Patent Application
20210280248
The present disclosure relates to high-density, high-efficiency, optically-enabled servers, and the components included therein. Specifically, disclosed embodiments utilize carbon nanotubes (CNTs) in the design of an optically-enabled blade server.
Businesses handling large data generally need higher computing power and memory to organize and process the data. Also, businesses processing high volumes of data use high-powered, low-cost computing modules that can be configured to perform specific processing tasks. These requirements have led to the emergence of technology related to aggregating and/or consolidating multiple computing modules or blade servers. A blade server is a self-contained computing device designed for a specific data processing task, such as high-density data storage. Typically, a blade server includes at least two processors and solid-state memory mounted on a single printed circuit board (PCB), often called a motherboard. Multiple blade servers are housed within a blade enclosure. The blade enclosure may include power supplies, cooling fans, electrical power connections, optical network data interconnections, and peripheral I/O devices communicating with the blade servers. In many instances, a data center can include blade servers and their associated enclosures placed within a rack. By increasing density and reducing cable lengths in each rack, a data center can accommodate hundreds or thousands of blade servers, called a hyperscale data center. However, conventional blade servers face several challenges. For example, network data interconnections provided by blade enclosures to the individual blades have been significantly slow and have failed to provide the optimal, high-speed optical data rate needed to meet current business needs. Furthermore, high-speed volatile memory included in blade servers consume significant amounts of power, which requires installing cooling units for thermal dissipation, which leads to added equipment costs and reduced operational efficiencies.
The accompanying drawings illustrate various embodiments of the principles described herein and are part of the specification. The illustrated embodiments are merely examples and do not limit the scope of the claims.
Embodiments of the present technology are directed at the design of a high-density, high-efficiency, optically-enabled server, and the components included therein. According to disclosed embodiments, the optically-enabled server can include a plurality of blade servers comprising non-volatile carbon nanotubes (CNT) based memory modules instead of traditional volatile silicon-based transistor and capacitor Dynamic Random Access Memory (DRAM). Replacing traditional volatile silicon-based memory with carbon nanotube based non-volatile memory results in at least a 3-fold energy savings over standard DDR4 Synchronous Dynamic Random Access Memory (SDRAM) memory of the same capacity. Further, the miniaturized CNTs enable improvements in processing density of the memory. For example, using 14 nm photolithography, CNT memory chips with storage capacities ranging between 16 Gigabit to 128 Gigabit can be obtained. A high-density, high-efficiency, optically-enabled server (a/k/a optical server) includes high-speed input and output optical network interface modules for providing external optical network 100 Gbps or 200 Gbps optical data streams to the optical blade servers.
The disclosed optical server also includes an optical network adapter module that is designed according to a form factor to fit within the space of an optical blade server (alternatively termed herein as “blade server”) and provide optical network connectivity by interfacing with an optically-enabled midplane included within an enclosure housing multiple optical blade servers. The optically-enabled midplane allows an optical blade server to be connected (e.g., to facilitate east-west Layer 2 traffic) directly to other components within the optical server chassis, and to other optical servers located within other server racks.
In some embodiments, the interfacing between the optical network adapter module and the optically-enabled midplane is based on using mezzanine connector slots within an optical blade server. Additionally, the disclosed optical blade server utilizes multiple optical interconnect modules (ICMs) located at the rear of the optical server to provide high-speed, non-blocking, optical connections between each blade server and external devices/networks (such as optical switches and/or high-speed networks). In some embodiments, the disclosed designs are based on using optical interconnects instead of copper-based interconnects. This can provide substantial energy savings, increased performance, and enhanced security.
In some embodiments, the designs disclosed herein eliminate hard drives and/or solid-state drives within an optical server. By eliminating use of hard drives and solid-state drives, advantageously, at least a 100-fold speed advantage over a single state-of-the-art blade server can be realized. In contrast to conventional designs, the disclosed designs utilize an optically-enabled midplane and/or an optically-enabled back plane that allow the optically-enabled server to be connected directly to other devices within a server chassis, and externally between racks of servers through optical connections for purposes of Layer 2 east/west data traffic. Thus, unlike a rack server which is an independent server within racks, the disclosed technology entails utilizing a collection of modular blade servers working with each other and housed inside a single chassis/enclosure.
In one aspect, the disclosed high-speed, optical blade server design enables additional space to be created on the motherboard for increased airflow and cooling. In one aspect, the disclosed high-speed optical server design can be configured to operate at ambient temperatures less than or equal to 80° Celsius. In one aspect, the disclosed high-speed optical server design using CNT-based memory is immune to ionizing radiation by virtue of CNTs being immune to effects of such radiations.
As used in this document, the singular forms “a,” “an,” and “the” include plural reference unless the context clearly dictates otherwise. Unless defined otherwise, all technical and scientific terms used herein have the same meanings as commonly understood by one of ordinary skill in the art. All publications mentioned in this document are incorporated herein by reference. All sizes recited in this document are by way of example only, and the disclosure is not limited to structures having the specific sizes or dimensions recited below.
The term “server” refers to the name given to any device and computer program that provides functionality for other programs or devices. This architecture commonly referred to as the client-server model, provides for a single overall computation to be distributed across multiple processes or devices. Servers can provide various functionalities, often called “services”, such as sharing data or resources among multiple users or performing computation for a client. A single server can serve multiple clients, and a single client can use multiple servers. A client process may run on the same device or may connect over a network to a server on a different device. Examples of a server include, but are not limited to: Application Server, Collaboration Server, Database Server, Edge Server, File Server, FTP Server, Game Server, Mail Server, Print Server, Windows server, Proxy Server, Real-Time Communication Server, Server Platforms, Web Server, etc.
The term “carbon nanotube” (CNT) refers to a honeycomb lattice rolled into a cylinder. The diameter of a carbon nanotube is of nanometer size and the length of the nanotube can be more than 1 μm. One of the significant physical properties of carbon nanotubes is their electronic structure which depends only on their geometry. Carbon nanotubes (CNTs) are composed of pure carbon in which atoms are positioned in a cylindrical form that exhibit novel properties that make them ideally useful in a wide variety of electronic and optical and physical applications. Single-wall CNTs (SWNTs) can be isolated and oriented, as is the case in CNT nonvolatile memory, to construct extremely dense switches, or used as trace materials on PCBs (either as SWNTS or as SWNT-bundles) to create extremely low impedance interconnects between components. In the case of memory, CNTs enable dramatic reduction in energy demand, 2- to 8-fold increase in storage density, greater reliability, instant on/off, perpetual storage of data, and no soft errors that are common to SDRAM. CNTs are inert to ionizing radiation, and so are radiation immune and well-suited for use in space-related applications. When CNTs are used in designing non-volatile memory, CNT-based memory provides high electrical and thermal conductivity. The thermal conductivity of CNTs makes them well-suited for thermal management (e.g., planar heat dissipation) for electronic devices, reducing the need for active cooling in most devices made therefrom.
The term “printed circuit board” (PCB) refers to the mechanical support and electrical connection of electronic components, or electrical components using conductive tracks, pads and other features etched from one or more sheet layers of a laminated substrate onto and/or between sheet layers of a non-conductive substrate. Components are generally soldered or adhesively bonded onto the PCB to both electrically connect and mechanically fasten them the components to the PCB. Printed circuit boards are used in most electronic products. PCBs can be single-sided, double-sided, or multi-layer. Multi-layer PCBs allow for much higher component density because circuit traces on the inner layers would otherwise take up surface space between components. For purposes of the disclosed technology, multilayer PCBs are considered having more than four trace planes incorporating surface mount technology.
The term “optical interconnect” refers to any system of transmitting and receiving optical signals from one part of an integrated circuit to another using light. Embodiments of the present disclosure provide the successor to electrical interconnects in order to address current and forthcoming needs for the transfer of large data volumes in constantly growing data centers and high-end computing systems. For purposes of this document, the term “optical interconnects” refers to high-speed optical connection into and out of network interface cards or optical network adapter modules mounted in server modules.
The term “module” can be regarded as generally synonymous with the term “card.” Thus, for example, a memory module and a memory card are used interchangeably.
Referring to
Referring to
In some embodiments, an ICM may include at least two computer processors, such as a first microprocessor and a second microprocessor. Each administrator module may include at least one processor configured to perform instructions stored in memory on the administrator module. The administrator modules may be configured to communicate (e.g., send/receive instructions and/or data) internally with one or more optical blade servers and other components inside the optical server.
Referring now to
In some exemplary embodiments, each optical blade server (200) includes two optical blind-mate connectors (209-1 and 209-2 shown in
It will be appreciated that the optical network adapter module (400) is different from a conventional optical network adapter module. For example, the optical network adapter module (400) does not include a PCIe connector for a PCIe slot, which is typically present in a conventional optical network adapter module. According to disclosed embodiments, in place of the PCIe connector, the mezzanine electrical connector (402) is used. The mezzanine electrical connector (402) located on the optical network adapter module (400) is designed to mate/align/interface with mezzanine connector slots (such as slots located on input/output network adapter modules (205-1 to 205-3) located on the motherboard of an optical blade server (200) shown in
Referring now to
The disclosed non-volatile DIMM (500) (e.g., includes 288 pins) is compatible with traditional volatile DIMMs. Current state-of-the-art (SOTA) blade servers or compute modules typically use volatile Synchronous Dynamic Random-Access Memory (SDRAM) multi-gigabit memory requiring a hard disk drive or solid-state drive to retain the computer processing units operating system and other component drivers. Several advantages of the disclosed CNT-based memory are as follows:
In
Referring to the secondary side indicated in
In some embodiments, the NRAM memory chip (501-1 to 501-18) has four internal bank groups comprising four memory banks each, providing a total of 16 banks. This enables use of an 8n-prefetch architecture with an interface designed to transfer two data words per clock cycle at the I/O pins. A single READ or WRITE operation for the NRAM memory chips (501-1 to 501-18) effectively includes a single 8n-bit-wide, four-clock data transfer at the internal NRAM core and eight corresponding n-bit-wide, one-half-clock-cycle data transfers at the I/O pins. In some embodiments, the NRAM memory chip uses two sets of differential signals: DQS_t and DQS_c to capture data and CK_t and CK_c to capture commands, addresses, and control signals. Differential clocks and data strobes ensure exceptional noise immunity for these signals and provide precise crossing points to capture input signals. According to some embodiments, memory cells (e.g., storing one bit of information) are arranged in a two-dimensional array. For example, if the array is of size 8×8, then the total number of bits that can be stored is 64. Each carbon nanotube memory cell has a word line that acts to control the cell. The signal that accesses the cell to either read or write data is applied to the word line. Perpendicular to the word line are bit lines. The data that is written into, or read from the memory, is found on the bit lines.
This embodiment of DIMM module (500) uses faster clock speeds than earlier DDR technologies, making signal quality more important than ever. For improved signal quality, the clock, control, command, and address buses are routed in a fly-by topology, where each clock, control, command, and address pin on each NRAM memory chip is connected to a single trace and terminated (rather than a tree structure, where the termination is off the module near the connector). The fly-by topology accounts for the timing skew between the clock and DQS signals by using the write leveling feature of JEDEC DDR4 specification.
The blind-mate connectors (209-1 and 209-2) line up with a set of blind-mate connectors that are located on an optically-enabled midplane (such as optically-enabled midplane (600) shown in
Some embodiments of the disclosed technology are now presented in clause-based format.
1. A high-density, high-speed optical server comprising:
2. The optical server of clause 1, further comprising:
3. The optical server of clause 1, wherein the plurality of ICMs include a set of ports supporting data transfer at either 100 Gbps or 200 Gbps data rate.
4. The optical server of clause 3, wherein the CNT-based non-volatile memory modules are compliant with IAW JEDEC DDR4 or DDR5 specifications for servers, desktops, and PCs.
5. The optical server of clause 3, wherein the plurality of ICMs include at least four (4) ICMs such that the ICMs support either 100 Gbps data rate or 200 Gbps data rate.
6. The optical server of clause 1, wherein the plurality of blade servers is arranged in at least two rows within the housing, with a first set of blade servers included in a first row and a second set of blade servers included in a second row.
7. The optical server of clause 6, wherein the first set of blade servers and the second set of blade servers each include six blade servers.
8. The optical server of clause 1, wherein the CNT-based non-volatile memory modules are inserted into dual in-line memory module (DIMM) slots disposed on the motherboard of the blade server.
9. The optical server of clause 1, wherein the (CNT)-based non-volatile memory modules comprise 128 GB, 256 GB, or 512 GB storage capacity.
10. The optical server of clause 1, wherein the plurality of CNT-based non-volatile memory modules is designed in accordance with a fly-by topology, wherein a clock, a control, a command, and an address pin on each CNT-based non-volatile memory module is connected to a single trace and terminated.
11. The optical server of clause 1, wherein the optical server is immune to ionizing radiation based on one or more characteristics of the plurality of CNT-based non-volatile memory modules CNTs.
12. The optical server of clause 1, wherein the optical server includes optical interconnections for Layer 2 east-west data traffic inside the optical server, eliminating a use of copper-based interconnections.
13. The optical server of clause 1, wherein the optical server is configured to operate at ambient temperatures less than or equal to 80° Celsius.
14. The optical server of clause 1, wherein the plurality of CNT-based non-volatile memory modules are immune to effects of ionizing radiations.
15. The optical server of clause 1, wherein the one or more external devices includes optical switches and high-speed optical networks.
Some of the embodiments described herein are described in the general context of methods or processes, which may be implemented in one embodiment by a computer program product, embodied in a computer-readable medium, including computer-executable instructions, such as program code, and executed by computers in networked environments. A computer-readable medium may include removable and non-removable storage devices including, but not limited to, Read-Only Memory (ROM), Random Access Memory (RAM), compact discs (CDs), digital versatile discs (DVD), etc. Therefore, the computer-readable media may include a non-transitory storage media. Generally, program modules may include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types. Computer- or processor-executable instructions, associated data structures, and program modules represent examples of program code for executing steps of the methods disclosed herein. The particular sequence of such executable instructions or associated data structures represents examples of corresponding acts for implementing the functions described in such steps or processes.
Some of the disclosed embodiments may be implemented as devices or modules using hardware circuits, software, or combinations thereof. For example, a hardware circuit implementation may include discrete analog and/or digital components that are, for example, integrated as part of a printed circuit board. Alternatively, or additionally, the disclosed components or modules may be implemented as an Application Specific Integrated Circuit (ASIC) and/or as a Field Programmable Gate Array (FPGA) device. Some implementations may additionally or alternatively include a digital signal processor (DSP) that is a specialized microprocessor with an architecture optimized for the operational needs of digital signal processing associated with the disclosed functionalities of this application. Similarly, the various components or sub-components within each module may be implemented in software, hardware or firmware. The connectivity between the modules and/or components within the modules may be provided using any one of the connectivity methods and media that is known in the art, including, but not limited to, communications over the Internet, wired, or wireless networks using the appropriate protocols.
The foregoing description of embodiments has been presented for purposes of illustration and description. The foregoing description is not intended to be exhaustive or to limit embodiments of the present disclosure(s) to the precise form disclosed, and modifications and variations are possible in light of the above teachings or may be acquired from practice of various embodiments. The embodiments discussed herein were chosen and described in order to explain the principles and the nature of various embodiments and their practical application to enable one skilled in the art to utilize the present disclosure(s) in various embodiments and with various modifications as are suited to the particular use contemplated. The features of the embodiments described herein may be combined in all possible combinations of methods, apparatus, modules, systems, and computer program products.
This application claims priority to U.S. Provisional Application Number 62/986,716 filed Mar. 8, 2020, which is incorporated herein by reference in its entirety.
| Number | Date | Country | |
|---|---|---|---|
| 62986716 | Mar 2020 | US |
