19b shows another embodiment of a rack of
servers employing a pre-installed interconnect comprising a
plurality of plastic cable waveguide configured to couple
millimeter-wave RF signals between EHF transceiver chips in
different chassis;
[0043] FIG. 19c shows an isometric frontal view of a server chassis
configured to be used in the server rack of FIG. 19a, according to
one embodiment; and
[0044] FIGS. 19d and 19e illustrate the location of EHF transceiver
chips relative to dielectric manifolds used for couple
millimeter-wave RF signals into and out of the plastic cable
waveguides.
DETAILED DESCRIPTION
[0045] Embodiments of apparatus and methods for rack level
pre-installed interconnect for enabling cableless server, storage,
and networking deployment are described herein. In the following
description, numerous specific details are set forth to provide a
thorough understanding of embodiments described and illustrated
herein. One skilled in the relevant art will recognize, however,
that the invention can be practiced without one or more of the
specific details, or with other methods, components, materials,
etc. In other instances, well-known structures, materials, or
operations are not shown or described in detail to avoid obscuring
aspects of the invention.
[0046] For clarity, individual components in the Figures herein may
also be referred to by their labels in the Figures, rather than by
a particular reference number. Additionally, reference numbers
referring to a particular type of component (as opposed to a
particular component) may be shown with a reference number followed
by "(typ)" meaning "typical." It will be understood that the
configuration of these components will be typical of similar
components that may exist but are not shown in the drawing Figures
for simplicity and clarity or otherwise similar components that are
not labeled with separate reference numbers. Conversely, "(typ)" is
not to be construed as meaning the component, element, etc. is
typically used for its disclosed function, implement, purpose,
etc.
[0047] In accordance with aspects of the embodiments disclosed
herein, Extremely High Frequency (EHF) wireless communication links
are used in place of conventional cabling techniques, resulting in
reductions in both system component costs and labor costs. The
Extremely High Frequency range is approximately 10 GHz-300 GHz. The
embodiments leverage recent advancements in very short length
millimeter-wave wireless transceiver chips to facilitate
contactless communication links for blade server and other
high-density module configurations applicable for data centers and
the like. As used herein, the terminology "millimeter-wave" means
the wavelength of the radio frequency signals is on the order of a
millimeter, which may include RF signals with sub-millimeter length
within the EHF range. Additionally, the embodiments facilitate use
of existing and future server blade and server module
configurations.
[0048] FIG. 1 illustrates radio frequency (RF) signal energy being
output by an antenna in a first EHF transceiver chip 100 operating
as a transmitter (Tx or TX) and being received by a second EHF
transceiver chip 102 that is operating as a receiver (Rx or RX). As
illustrated by the darker shading representing higher energy
density, the electromagnetic field strength of the RF signal
dissipates with distance from the transmitter.
[0049] In one embodiment, each of EHF chips 100 and 102 comprise
EHF chips manufactures by WaveConnex, Inc., Mountainview, Calif. In
one embodiment, the EHF chips illustrated in the Figures herein
comprise a WaveConnex WCX102 (or WCX102b) transceiver chip. Details
of the structure and operations of the millimeter-wave technology
implemented in the WaveConnex chips are disclosed in U.S. Pat. No.
8,554,136 entitled "TIGHTLY-COUPLED NEAR-FIELD COMMUNICATION-LINK
CONNECTOR-REPLACEMENT CHIPS," and U.S. application Ser. Nos.
13/471,052 (U.S. Pub. No. 2012/0286049 A1) and 13/471,058 (U.S.
Pub. No. 2012/0290760 A1), both entitled "SCALABLE HIGH-BANDWIDTH
CONNECTIVITY."
[0050] FIG. 2 shows a block diagram 200 of an embodiment of an EHF
transceiver chip. The basic chip blocks includes a Tx baseband
block 202, RF blocks 204, and an Rx baseband block 206. The RF
blocks include an EHF transmitter block 208, an EHF receiver block
210, and an antenna 212. The EHF chip is configured to receive a
stream of data to be transmitted from an external component using a
differential signal at pins TXinP (positive) and TXinN (negative).
The input transmitted digital stream is processed by Tx baseband
block 202 and EHF transmitter block 208 to create a modulated RF
signal that is radiated output from antenna 212. Antenna 212 also
receives signals transmitted from a paired EHF transceiver of
similar configuration (not) shown, with the received signals
processed by EHF receiver block 210 and Rx baseband block 206 to
generate a received bitstream encoded using differential signaling
that is output at the RXoutP and RXoutN pins. In one embodiment,
the EHF transceiver chip employs a 60 GHz carrier that is generated
on-chip, with the modulated signal sent to antenna 212 for
transmission.
[0051] The EHF transceiver chip includes multiple control inputs
214 that are used for various control and configuration purposes.
The control inputs enable the transceiver chip to be configured in
two operating modes, including a high-speed mode, intended for use
with DC balanced differential signals that is suitable for signals
running from 100 Mb/s to 6.0 Gb/s and features support for
envelope-based Out-of-Band (00B) signaling used in
Serial-Attached-SCSI (SAS) and Serial Advanced Technology
Attachment (SATA), as well as electrical idle and Low Frequency
Periodic Signaling (LFPS) signals in Peripheral Component
Interconnect Express (PCIe) and Universal Serial Bus version 3.0
(USB 3.0).
[0052] The EHF transceiver chips are configured to facilitate very
short range wireless communication links between pairs of
transceiver chips in various orientations. For example, a pair of
chips may be configured with the top surfaces opposite one another
as shown by the vertical launch configuration of FIG. 3a. As shown
in FIG. 3b, the antennas of a pair of EHF transceiver chips do not
need to be in alignment. FIG. 3c shows a configuration under which
a pair of EHF transceiver chips 100 and 102 are in substantially
the same plane. In addition to this configuration, a pair of EHF
transceiver chips can be in respective parallel planes that are
closely spaced (e.g., within a 5-15 millimeters). As shown in FIG.
3d, a diagonal launch configuration is also supported.
[0053] In accordance with further aspects of some embodiments, EHF
transceiver chips are configured to support a plurality of very
short length millimeter-wave wireless links between circuitry and
components in physically separate enclosures, such as chassis
employed in standard 19'' racks. By way of example and without
limitation, a configuration 400 is shown in FIGS. 4a and 4b under
which circuitry on a server blade 402 in a blade server chassis 404
is linked in communication with a disk drive 406 in a storage array
chassis 408. In further detail, server blade 402 includes a main
board 410 to which an EHF transceiver chip 412 is mounted. As an
option, an EHF transceiver chip may be mounted to a daughter board
or otherwise comprise part of a multi-board module. In the
illustrated embodiment of FIG. 4a, server blade 402 is either
mounted within an enclosure including a cover plate 414 or is
coupled to the cover plate 414 in which a hole 416 is formed.
Similarly-sized holes 418 and 420 are respectively formed in the
sheet metal baseplate 422 of blade server chassis 404 an in a top
plate 424 of storage array 408. Preferably, holes 416, 418, and 420
are substantially aligned to form an open pathway 426 through cover
plate 414, baseplate 422 and top plate 424, enabling transmission
of RF energy between EHF transceiver chip 412 and an EHF
transceiver 428 mounted to a backplane 430 in storage array chassis
408. Baseplate 430 includes a plurality of Serial ATA (SATA)
connectors 432 to which disk drive 406 is connected.
[0054] In one embodiment, EHF transceiver chip 428 is configured to
perform signaling to support a SATA interface to facilitate
communication between disk drive 406 and the EHF transceiver chip
using the SATA protocol. Accordingly, configuration 400 enables
circuitry on server blade 410 to write data to and read data from a
disk drive 406 in a separate chassis via an EHF millimeter-wave
bi-directional wireless link 434. As a result, configuration 400
removes the need for use of physical cabling between blade server
chassis 404 and storage array chassis 408.
[0055] Exemplary variations of configuration 400 are shown in FIGS.
4c, 4d, 4e, and 4f. Under a configuration 400c of FIG. 4c, a server
blade 402a is mounted within an enclosure including a plastic cover
plate 415 or otherwise cover plate 415 is attached to main board
410. Unlike metals, which generally attenuate RF signals in the EHF
frequency range, various plastics may be employed that provide
substantially insignificant attenuation. Accordingly, there is no
hole formed in cover plate 415 in the illustrated embodiment.
Alternatively, a hole could be formed in cover plate 415 depending
on the attenuation of the cover plate material in the EHF frequency
range.
[0056] Under configurations 400d, 400e and 400f of respective FIGS.
4d, 4e, and 4f, the storage array chassis 408 is placed above blade
server chassis 404, and the rest of the components are generally
flipped vertically. Configuration 400d is similar to configuration
400, and includes the passing of EHF millimeter-wave bi-directional
wireless link 434 via an open pathway formed by holes 416, 418, and
420 through cover plate 414, baseplate 422 and top plate 424.
[0057] Under some blade server chassis configuration, blade servers
or server modules are inserted vertically and may be "hot-swapped"
without having to power down the entire chassis. In addition, there
are similar blade server chassis configurations under which the
chassis does not include a top or cover plate, since this would
need to be removed to remove or install server blades or modules.
Such a configuration 400e is shown in FIG. 4e, wherein blade server
chassis 404a does not include a cover plate. In this instance, RF
signals to facilitate EHF millimeter-wave bi-directional wireless
link 434 only need to pass through two metal sheets corresponding
to base plate 424 of storage array chassis 408 and a cover plate
423 of a server blade 402b. The also reduces the distance between
EHF transceiver chips 412 and 428.
[0058] Under configuration 400f shown in FIG. 4f, the RF signals
only need to pass through a single metal sheet corresponding to the
base plate 424 of storage array chassis 408. In this configuration,
a blade server 402c does not include a cover plate. Optionally, the
server blade or module could include a plastic cover plate (not
shown) through which a hole may or may not be formed. As with blade
server chassis 404a in FIG. 4e, blade server chassis 404b does not
include a cover plate.
[0059] FIGS. 5a-5f illustrate various configurations under which
server blades in a blade server chassis are enabled to wirelessly
communicate with networking and/or switching components in a
separate chassis. For example, FIGS. 5a and 5b illustrate a
configuration 500 under which a server blade 502 in a blade server
chassis 504 is enabled to communicate with networking circuitry on
a backplane 506 in a network/switch chassis 508 via a an EHF
millimeter-wave bi-directional wireless link 510 facilitated by a
pair of EHF transceiver chips 512 and 514. As before, EHF
transceiver chip 512 is mounted to a main board 516 (or
daughterboard or similar) of server blade 502, which includes
either an enclosure having a cover plate 518 or cover plate 518 is
coupled to main board 516. The other two sheet metal layers
illustrated in FIGS. 5a and 5b correspond to a blade server chassis
cover plate 520 and a bottom plate 522 of network/switch chassis
508. Respective holes 524, 526, and 528 are formed in cover plate
518, cover plate 520, and bottom plate 522, thereby creating an
open pathway 530 through which EHF millimeter-wave bi-directional
wireless link 510 RF signals propagate.
[0060] Configuration 500c of FIG. 5c illustrates a server blade or
module 502a that employs a plastic cover plate 519 rather than a
metal cover plate. As above, depending on the attenuation of EHF RF
signals by the plastic material, a hole through the cover plate may
or may not need to be formed. Under a configuration 500d of FIG.
5d, server blade or module 502b does not employ a cover plate.
Under a configuration 500e shown in FIG. 5e, blade server chassis
504a does not employ a cover plate, while server blade or module
502c employs a metal cover plate 521 with a hole 527 formed through
it. Optionally, cover plate 521 could be made of plastic and may or
may not include a hole (not shown). Under a configuration 500f of
FIG. 5f, neither blade server chassis 504a nor server blade/module
502d employ a cover plate. Thus, the RF signals for EHF
millimeter-wave bi-directional wireless link 510 only need to pass
through a single metal sheet corresponding to bottom plate 522 of
network/switch chassis 508.
[0061] As with any RF signal, the strength of the EHF
millimeter-wave signal is a function of the RF energy emitted for
the RF source (e.g., antenna) and the spectral attributes of the
signal. In turn, the length of the wireless link facilitated
between a pair of EHF transceiver chips will depend on the amount
of RF energy received at the receiver's antenna and signal
filtering and processing capabilities of the EHF receiver
circuitry. In one embodiment, the aforementioned WCX100 chip
supports multiple power output levels via corresponding control
inputs via one or more of control pins 214. Under one embodiment,
the distance between EHF transceiver chips is 2-15 mm, noting that
this is merely exemplary and non-limiting. Generally, higher data
transmission link bandwidth may be achieved when the link's pair of
EHF transceiver chips are closer together and/or using more
power.
[0062] To verify link performance capabilities and expectations
under some of the embodiments disclosed herein, computer-based RF
modeling was performed. Under one approach, the computational
software (ANSYS HFSS) generated a visual representation of the
signal strength of the RF signals emitted from a transmitting EHF
transceiver chip. The models also considered the effect of the
metal sheets/plates between pairs of EHF transceiver chips under
various configurations.
[0063] A snapshot 600 of the RF energy pattern for a configuration
under which the RF signal emitted from a transmitting EHF
transceiver chip 602 is modeled as passing through two metal sheets
604 and 606 before being received by a receiving EHF transceiver
chip 608 is shown in FIG. 6. The model graphically illustrates the
electro-magnetic field energy level in decibels. FIG. 7 shows a
snapshot 700 of the RF energy pattern for a configuration under
which the RF signal emitted from a transmitting EHF transceiver
chip 702 is modeled as passing through three metal sheets 704, 706
and 708 before being received by a receiving EHF transceiver chip
710. In addition to these configurations, various other
configurations were modeled, including variations in the size of
the holes in the metal sheets/plates, the number of metal
sheets/plates, the distance between the pair of EHF transceiver
chips, the orientation of the EHF transceiver chips (e.g., vertical
launch, side launch, diagonal launch, amount of alignment offset,
etc.).
[0064] Generally, the teachings and principles disclosed herein may
be implemented to support wireless communication between components
in separate chassis that are adjacent to one another (e.g., one
chassis on top of another chassis in the rack). Various
non-limiting examples of configurations supporting wireless
communication between chassis using EHF transceiver chips are shown
in FIGS. 8a, 8b, 9a, 9b, 10a, 10b, 11a, 11b, 12a, 12b, 13a, 13b,
14a, and 14b. It will be understood that the configurations shown
in these figures are simplified to emphasize the millimeter-wave
wireless communication facilitated through use of EHF transceiver
chips. Accordingly, these illustrative embodiments may show more or
less EHF transceiver chips than might be implemented, and actual
implementations would include well-known components that are not
shown for convenience and simplicity in order to not obscure the
inventive aspects depicted in the corresponding Figures. In
addition, the illustrated embodiment may not be to scale, and may
present partial or transparent components to reveal other
components that would otherwise be obscured.
[0065] In a configuration 800 of FIGS. 8a and 8b, an array of
server modules 802 are mounted to backplanes 804, 806, 808, and 810
in an upper server chassis 812. In one exemplary embodiment, server
modules 802 comprising Intel.RTM. Avoton.TM. servers modules.
Meanwhile, a plurality of storage drives 814 are coupled to
backplanes 816, 818, 820, and 822 in a lower storage chassis 824.
Each of backplanes 804, 806, 808, and 810 contain an array of
downward-facing EHF transceiver chips 826, while each of backplanes
816, 818, 820, and 822 contain an array of upward-facing EHF
transceiver chips 828, wherein the arrays of the EHF transceiver
chips are configured such that the downward-facing EHF transceiver
chips are aligned respective upward-facing EHF transceiver chips on
a pairwise basis. In addition to what is shown in FIGS. 8a and 8b,
there would be arrays of holes (not shown) in a bottom 830 of upper
chassis 812 and a cover plate (not shown) of lower storage chassis
824.
[0066] FIGS. 9a and 9b show a configuration 1000 under which a
server chassis 812a comprising a modified version of server 812 is
installed above a storage chassis 824. As illustrated, server
chassis 812 now further includes four fabric backplanes 904, 906,
908 and 910 disposed below backplanes 804, 806, 808, and 810. Each
of fabric backplanes 904, 906, 908 and 910 includes an array of
upward-facing EHF transceiver chips 912 mounted to its topside an
array of downward-facing EHF transceiver chips 914 mounted to its
underside. Upward-facing EHF transceiver chips 828 are configured
to be substantially aligned with downward-facing transceiver chips
826 on backplanes 804, 806, 808, and 810. Similarly,
downward-facing EHF transceiver chips 914 are configured to be
substantially aligned with upward-facing transceiver chips on 828
on backplanes 816, 818, 820, and 822.
[0067] FIG. 10a shows a configuration 1000 under which a middle
server chassis 1002 is sandwiched between an upper storage chassis
1004 and a lower storage chassis 1006, with further details of
middle server chassis 1002 depicted in FIG. 10b. Server chassis
1002 includes server board assemblies 1026, each including a
backplane 1010 to which various components are mounted on a topside
thereof including a processor 1012, a plurality of memory modules
1014, and an InfiniBand host bus adaptor (HBA) 1016. Processor 1012
is generally illustrative of one or more processors that may be
included with each server board assembly 1008. An array of EHF
transceiver chips 1020 are mounted to the underside of each
backplane 1010. In addition, there would be a hole pattern having a
configuration similar to the EHF transceiver chips 1020 in the base
plate 1022 of a chassis frame 1024 (not shown for clarity).
[0068] As shown in FIG. 10a, server chassis 1002 also includes an
upper set of four backplanes 1026, each including an array of
upward-facing EHF transceiver chips 1028. In one embodiment,
backplanes 1026 are communicatively coupled to backplanes 1010 via
some form of physical connections, such as but not limited to
connectors between pairs of backplanes or ribbon cables. In the
illustrated embodiments, pairs of upper and lower backplanes are
each connected to an HBA 1016 that further supports communication
between the backplanes.
[0069] Lower storage chassis 1006 is generally configured in a
similar manner to lower storage chassis 824, except the shape of
each of four backplanes 1030 is different than backplanes 816, 818,
820, and 822. As before, an array of upward-facing EHF transceiver
chips 1032 is mounted to each backplane 1030, while a plurality of
storage drives 1034 are coupled to an underneath side of the
backplanes via applicable connectors. There also would be an array
of holes in the cover plate of lower storage chassis 1006 (not
shown) that would be aligned with the array of EHF transceiver
chips 1032.
[0070] Upper storage chassis 1004 is generally configured in a
similar manner to lower storage chassis 1006, but with its vertical
orientation flipped. As a result, each of four backplanes 1036
include a plurality of storage devices 1038 coupled to its topside,
and an array of downward-facing EHF transceiver chips 1040. Upper
storage chassis 1004 would also have an array of storage drives
1038.
[0071] FIGS. 11a and 11b illustrate a configuration 1100 including
a 6U blade server chassis 1102 disposed under a switch chassis 1104
and above a storage array 1006. As shown in FIG. 11b, arrays of
holes are formed in a cover plate 1108 of storage array 1006 and in
a base plate 1110 of switch chassis 1004. Each of a plurality of
server blades 1112 installed in blade server chassis 1102 includes
a frame having an upper plate 1114 and a lower plate 1116 through
which a plurality of holes are formed adjacent to EHF transceiver
chips along the top and bottom edges of the server blade's main
board (not shown), which is mounted to the frame. In addition, the
cover and base plate of blade server chassis 1102 (not shown) will
also include a plurality of holes that are substantially aligned
with the holes in upper plate 1114 and lower plate 1116 when blade
servers 1112 are installed in server chassis 1002.
[0072] FIGS. 12a and 12b show further details of storage array
1106, according to one embodiment. A plurality of storage drives
1200 are mounted to and communicatively coupled with (e.g., via
SATA connectors) vertical boards 1202. In turn, the vertical boards
1202 are communicatively coupled with a backplane 1204 including an
array of EHF transceiver chip 1206. In the illustrated embodiment,
storage drives comprise 21/2 inch drives that are mounted back to
back. Storage drives having other form factors, such as 31/2 inch
drives may be used in other embodiments.
[0073] FIG. 12c shows an embodiment of a storage array 1106a. Under
the illustrated configuration, EHF transceiver chips 1208 are
mounted toward the top of vertical boards 1202. In the illustrated
embodiment one EHF transceiver chip 1208 is implemented for each
drive; however, this is merely exemplary, as multiple EHF
transceiver chips may be used for one or more drive. Also in the
illustrated embodiment the EHF transceiver chips 1208 are mounted
on a single side of vertical boards 1202; optionally, EHF
transceiver chips may be mounted on both sides of the vertical
boards.
[0074] As shown in FIG. 12d, in one embodiment a plurality of SATA
connectors 1210 are mounted to a backplane 1212 having an array of
EHF transceiver chips 1214. In one configuration, backplane 1212 is
disposed in the bottom of a chassis with SATA connectors 1210
pointing upward and EHF transceiver chips 1214 pointing downward.
In another configuration, backplane 1212 is inverted and disposed
toward the top of a chassis with EHF transceiver chips 1214
pointing upward and SATA connectors 1210 pointing downward.
[0075] EHF transceiver chips may be implemented in networking
related chassis, such as switch chassis and network chassis or a
network/switch chassis that includes components supporting
networking and switching functions. Generally, a network/switch
chassis may employ a single backplane or two backplanes arrayed
with EHF transceiver chips, such as illustrated by a network/switch
backplane 1300 in FIGS. 13a and 13b. In this example, a plurality
of Ethernet network connectors 1302 comprising RJ45 Ethernet jacks
are mounted on a topside of network/switch backplane 1300, while an
array of EHF transceiver chips 1304 are mounted on the underside of
the backplane. Wire traces in network/switch backplane 1300 are
routed to network connectors 1302 and a multi-port network/switch
chip 1304. Although only a single multi-port network/switch chip is
shown, it will be understood that multiple chips of similar
configuration may be implemented on a network/switch backplane, and
that network ports and switch operations may also be implemented on
separate chips or otherwise using separate circuitry and logic.
Multi-port network/switch chip 1304 also is connected via wire
traces in network/switch backplane 1300 to EHF transceiver chips
1306. In addition, applicable interface circuitry and
signal-conditioning circuitry (not shown) may be implemented using
techniques and principles well-known in the art.
[0076] The terminology network/switch is meant to convey the
apparatus may be implemented for networking and switching
functions. Depending on the particular system needs or
architecture, a network/switch chassis may include various numbers
of external network ports that are used to interface with other
servers, storage devices, etc. in other chassis and/or other racks,
such as 4, 8 12, 16, 24, etc. In some implementations, a
network/switch backplane may be configured to support switching
functionality related to internal communications in a manner
similar to some switch cards used in data centers and the like.
[0077] FIGS. 14a and 14b respectively show exemplary 1U
network/switch chassis that support wireless connections with a
chassis below (for network/switch chassis 1400a) and with both a
chassis above and below (for a network/switch chassis 1400b). As
shown in FIG. 14a, a network/switch backplane 1300 is mounted
within a 1U chassis frame 1402, with network connectors 1302
mounted toward the rear of the chassis frame. Network/switch
chassis 1400b further adds a second network/switch backplane 1300a
this is mounted such that EHF transceiver chips 1404 are just below
a top) of 1U chassis frame 1402 in which a plurality of holes 1308
are defined proximate to each EHF transceiver chip.
[0078] A server module 1500 configured to facilitate wireless
communication with components in another chassis is shown in FIG.
15. Server module 1500 includes four CPU subsystems comprising
Systems on a Chip (SoCs) 1502a, 1502b, 1502c, and 1502d, each
coupled to respective memories 1504a, 1504b, 1504c, and 1504d. Each
of SoCs 1502a, 1502b, 1502c, and 1502d is also communicatively
coupled to PCIe interface 1506 via a respective PCIe link. Each of
SoCs 1502a, 1502b, 1502c, and 1502d also has access to an
instruction storage device that contains instructions used to
execute on the processing cores of the SoC. Generally, these
instructions may include both firmware and software instructions,
and may be stored in either single devices for a module, or each
SoC may have its own local firmware storage device and/or local
software storage device. As another option, software instructions
may be stored on one or more mass storage modules and accessed via
an internal network during module initialization and/or ongoing
operations.
[0079] Each of the illustrated components are mounted either
directly or via an applicable socket or connector to a printed
circuit board (PCB) 1510 including wiring (e.g., layout traces and
vias) facilitating transfer of signals between the components. This
wiring includes signal paths for facilitating communication over
each of the PCIe links depicted in FIG. 15. PCB 1510 also includes
wiring for connecting selected components to corresponding pin
traces on an edge connector 1512. In one embodiment, edge connector
1512 comprises a PCIe edge connector, although this is merely
illustrative of one type of edge connector configuration and is not
to be limiting. In addition to an edge connector, an arrayed pin
connector may be used, and the orientation of the connector on the
bottom of PCB 1510 in FIG. 15 is exemplary, as an edge or arrayed
pin connector may be located at an end of the PCB, which is a
common configuration for a blade server.
[0080] As further shown in FIG. 15, server module 1500 includes a
pair of EHF transceiver chips 1508 that are mounted toward the top
edge of PCB 1510. This configuration is similar to that shown by
Server Blade/Module 402 in FIG. 4d, Server Blade/Module 402b in
FIG. 4e, and Server Blade/Module 402c in FIG. 4f. In general, a
server module that supports communication via millimeter-wave
wireless links may employ one or more EHF transceiver chips, which
may be mounted on one or both sides of the modules main board
and/or a daughterboard or the like.
[0081] FIG. 16 illustrates an example of combining multiple
individual millimeter-wave EHF links in parallel to support
increased transfer rates across communication interfaces. In the
illustrated embodiment, a server module 1500a includes a four lane
(4x) PCIe interface 1506, and is coupled to a PCIe connector 1600
supporting four (or more) PCIe lanes. Pins corresponding to
respective PCIe differential signal pairs are coupled to the
differential TX input pins on each of a first set of EHF
transceiver chips 1602, which are wirelessly linked in
communication with EHF transceiver chips 1604 on a pairwise basis.
In turn, the differential RX output pins on EHF transceiver chips
1604 are coupled to differential signal pair I/O pins on a PCIe
interface chip 1606. In general, the technique illustrated in FIG.
16 may be used to support an nx PCIe link wherein n is an integer
number of lanes greater than one. For example, standard PCIe
multi-lane links may be implemented, such as 2x, 4x, 8x, 16x, etc.
PCIe links.
[0082] Rack level Pre-installed Interconnect for enabling cableless
Server/Storage/Networking deployment
[0083] According to further aspects of the disclosure, EHF
transceiver chips may be combined with pre-configured and/or
installed waveguides made of plastic or similar material to
facilitate implementation of rack level pre-installed interconnects
that replace conventional cabling. This provides both a cost
savings from both a materials and labor consideration. It further
reduces or eliminates cabling errors, and facilitates tighter rack
spacing.
[0084] During experimentation with the EHF transceiver chips, the
inventors discovered that the millimeter-wave RF signals
transmitted from a chip could be coupled to one end of a plastic
tie-wrap and transmitted out the other end if the ends were cut-off
cleanly. Moreover, the plastic tie-wraps could be bent with minimal
attenuation of the RF signals, thus functioning like an optical
waveguide at millimeter-wave frequencies.
[0085] Extending this concept further, it was realized that server
racks could be "pre-wired" such that when a server chassis was
installed in a rack slot it could be automatically connected to
components in other rack slots via pre-defined connection point in
the rack comprising ends of plastic "cable" waveguides, also
referred to herein as simply plastic waveguides. Optionally,
dielectric "manifolds" may be coupled to or integrated in the
plastic waveguides to further facilitate coupling of
millimeter-wave RF signals between the EHF transceiver chips and
the plastic waveguide connection points.
[0086] Various exemplary configurations illustrating transfer of
millimeter-wave RF signals between pairs of EHF transceiver chips
100 and 102 using plastic waveguides are shown in FIGS. 3e-3h. In
FIG. 3e, a pair of EHF transceiver chips 100 and 102 have a similar
horizontal configuration and a plastic waveguide 300 has an
elongated "U"-shape. In this example, EFT transceiver chips 100 and
102 are substantially in the same plane and the "legs" of plastic
waveguide 300 are substantially equal; however, it will be
recognized that EHF transceiver chips having similar orientations
may be in different planes, and a plastic waveguide's legs may have
different lengths.
[0087] FIG. 3f depicts a configuration under which millimeter-wave
RF signals for EHF transceiver chips 100 and 102 differ by 90
degrees and are coupled via an elongated U-shaped plastic waveguide
302 having legs of different lengths. Similarly, the EHF
transceiver chips 100 and 102 in FIG. 3g differ by 90 degrees, and
have their millimeter-wave RF signals coupled via an L-shaped
plastic waveguide 304. Another configuration employing an L-shaped
plastic waveguide 306 coupling millimeter-wave signals between EHF
transceiver chips 100 and 102 having the same orientation but in
different planes is shown in FIG. 3h.
[0088] In addition to facilitating communication between pairs of
EHF transceiver chips, plastic waveguide may be configured to
facilitate communication between multiple EHF transceiver chips.
For example, FIG. 3i depicts a configuration under which
millimeter-wave RF transmitter signals for three EHF transceiver
chips 100, 102, and 104 are couple via a plastic waveguide 308 with
three legs. More generally, this scheme may be extended to
facilitate communication between n EHF transceiver chips using
plastic waveguides with n legs. As before, the length of the legs
may be the same or may differ.
[0089] Generally, the legs or other receiving members of a plastic
waveguide may either extend into a chassis in which an EHF
transceiver chip is disposed, or the millimeter-wave RF signals may
pass through holes in a chassis baseplate, top-plate, or walls in a
manner similar to shown in FIGS. 4a-4f, 5a-5f, 6 and 7. For
instance, FIG. 3i-a shows a configuration under which
millimeter-wave RF signals transmitted from and received by EHF
transceiver chips 100, 102, and 104, pass through holes 310 in a
sheet metal plate 312 and are coupled to the ends of legs 314, 316,
and 318 of plastic waveguide 308. Meanwhile, FIG. 3i-b illustrates
a configuration under which legs 314, 316, and 318 of plastic
waveguide 308 extend through holes 310 in sheet metal plate
312.
[0090] FIGS. 17a-17g show various levels of details of an
embodiment employing a pair of rack level pre-installed
interconnects 1700a and 1700b. As shown in FIGS. 17a and 17b,
pre-installed interconnects 1700a and 1700b include a plurality of
plastic waveguides 1702 having a top portion including a plurality
of legs 1704 and a vertical portion that extends down the sides of
a rack 1706 having a pair of "top of rack" switches 1708 and 1710
(also shown as "Switch A" and "Switch B," respectively), and having
a plurality of slots in which respective server chassis 1712 are
installed. In the illustrated embodiment, the plastic waveguides
are configured spaced at substantially fixed spacing that is
maintained by a plurality of guides 1714. In one embodiment, guides
1714 are mounted to the sides of the rack (which are not shown so
as not obscure details that would otherwise be hidden). In one
embodiment, pre-installed interconnect 1700a and 1700b may be
assembled as shown in FIG. 17b prior to installation to the rack.
Optionally, guides 1714 may first be mounted to the rack, and then
plastic waveguides 1702 may be installed in slots in guides
1714.
[0091] FIGS. 17c-17e show further details of server chassis 1712,
according to one embodiment. A plurality of holes 1716 are formed
in side plates 1718 and 1720 of each server chassis 1712, wherein
the pattern of the holes are configured to be proximate to
respective plastic waveguides when the server chassis is installed
in the rack. Meanwhile, an EHF transceiver chip is located
proximate to one or more of holes 1716. Generally, an EHF
transceiver chip may be mounted to a vertical board that is
oriented parallel to side plates 1718 and 1720, or perpendicular to
the side plates. This latter configuration is shown in the
embodiment of FIG. 17d, wherein each of multiple microserver boards
1722 include an EHF transceiver chip 1724 that is located toward
the edge of the board proximate to a respective hole 1716. This
configuration enables server chassis to be installed and/or
replaced without requiring any wiring to the server chassis.
[0092] As shown in FIGS. 17e and 17f, rack level pre-installed
interconnects may be used to facilitate communication between
components in multiple racks. FIG. 17e depicts four racks 1706a,
1706b, 1706c and 1706d, each with a pair of top of the rack
switches 1708 and 1710. As shown in FIG. 17f, a portion of plastic
waveguides 1726 that are communicatively coupled with server
chassis 1712 along the right side of rack 1706c are routed to the
top plate of top of rack switch 1710d of rack 1706d. Similarly, a
portion of plastic waveguides 1728 that are communicatively coupled
with server chassis 1712 along the left side of rack 1706d are
routed to the top plate of top of rack switch 1710c of rack 1706c.
Switching functionality provided by each of top of the rack
switches 1708 and 1710 facilitate coupling of signals between
server chassis 1712 in racks 1706c and 1706d. This scheme is
similarly extended to facilitate communication between server
chassis in each of racks 1706a, 1706b, 1706c, and 1706d.
[0093] FIGS. 18a and 18b show further details of a top of the rack
switch 1708 or 1710, according to one embodiment. The top of the
rack switch includes a circuit board 1800 in which a plurality of
EHF transceiver chips 1802 are installed, each disposed opposite a
respective hole 1804 formed in a top plate 1806 or a chassis 1808.
A plurality of network connectors 1810 are mounted to the rear side
of circuit board 1800, with various wire traces and via connecting
the EHF transceiver chips to networks switch circuitry, as
illustrated by a network switch chip 1812. Generally, the location
of the EHF transceiver chips and holes will correspond to the
location of legs 1704 the plastic cable waveguides of pre-installed
interconnects 1700a and 1700b. Depending on the implementation, the
end of legs 1704 may or may not extend through the holes 1804.
[0094] Another embodiment of a rack level pre-installed
interconnect scheme is shown in FIGS. 19a-19e. As shown in FIG.
19a, a plurality of server chassis 1900 are installed in respective
slots in a rack 1902, with sets of plastic waveguides 1904 and 1906
extending up and down the left and right sides of rack 1902. As
shown in FIG. 19b, plastic waveguides 1906 are configured in an
overlapping configuration under which there are a total of 32
plastic waveguides at the top of each side of the rack.
[0095] FIG. 19c shows details of one embodiment of server chassis
1900, which includes four server modules each including a processor
1910, a plurality of memory modules 1912, and a network interface
controller card (NIC) 1914. As further detailed in FIGS. 19d and
19e, each NIC 1914 includes a plurality of EHF transceiver chips
1916, 1918, 1920, and 1922. Each of these EHF transceiver chips is
disposed proximate to a respective dielectric manifold 1924, 1926,
1928, and 1930, which in turn are respectively coupled to a plastic
waveguide 1932, 1934, 1936, and 1938. As further shown, these
plastic waveguides are generally flat in shape and are configured
in a stacked manner. For convenience, the side plate of server
chassis 1900 is not shown, and the side panel 1940 of rack 1902 is
shown as being transparent. In an actual implementation, both of
the server chassis side plates and the rack side panels would
include holes having a pattern that matches the configuration of
the EHF transceiver chips and dielectric manifolds.
[0096] Generally, dielectric manifolds such as illustrated in FIGS.
19d and 19e may be used to facilitate coupling of millimeter-wave
RF signals to and from plastic waveguides. The particular materials
and configuration of the dielectric manifolds may generally depend
on the particular RF frequency employed by the EHF transceiver
chips. The plastic waveguides themselves have dielectric
characteristics relative to the millimeter-wave RF signals, and
thus in some embodiments the dielectric manifolds may be integrally
formed with the plastic waveguides using a single plastic material.
In other embodiments, the dielectric manifolds may be made of a
different material than the plastic waveguide material.
[0097] In general, the plastic multiple waveguides may be
configured in a variety of different cross-sections ranging from
substantially flat to round. Additionally, the cross-section may
also vary along its length. Also, multiple plastic waveguides may
be combined in a bundle along a portion of their length, such as
exemplified by the stacked configuration illustrated in FIGS.
19a-19e, as well as other bundled configurations. Preferably, areas
of the waveguides that would be in contact in the bundle are
separated by a conductive film or sheet, or otherwise means are
provided for preventing millimeter-wave RF signals from being
coupled between plastic waveguides.
[0098] Embodiments implementing the principles and teachings here
provide several advantages over conventional approach. First, by
facilitating millimeter-wave wireless links between EHF transceiver
chips disposed in separate chassis, components that are directly or
indirectly commutatively coupled to EHF transceiver chips are
enabled to pass data to and receive data from components in other
chassis without using a wire or optical cable connected between the
chassis. This results in a cost savings, and also prevents wiring
errors such as might result when connecting a large number of
cables between chassis in a rack. Since the EHF transceiver chips
are mounted to backplanes and other circuit boards, their
implementation can be mass produced at a relatively low marginal
cost (compared to similar components without the chips).
Additionally, since no cable connections are required chassis can
be easily removed for racks for maintenance such as replacement or
upgrade of server blades or modules without having to disconnect
and then reconnect the cables or otherwise need to employ extra
cable lengths to allow for maintenance of chassis components.
[0099] In addition to using separate types of EHF communication
techniques (e.g., EHF transceiver chip-to-EHF transceiver chip, or
chip-to-waveguide-to chip) for a given chassis, the two techniques
may be combined for coupling signals to and from the chassis. For
example, a top of rack switch could couple signals to another top
of rack switch in another rack using plastic cable waveguides,
while coupling signals to a chassis below it using chip-to-chip
signal coupling.
[0100] The following examples pertain to further embodiments. In an
embodiment, a method is implemented that facilitates transfer of
data between components in separate chassis using EHF transceiver
chips and plastic cable waveguides. In accordance with the method,
EHF transceiver chips are operatively coupled to first and second
components in respect first and second chassis. A plastic cable
waveguide is coupled to at least one of the first or second
chassis, or a rack in which the first and second chassis are
installed. The plastic cable waveguide includes first and second
respective millimeter-wave RF coupling means located proximate to
each of the first and second EHF transceiver chips. Communication
between the first and second components is facilitated by
transmitting a millimeter-wave RF signal from the first EHF
transceiver chip to the second EHF transceiver chip via the plastic
cable waveguide. Moreover, bi-direction communication between the
components is supported by also transmitting millimeter-wave RF
signals from the second EHF transceiver chip to the first EHF
transceiver chip via the plastic cable waveguide.
[0101] In an embodiment of the method, the first and second chassis
are installed in the same rack. In another embodiment of the
method, the first and second chassis area installed in separate
racks. In embodiments of the method the EHF transceiver chips use a
60 GHz carrier frequency and support a transfer bandwidth of up to
6 gigabits per second.
[0102] In an embodiment of the method, the millimeter-wave RF
signal is transmitted via the plastic cable waveguide by
transmitting a millimeter-wave RF signal from an antenna of the
first EHF transceiver chip toward a first end of the plastic cable
waveguide, which comprises a millimeter-wave radio frequency (RF)
coupling means and is configured to couple the millimeter-wave RF
signal into the plastic cable waveguide. In an embodiment, at least
one of the first and second millimeter-wave RF coupling means
comprises a dielectric manifold that is coupled to the plastic
cable waveguide.
[0103] In another embodiment of the method the plastic cable
waveguide includes a plurality of legs along a portion of its
length, each comprising a respective millimeter-wave RF coupling
means. A respective EHF transceiver chip is disposed proximate to
each of the plurality of legs, and millimeter-wave RF signals are
coupled into each of the plurality of legs transmitted from the
respective EHF transceiver chip disposed proximate to that leg.
Similarly, millimeter-wave RF signals are coupled out of each of
the plurality of legs toward the respective EHF transceiver chip
disposed proximate to that leg.
[0104] In accordance with further embodiments, apparatus are
configured with means for performing the foregoing method
operations. In an embodiment of an apparatus, the apparatus
comprises a first chassis including a first component having a
first EHF transceiver chip operatively coupled in communication
therewith, and a second chassis including a second component having
a second EHF transceiver chip operatively coupled in communication
therewith. A first plastic waveguide is operatively coupled to at
least one of the first and second chassis, having a first end
proximate to the first EHF transceiver chip and a second end
proximate to the second EHF transceiver chip. The first plastic
waveguide is configured to facilitate a bi-directional
millimeter-wave communication link between the first and second EFH
transceiver chips when the first and second components are
operating.
[0105] In an embodiment of the apparatus, the first EHF transceiver
chip is located within a chassis frame of the first chassis, and
the chassis frame includes a hole proximate to the first EHF
transceiver chip that is configured to enable millimeter-wave RF
signals transmitted from and received by the first EHF transceiver
chip to be passed through the hole. In one embodiment the first
component comprises a network interface component or network
adaptor. In another embodiment, the first component comprises a
server blade or server module to which the first EHF transceiver
chip is coupled. In yet another embodiment, the first component
comprises a backplane to which the first EHF transceiver chip is
mounted. In embodiments of the method the EHF transceiver chips use
a 60 GHz carrier frequency and support a transfer bandwidth of up
to 6 gigabits per second.
[0106] In an embodiment of another apparatus, a chassis frame
includes a metal top plate in which a plurality of holes are
formed, and a backplane, mounted to the chassis frame proximate to
the metal top plate, having a plurality of EHF transceiver chips
mounted thereto, wherein the plurality of EHF transceiver chips are
aligned with the plurality of holes formed in the metal top plate.
The apparatus further includes at least one plastic waveguide
having a plurality of legs, each leg disposed proximate to a
respective EHF transceiver chip. The components are configured such
that upon operation of the apparatus, millimeter-wave RF signals
transmitted from each EHF transceiver chip is coupled into the
plastic waveguide via the leg that is disposed proximate to the EHF
transceiver chip.
[0107] In one embodiment, the backplane further comprises switching
circuitry that is communicatively coupled to the plurality of EHF
transceiver chips and a plurality of network connectors
commutatively coupled to the switching circuitry. In one exemplary
use of this embodiment, the apparatus is implemented as a top of
the rack switch. In an embodiment, at least one of the plurality of
the legs extends through a respective hole in the metal top plate.
In an embodiment, the plurality of EHF transceiver chips are
configured in a plurality of rows, and the apparatus further
comprises a respective plastic waveguide having a plurality of legs
for each row, wherein each leg of the respective plastic waveguide
is disposed proximate to a respective EHF transceiver chip in the
row.
[0108] In another embodiment of an apparatus, a plurality of
plastic cable waveguides are coupled to a plurality of guides, and
each of the guides configured to be mounted to at least one of a
server rack or a member coupled to the server rack. Each plastic
cable waveguide are configured for coupling millimeter-wave RF
signals between the plastic cable waveguide and an EHF transceiver
chip.
[0109] In an embodiment of the apparatus, the millimeter-wave RF
signals employ a 60 GHz carrier frequency. In an embodiment of the
apparatus, the EHF transceiver chips and plastic cable waveguides
are configured to support communication bandwidths of up to 6
gigabits per second. In an aspect of some embodiment, multiple
plastic cable waveguides are bundled together along a portion of
their length. In one embodiment, the plastic cable waveguides are
bundled in a stacked configuration.
[0110] In an embodiment, the apparatus further includes a server
rack to which the plurality of guides are operatively coupled and
having a plurality of server chassis slots configured to receive
respective server chassis, where at least one server chassis slot
includes at least one aperture through which millimeter-wave RF
signals are enabled to pass. The installation is configured such
that when a server chassis including a plurality of EHF transceiver
chips is installed in a server chassis slot including at least one
aperture a respective means for coupling millimeter-wave RF signals
is disposed proximate to a respective EHF transceiver chip included
with the server chassis. In accordance with an aspect of this
embodiment, at least a portion of the means for coupling the
millimeter-wave RF signals are configured in a pattern and the at
least one aperture comprises a plurality of holes in a side panel
of the server rack having a pattern that matches the pattern.
[0111] In accordance with another aspect of the apparatus, at least
one of the plurality of plastic cable waveguides is configured to
couple millimeter-wave RF signals between a first EHF transceiver
chip disposed in the server rack to a second EFH transceiver chip
disposed in another server rack. In an embodiment of the apparatus,
at least one of the plurality of plastic cable waveguide includes a
plurality of legs, wherein each leg is configured for coupling
millimeter-wave RF signals into and out of that plastic cable
waveguide. In an embodiment, dielectric manifolds are coupled to at
least one plastic cable waveguide or integrally formed with at
least one plastic cable waveguide.
[0112] In an embodiment of another method, a plurality of plastic
cable waveguides are coupled to a server rack having a plurality of
slots configured to receive a respective server chassis, at least a
portion of the server chassis including at least one EHF
transceiver chip, wherein when the at least a portion of the server
chassis are installed in the server rack the plastic cable
waveguides are configured to couple millimeter-wave RF signals
between transceiver chips in the server chassis. In an embodiment,
the plurality of plastic cable waveguides are pre-installed, and at
least one server chassis is enabled to be installed and removed
from a corresponding slot in the server rack without requiring
physical connection or disconnection of any wire or optical cables.
In an exemplary implementation, the plastic cable guides are
configured to communicatively couple signals from server chassis in
a rack to a top of the rack switch in the rack.
[0113] In an embodiment, the method further includes operatively
coupling a second plurality of plastic cable waveguides to at least
one of two adjacent server racks, at least one of the racks
including server chassis and a switch chassis including a plurality
of EHF transceiver chips, wherein the second plurality of
waveguides are configured to couple millimeter-wave RF signals
between transceiver chips in at least one of server chassis and
switch chassis in separate racks. In an exemplary implementation
each of the adjacent server racks includes a top of the rack switch
including a plurality of EHF transceiver chips. In accordance with
an embodiment of this implementation, a first plurality of plastic
cable waveguides are employed to communicatively couple signals
from server chassis in a first rack to a top of the rack switch in
the first rack, while a second plurality of plastic cable
waveguides are employed to communicatively couple signals between
the top of the rack switch in the first rack to the top of the rack
switch in the adjacent rack.
[0114] Although some embodiments have been described in reference
to particular implementations, other implementations are possible
according to some embodiments. Additionally, the arrangement and/or
order of elements or other features illustrated in the drawings
and/or described herein need not be arranged in the particular way
illustrated and described. Many other arrangements are possible
according to some embodiments.
[0115] In each system shown in a figure, the elements in some cases
may each have a same reference number or a different reference
number to suggest that the elements represented could be different
and/or similar. However, an element may be flexible enough to have
different implementations and work with some or all of the systems
shown or described herein. The various elements shown in the
figures may be the same or different. Which one is referred to as a
first element and which is called a second element is
arbitrary.
[0116] In the description and claims, the terms "coupled" and
"connected," along with their derivatives, may be used. It should
be understood that these terms are not intended as synonyms for
each other. Rather, in particular embodiments, "connected" may be
used to indicate that two or more elements are in direct physical
or electrical contact with each other. "Coupled" may mean that two
or more elements are in direct physical or electrical contact.
However, "coupled" may also mean that two or more elements are not
in direct contact with each other, but yet still co-operate or
interact with each other.
[0117] An embodiment is an implementation or example of the
inventions. Reference in the specification to "an embodiment," "one
embodiment," "some embodiments," or "other embodiments" means that
a particular feature, structure, or characteristic described in
connection with the embodiments is included in at least some
embodiments, but not necessarily all embodiments, of the
inventions. The various appearances "an embodiment," "one
embodiment," or "some embodiments" are not necessarily all
referring to the same embodiments.
[0118] Not all components, features, structures, characteristics,
etc. described and illustrated herein need be included in a
particular embodiment or embodiments. If the specification states a
component, feature, structure, or characteristic "may", "might",
"can" or "could" be included, for example, that particular
component, feature, structure, or characteristic is not required to
be included. If the specification or claim refers to "a" or "an"
element, that does not mean there is only one of the element. If
the specification or claims refer to "an additional" element, that
does not preclude there being more than one of the additional
element.
[0119] As used herein, a list of items joined by the term "at least
one of" can mean any combination of the listed terms. For example,
the phrase "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.
[0120] As discussed above, various aspects of the embodiments
herein may be facilitated by corresponding software and/or firmware
components and applications, such as software running on a server
or firmware executed by an embedded processor on a network element.
Thus, embodiments of this invention may be used as or to support a
software program, software modules, firmware, and/or distributed
software executed upon some form of processing core (such as the
CPU of a computer, one or more cores of a multi-core processor), a
virtual machine running on a processor or core or otherwise
implemented or realized upon or within a machine-readable medium. A
machine-readable medium includes any mechanism for storing or
transmitting information in a form readable by a machine (e.g., a
computer). For example, a machine-readable medium may include a
read only memory (ROM); a random access memory (RAM); a magnetic
disk storage media; an optical storage media; and a flash memory
device, etc.
[0121] The above description of illustrated embodiments of the
invention, including what is described in the Abstract, is not
intended to be exhaustive or to limit the invention to the precise
forms disclosed. While specific embodiments of, and examples for,
the invention are described herein for illustrative purposes,
various equivalent modifications are possible within the scope of
the invention, as those skilled in the relevant art will
recognize.
[0122] These modifications can be made to the invention in light of
the above detailed description. The terms used in the following
claims should not be construed to limit the invention to the
specific embodiments disclosed in the specification and the
drawings. Rather, the scope of the invention is to be determined
entirely by the following claims, which are to be construed in
accordance with established doctrines of claim interpretation.
* * * * *