CN212647385U

CN212647385U - AI server of 2U support nine two wide GPU cards

Info

Publication number: CN212647385U
Application number: CN202021809621.1U
Authority: CN
Inventors: 程佶; 李增光
Original assignee: Shenzhen Cloud Kirin Computer System Co ltd
Current assignee: Shenzhen Cloud Kirin Computer System Co ltd
Priority date: 2020-08-26
Filing date: 2020-08-26
Publication date: 2021-03-02
Anticipated expiration: 2030-08-26

Abstract

The utility model is suitable for a computer server technical field. The utility model discloses a 2U AI server supporting nine double-width GPU cards, which comprises a server main body (10); a mainboard, a CPU, a memory bank and a hard disk module are arranged in the server main body (10); the front part of the server main body (10) is provided with three double GPU cards (11); the rear part of the server main body (10) is provided with three single GPU cards (12) and two power supply modules (13); a fan wall (14) and a hard disk rack (15) are arranged in the middle of the server main body (10); the beneficial effects of the utility model reside in that: the design is provided with nine double-width GPU cards, the two double-width GPU cards become the AI server with the highest GPU card density, and more GPU cards are supported than the existing 2U AI server; moreover, the design structure is compact, high temperature resistance and good heat dissipation performance are achieved, the operation speed of the server is high, and the stability is higher.

Description

AI server of 2U support nine two wide GPU cards

Technical Field

The utility model relates to a computer server technical field, in particular to AI server of nine two wide GPU cards of 2U support.

Background

With the mature and landing of technologies such as 5G, Internet of things, artificial intelligence and big data, enterprises are also confronted with server replacement and upgrade to meet new scene requirements. The AI server is different from a common server which only adopts a CPU, adopts a heterogeneous form, generally adopts a combination mode of the CPU and the GPU, and can realize powerful parallel computation by adding the GPU card. The AI server has wide application in medical treatment, search engine, game, E-commerce, finance, security protection and other industries. Currently, the highest GPU density of AI servers is: and the 2U AI server supports 8 double-width GPU cards at most. The larger the number of average GPUs supported by each U space in the AI server is, the higher the AI computational power provided by the AI server is, that is, more AI applications can be run in the same cabinet space, so that more electricity charges, cabinet lease fees and hardware purchase costs can be saved.

The existing 2U AI server has the following defects that the 2U AI server supports 8 double-width GPU cards to the maximum, and the GPU density cannot keep up with the market demand of the rapidly developed artificial intelligence; secondly, the server has poor heat dissipation performance, and the operating environment temperature is required to be less than 35 ℃; thirdly, the server is not compact enough in structure and large in size; fourthly, the server is slow in running speed and poor in stability.

SUMMERY OF THE UTILITY MODEL

The utility model discloses the main technical problem who solves provides a 2U supports AI server of nine two wide GPU cards, and it is equipped with nine two wide GPU cards, and compact structure, high temperature resistant, heat dispersion are good, and the functioning speed is fast, and stability is higher.

In order to solve the above problem, the utility model provides a 2U supports AI server of nine pieces of double-width GPU cards, including the server main part; the server main body is a flat rectangular body, and a mainboard, a CPU, a memory bank and a hard disk module are arranged inside the server main body (10); the front part of the server main body is provided with three double GPU cards; the double GPU cards are two double-width GPU cards; the rear part of the server main body is provided with three single GPU cards and two power supply modules; the single GPU card is a double-width GPU card; the power supply module is provided with a power supply, a power supply interface and a power supply fan; the middle part of the server main body is provided with a fan wall and a hard disk rack; the hard disk module is arranged on the hard disk frame; twelve heat dissipation fans are arranged on the fan wall; two USB interfaces, a VGA interface, an indicator light, a switch and a network interface are arranged at the back of the server main body, and the two USB interfaces, the VGA interface, the indicator light, the switch and the network interface are simultaneously integrated on the mainboard; the CPU, the memory bank, the hard disk module, the double GPU cards, the single GPU card, the power module and the fan wall are electrically connected with the mainboard.

Furthermore, the main board is an industrial main board, the working temperature range can reach 0-55 ℃, and the main board has the characteristics of moisture resistance, salt mist resistance and mildew resistance.

Furthermore, the hard disk module can be six specifications of 2.5-inch hard disks, and can also be four specifications of 3.5-inch hard disks.

Further, the memory bank is a DDR4 RDIMM memory bank.

Furthermore, the USB interface is used for connecting external equipment of the USB interface; the VGA interface is used for connecting a display; the network interface is used for connecting a network; the power interface is used for connecting a power supply.

Further, the power supply is a redundant power supply.

Furthermore, the heat dissipation fan and the power supply fan are used for heat dissipation.

Furthermore, the GPU card is a graphics processor, namely, a display core, a visual processor, and a display chip, and is used for a microprocessor for performing image and graphics related operations on a personal computer, a workstation, a game machine, and a mobile device.

The beneficial effects of the utility model reside in that: the design is provided with nine double-width GPU cards, the two double-width GPU cards become the AI server with the highest GPU card density, and more GPU cards are supported than the existing 2U AI server; moreover, the design structure is compact, high temperature resistance and good heat dissipation performance are achieved, the operation speed of the server is high, and the stability is higher.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the description are some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.

Fig. 1 is a schematic cross-sectional view of the present invention.

Fig. 2 is a rear view of fig. 1.

Fig. 3 is a schematic diagram of the dual GPU card of the present invention.

Fig. 4 is a schematic diagram of the single GPU card of the present invention.

Fig. 5 is a schematic view of the fan wall of the present invention.

The objects, features and advantages of the present invention will be further explained with reference to the accompanying drawings.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some embodiments, not all embodiments, of the present invention. Based on the embodiments in the present invention, all other embodiments obtained by a person skilled in the art without creative efforts belong to the protection scope of the present invention.

The first embodiment is as follows:

as shown in fig. 1 to 5, the 2U AI server supporting nine double-wide GPU cards includes a server main body 10; the server main body 10 is a flat rectangular body, and a main board, a CPU, a memory bank and a hard disk module are arranged inside the server main body 10; the front part of the server main body 10 is provided with three double GPU cards 11; the double GPU cards 11 are two double-width GPU cards; the rear part of the server main body 10 is provided with three single GPU cards 12 and two power supply modules 13; the single GPU card 12 is a double-width GPU card; the power module 13 is provided with a power supply, a power interface 131 and a power fan 132; a fan wall 14 and a hard disk rack 15 are arranged in the middle of the server main body 10; the hard disk module is arranged on the hard disk frame 15; the fan wall 14 is provided with twelve heat dissipation fans 141; two USB interfaces 16, a VGA interface 17, an indicator light 18, a switch 19 and a network interface 20 are arranged behind the server main body 10, and the two USB interfaces 16, the VGA interface 17, the indicator light 18, the switch 19 and the network interface 20 are integrated on a mainboard at the same time; the CPU, the memory bank, the hard disk module, the double GPU cards 11, the single GPU card 12, the power module 13 and the fan wall 14 are electrically connected with the mainboard.

As shown in figure 1, the main board is an industrial grade main board, the working temperature range can reach 0-55 ℃, and the main board has the characteristics of moisture resistance, salt mist resistance and mildew resistance.

As shown in fig. 1, the hard disk module may be six 2.5 inch hard disks or four 3.5 inch hard disks.

As shown in fig. 1, the memory bank is a DDR4 RDIMM memory bank.

As shown in fig. 2, the USB interface 16 is used for connecting an external device of the USB interface; the VGA interface 17 is used for connecting a display; the network interface 20 is used for connecting a network; the power interface 131 is used for connecting a power source.

As shown in fig. 2, the power supply is a redundant power supply.

As shown in fig. 1 and 5, the heat dissipation fan 141 and the power supply fan 132 are used for dissipating heat.

As shown in fig. 1 to 4, the GPU card is a graphics processor, namely, a display core, a visual processor, a display chip, and is used for a microprocessor for performing image and graphics related operations on a personal computer, a workstation, a game machine, and a mobile device; a specific double-wide GPU card refers to an image processor that employs two cores.

The above embodiments are only used to illustrate the technical solution of the present invention, and not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it should be understood by those skilled in the art that: it is to be understood that modifications may be made to the above-described arrangements in the embodiments or equivalents may be substituted for some of the features of the embodiments without departing from the spirit or scope of the present invention.

Claims

1. The AI server of the 2U supporting nine double-width GPU cards is characterized by comprising a server main body (10); the server main body (10) is a flat rectangular body, and a mainboard, a CPU, a memory bank and a hard disk module are arranged inside the server main body (10); the front part of the server main body (10) is provided with three double GPU cards (11); the double GPU cards (11) are two double-width GPU cards; the rear part of the server main body (10) is provided with three single GPU cards (12) and two power supply modules (13); the single GPU card (12) is a double-width GPU card; the power supply module (13) is provided with a power supply, a power supply interface (131) and a power supply fan (132); a fan wall (14) and a hard disk rack (15) are arranged in the middle of the server main body (10); the hard disk module is arranged on a hard disk frame (15); twelve heat dissipation fans (141) are arranged on the fan wall (14); two USB interfaces (16), a VGA interface (17), an indicator lamp (18), a switch (19) and a network interface (20) are arranged behind the server main body (10), and the two USB interfaces (16), the VGA interface (17), the indicator lamp (18), the switch (19) and the network interface (20) are integrated on the mainboard at the same time; the CPU, the memory bank, the hard disk module, the double GPU cards (11), the single GPU card (12), the power supply module (13) and the fan wall (14) are electrically connected with the mainboard.

2. The AI server of claim 1, wherein the 2U motherboard is an industrial motherboard with operating temperature ranging from 0 ℃ to 55 ℃ and has moisture, salt fog and mold resistant properties.

3. The AI server of claim 2 wherein said 2U support nine dual-width GPU cards is a six size 2.5 inch hard disk drive or four size 3.5 inch hard disk drive.

4. The AI server of claim 2, wherein the memory banks are DDR4 RDIMM memory banks.

5. The AI server of claim 2, wherein the USB interface (16) is used for connecting an external device of the USB interface; the VGA interface (17) is used for connecting a display; the network interface (20) is used for connecting a network; the power interface (131) is used for connecting a power supply.

6. The AI server of claim 1, wherein the power supply is a redundant power supply.

7. The AI server of claim 1, wherein the 2U heat dissipation fan (141) and the power supply fan (132) are configured to dissipate heat.

8. The AI server of claim 1, wherein the GPU card is a graphics processor, i.e. a display core, a vision processor, a display chip, and a microprocessor for image and graphics related operations on personal computers, workstations, game consoles, and mobile devices.