Blade type 8-card graphics card GPU (Graphics Processing Unit) server

Abstract

The utility model discloses a blade type 8-card graphics card GPU (Graphics Processing Unit) server, and relates to the technical field of GPU servers. The device comprises a protective shell, a top cover is installed at the top of the protective shell, a side heat dissipation panel is arranged at the front ends of the protective shell and the top cover, a plurality of partition baffles are installed at the lower end of the side heat dissipation panel, and a GPU card is installed between the partition baffles and the protective shell. Compared with the prior art, the GPU server has the advantages that the side edge limiting plates, the guiding mechanisms and the arc guiding plates are additionally arranged in the GPU server, workers can be helped to conduct correct and rapid installation, and in the disassembling process, the rebounding device can provide reverse thrust, so that the workers are helped to conduct disassembling faster and better.

Description

Technical Field

[0001] The utility model relates to the technical field of GPU servers, in particular to a blade-type 8-card GPU server. Background Art

[0002] GPU servers are GPU-based computers and their supporting equipment, which are used in a variety of scenarios such as video encoding and decoding, deep learning, scientific computing, etc., and provide fast, stable, and flexible computing services. They have excellent graphics processing capabilities and high-performance computing capabilities, which can effectively relieve computing pressure and improve the computing processing efficiency and competitiveness of products.

[0003] Existing patent technology 201720802333.5 A GPU server includes: a chassis, in which a fan module is arranged near the middle; at least eight GPU cards are evenly divided into two groups, respectively arranged on both sides of the fan module; at least two PLX / PEX chips, which are arranged at the bottom of the chassis, each PLX / PEX chip allows at least four GPU cards to be interconnected, so that the interconnected GPU cards perform parallel calculations; an OCUlink connector, which is arranged at one end of the chassis, electrically connected to the at least two PLX / PEX chips, and tasks are assigned to each GPU card for calculation through the PLX / PEX chip, and then the GPU calculation results are obtained, and the OCUlink connector is used to input or output data.

[0004] However, when the staff installs and uses it, the installation of the GPU card is still inconvenient and cannot be installed quickly. Moreover, when it needs to be disassembled and replaced later, it still cannot be disassembled quickly, which is inconvenient. Utility Model Content

[0005] The purpose of the utility model is to provide a blade-type 8-card graphics card GPU server in view of the shortcomings and defects in the prior art. The GPU server adds side limit plates, guide mechanisms and arc guide plates inside to help workers to install correctly and quickly. During disassembly, the rebounder can provide reverse thrust to help workers to disassemble faster and better.

[0006] To achieve the above-mentioned purpose, the utility model adopts the following technical scheme: a blade-type 8-card graphics card GPU server, which includes a protective shell 1, a top cover 2 is installed on the top of the protective shell 1, and the upper side heat dissipation panel 3 at the front end of the protective shell 1 and the top cover 2, and a plurality of partition baffles 6 are installed at the lower end of the side heat dissipation panel 3, and a GPU card 7 is installed between the partition baffle 6 and the protective shell 1, the protective shell 1 includes a main shell 14, a supporting cross plate 11 is installed at one end of the inner side of the main shell 14, one end of the supporting cross plate 11 is connected to the partition baffle 6, and a rebounder 12 and a pressure spring 13 are installed on the end of the supporting cross plate 11 connected to the partition baffle 6, a rebounder 12 and a pressure spring 13 are installed on the rebounder 12 and the pressure spring 13, and a rebound plate 15 is installed on the rebounder 12 and the pressure spring 13, and a side limit plate 16, a guide mechanism 17 and an arc guide plate 18 are installed on the partition baffle 6 and the main shell 14 from the inside to the outside.

[0007] Furthermore, at least two heat dissipation panels 3 are provided and are distributed at the front ends of the protective housing 1 and the top cover 2 .

[0008] Furthermore, the side edges of the front end of the protective housing 1 are respectively provided with side connection ports 4 for the convenience of connection control by the staff and lifting handles 5 for the convenience of carrying and lifting by the staff.

[0009] Furthermore, the side limit plates 16 , the guide mechanisms 17 and the arc guide plates 18 are provided in a plurality of groups and are symmetrically distributed on the inner side of the main housing 14 and the partition baffle 6 .

[0010] Furthermore, a card body cavity I is provided between the partition baffle 6 and the main shell 14 , and a card body cavity I is also provided between two adjacent partition baffles 6 .

[0011] Furthermore, the side limit plate 16 is composed of a T-shaped support plate 16-1 and a buffer pad 16-2. One end of the T-shaped support plate 16-1 is installed on the protective shell 1 or the partition baffle 6, and the other end of the T-shaped support plate 16-1 is provided with a buffer pad 16-2.

[0012] Furthermore, the guiding mechanism 17 includes an elastic component 17 - 1 , a roller fixing block 17 - 2 is installed at one end of the elastic component 17 - 1 , and a rolling roller 17 - 3 is rollingly connected to the roller fixing block 17 - 2 .

[0013] After adopting the above technical solution, the beneficial effect of the utility model is as follows: the GPU server, by adding side limit plates, guide mechanisms and arc guide plates inside, can help workers to install correctly and quickly, and when disassembling, its rebounder can provide reverse thrust to help workers disassemble faster and better. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the embodiments of the utility model or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the utility model. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0015] Figure 1 It is a structural schematic diagram of the utility model.

[0016] Figure 2 It is a schematic diagram of part of the internal structure of the utility model.

[0017] Figure 3 yes Figure 2 Enlarged structural diagram at A in the middle.

[0018] Figure 4 It is a structural schematic diagram of the side limit plate 16 in the utility model.

[0019] Figure 5 It is a structural schematic diagram of the guiding mechanism 17 in the utility model.

[0020] Explanation of the accompanying drawings: protective shell 1, top cover 2, heat dissipation panel 3, side connection port 4, lifting handle 5, partition baffle 6, GPU card 7, supporting cross plate 11, rebounder 12, pressure spring 13, main shell 14, rebound plate 15, side limit plate 16, guide mechanism 17, arc guide plate 18, card body cavity I, T-shaped support plate 16-1, buffer pad 16-2, elastic component 17-1, roller fixing block 17-2, rolling roller 17-3. DETAILED DESCRIPTION

[0021] See also Figure 1-Figure 5 As shown, the technical solution adopted in this specific embodiment is: it includes a protective shell 1, a top cover 2 is installed on the top of the protective shell 1, and an upper side heat dissipation panel 3 is installed at the front end of the protective shell 1 and the top cover 2, a plurality of partition baffles 6 are installed at the lower end of the side heat dissipation panel 3, and a GPU card 7 is installed between the partition baffle 6 and the protective shell 1, the protective shell 1 includes a main shell 14, a supporting cross plate 11 is installed at one end of the inner side of the main shell 14, one end of the supporting cross plate 11 is connected to the partition baffle 6, and a rebounder 12 and a pressure spring 13 are installed on the end of the supporting cross plate 11 connected to the partition baffle 6, a rebounder 12 and a pressure spring 13 are installed on the rebounder 12 and the pressure spring 13, and a rebound plate 15 is installed on the rebounder 12 and the pressure spring 13, and a side limit plate 16, a guide mechanism 17 and an arc guide plate 18 are installed on the partition baffle 6 and the main shell 14 from the inside to the outside.

[0022] More specifically, at least two heat dissipation panels 3 are provided, and are distributed at the front ends of the protective housing 1 and the top cover 2. The heat dissipation panels 3 can ensure heat dissipation inside the device, thereby extending the service life of the device and ensuring a safe use environment for the internal electronic components.

[0023] More specifically, the side edges of the front end of the protective housing 1 are respectively provided with side connection ports 4 for the convenience of connection control by the staff and lifting handles 5 for the convenience of carrying and lifting by the staff.

[0024] More specifically, the side limit plates 16, the guide mechanism 17 and the arc guide plates 18 are provided in several groups and are symmetrically distributed on the inner side of the main housing 14 and the partition baffle 6. When the staff needs to install the GPU card 7, the arc guide plates 18 can guide the GPU card 7 to ensure that the GPU card 7 contacts the guide mechanism 17 smoothly and is guided to the side limit plates 16 by the guide mechanism 17, so that the staff can quickly install the GPU card 7.

[0025] More specifically, a card body cavity I is provided between the partition baffle 6 and the main housing 14, and a card body cavity I is also provided between two adjacent partition baffles 6. The card body cavity I facilitates the installation and storage of the GPU card 7.

[0026] More specifically, the side limit plate 16 is composed of a T-shaped support plate 16-1 and a buffer pad 16-2. One end of the T-shaped support plate 16-1 is mounted on the protective housing 1 or the partition baffle 6, and the other end of the T-shaped support plate 16-1 is provided with a buffer pad 16-2. The buffer pad 16-2 can increase the buffering capacity of the side limit plate 16, thereby ensuring the safety of the GPU card 7.

[0027] More specifically, the guide mechanism 17 includes an elastic component 17-1, one end of which is mounted with a roller fixing block 17-2, and a rolling roller 17-3 is rollingly connected to the roller fixing block 17-2. When guiding the GPU card 7, the rolling roller 17-3 can guide the GPU card 7, while the elastic component 17-1 can prevent the GPU card 7 from being squeezed and damaged.

[0028] The working principle of the utility model is as follows: when it is necessary to install the GPU card 7 into the protective shell 1, the staff only needs to install and insert the GPU card 7 into the protective shell 1, and the arc guide plate 18 will perform preliminary guidance, guide the GPU card 7 to the guide mechanism 17, and guide it to the side limit plate 16 through the guide mechanism 17. At this time, the staff only needs to push inward, and the GPU card 7 can squeeze the rebounder 12, thereby completing the installation of the GPU card 7. After the installation is completed, the staff should reinforce and stabilize the GPU card 7. When it is necessary to disassemble and replace the GPU card 7, the staff only needs to push the GPU card 7 to squeeze the rebounder 12 and the pressure spring 13. The pressure spring 13 and the rebounder 12 can provide reverse thrust to push out the GPU card 7, making it convenient for the staff to disassemble and replace the GPU card 7.

[0029] The above is only used to illustrate the technical solution of the utility model rather than to limit it. Other modifications or equivalent substitutions made to the technical solution of the utility model by ordinary technicians in this field should be included in the scope of the claims of the utility model as long as they do not depart from the spirit and scope of the technical solution of the utility model.

Claims

1. A blade-type 8-card GPU server, characterized by: It comprises a protective shell (1), a top cover (2) is installed on the top of the protective shell (1), and an upper side heat dissipation panel (3) at the front end of the protective shell (1) and the top cover (2), and a plurality of partition baffles (6) are installed at the lower end of the side heat dissipation panel (3), and a GPU card (7) is installed between the partition baffle (6) and the protective shell (1); The protective shell (1) comprises a main shell (14), a supporting transverse plate (11) is installed at one end of the inner side of the main shell (14), one end of the supporting transverse plate (11) is connected to the partition baffle (6), and a rebounder (12) and a pressure spring (13) are installed on the end of the supporting transverse plate (11) connected to the partition baffle (6), a rebounder (12) and a pressure spring (13) are installed on the rebounder (12) and the pressure spring (13), and a side limit plate (16), a guide mechanism (17) and an arc guide plate (18) are installed on the partition baffle (6) and the main shell (14) from the inside to the outside.

2. The blade-type 8-card GPU server according to claim 1, characterized in that: At least two heat dissipation panels (3) are provided and are distributed at the front ends of the protective housing (1) and the top cover (2).

3. The blade-type 8-card GPU server according to claim 1, characterized in that: The side edges of the front end of the protective housing (1) are respectively provided with side connection ports (4) for the convenience of connection control by the staff and lifting handles (5) for the convenience of carrying and lifting by the staff.

4. The blade-type 8-card GPU server according to claim 1, characterized in that: The side limit plates (16), the guide mechanisms (17) and the arc guide plates (18) are provided in a plurality of groups and are symmetrically distributed on the inner side of the main housing (14) and the partition baffle (6).

5. The blade-type 8-card GPU server according to claim 1, characterized in that: A card body cavity (I) is provided between the partition baffle (6) and the main shell (14), and a card body cavity (I) is also provided between two adjacent partition baffles (6).

6. The blade-type 8-card GPU server according to claim 1, characterized in that: The side limit plate (16) is composed of a T-shaped support plate (16-1) and a buffer pad (16-2); one end of the T-shaped support plate (16-1) is mounted on the protective housing (1) or the partition baffle (6), and the other end of the T-shaped support plate (16-1) is provided with a buffer pad (16-2).

7. The blade-type 8-card GPU server according to claim 1, characterized in that: The guiding mechanism (17) comprises an elastic component (17-1), one end of which is mounted a roller fixing block (17-2), and a rolling roller (17-3) is rollingly connected to the roller fixing block (17-2).

Images