Multi-three-width GPU (Graphics Processing Unit) card case with temperature monitoring function and server

Abstract

The utility model provides a multi-three-width GPU (Graphics Processing Unit) card case with a temperature monitoring function and a server, which belong to the technical field of three-slot graphics cards and comprise a monitoring module and a plurality of rack node units, the rack node unit is provided with a plurality of connecting cards and a heat dissipation device, the connecting cards are connected with an image processing module, the image processing module is provided with a plurality of mounting slots, and the mounting slots are connected with three-width GPU cards; the monitoring module is electrically connected with the three-width GPU card, detects operation temperature information of the three-width GPU card, and controls the heat dissipation device to operate based on the operation temperature information through being electrically connected with the heat dissipation device. The running temperature of the three-width GPU card is monitored in real time through the monitoring module, stable running of the three-width GPU card is guaranteed, and the stability and reliability of the server are improved. The design of the case supports a plurality of rack node units, and a plurality of three-width GPU cards can be installed in each unit, so that flexible expansion can be performed according to actual requirements, and calculation requirements of different scales and complexity can be met.

Description

Technical Field

[0001] The utility model belongs to the technical field of three-slot graphics cards, and in particular relates to a multi-triple-width GPU card chassis and a server with temperature monitoring. Background Art

[0002] Current GPUs (Graphics Processing Units) are microprocessors used to process images and graphics-related operations. GPUs contain hundreds or even thousands of streaming processors (CUDACores), which can process multiple data points simultaneously, significantly increasing processing speed.

[0003] To meet current usage requirements, servers currently utilize triple-width GPU cards for data processing. Because triple-width GPUs can accommodate more transistors, memory, and other computing resources, they typically offer higher computing performance. This performance boost enables triple-width GPUs to excel in processing large amounts of data, complex calculations, and graphics rendering.

[0004] However, as GPU performance continues to improve, its power consumption and heat generation are also increasing significantly, affecting overall performance and stability. While GPU chassis in related technologies are equipped with cooling devices, they often lack temperature monitoring, making it impossible to sense the operating temperature of the triple-wide GPU card in real time and adjust cooling accordingly. This results in the cooling system being unable to effectively cool the triple-wide GPU card in a timely manner when the card is heavily loaded and its temperature rises, thus affecting the GPU performance and the stability of server operation. Ensuring effective temperature monitoring and cooling control for triple-wide GPU cards is a technical issue that urgently needs to be addressed. Utility Model Content

[0005] The utility model provides a multi-triple-width GPU card chassis with temperature monitoring, which realizes real-time monitoring of the operating temperature of the triple-width GPU cards and ensures the stability and reliability of the operation of the triple-width GPU cards.

[0006] The triple-width GPU card chassis includes: a monitoring module and multiple rack node units;

[0007] The rack node unit is provided with multiple connection cards and a heat dissipation device, the connection card is connected to the image processing module, the image processing module is provided with multiple installation slots, and the installation slots are connected to the triple-width GPU card;

[0008] The monitoring module is electrically connected to the triple-wide GPU card through the connection card, the image processing module and the installation slot in sequence, detects the operating temperature information of the triple-wide GPU card, and is electrically connected to the heat dissipation device to control the operation of the heat dissipation device based on the operating temperature information.

[0009] It should be further explained that the monitoring module includes: a data processing unit, a communication unit and a first communication interface;

[0010] The data processing unit is connected to the rack node unit through the communication unit and the communication interface in turn.

[0011] It should be further explained that the rack node unit is also provided with a second communication interface adapted to the first communication interface.

[0012] It should be further explained that the monitoring module also includes: an aggregation unit;

[0013] The data processing unit is electrically connected to each connection card through the aggregation unit.

[0014] It should be further noted that it also includes: a power supply module for supplying power to each rack node unit;

[0015] The monitoring module also includes: a bidirectional buffer;

[0016] The data processing unit is electrically connected to the power module through a bidirectional buffer to obtain voltage and current information of the output end of the power module.

[0017] It should be further explained that the connection card is electrically connected to the image processing module via a PCIe Gen4 x16 interface or a PCIe Gen5 x16 interface.

[0018] It should be further noted that the data processing unit and the communication unit are electrically connected via an I²C bus;

[0019] The aggregation unit is electrically connected to each connection card through a PCIe Gen4 x16 interface or a PCIe Gen5 x16 interface.

[0020] It should be further explained that the rack node unit is also provided with a temperature sensor for detecting the temperature of the triple-width GPU card; the monitoring module further includes: a signal processing circuit;

[0021] The data processing unit is connected to the temperature sensor through the signal processing circuit, the communication unit, and the communication interface in sequence;

[0022] The signal processing circuit includes: a resistor R1, a resistor R2, a resistor R3, a resistor R4, a resistor R5, a resistor R6, a resistor R7, a resistor R8, a capacitor C1, a capacitor C2, a diode D1, an operational amplifier U2, and a connection terminal U1;

[0023] The input end of the connection terminal U1 is connected to the temperature sensor; the output end of the connection terminal U1 is connected to the first end of the resistor R1; the second end of the resistor R1 is respectively connected to the first end of the resistor R2, the first end of the capacitor C1, and the first end of the resistor R4; the second end of the capacitor C1 is respectively connected to the first end of the resistor R5, the second end of the resistor R6, and the first end of the resistor R7; the first end of the resistor R6 is connected to the power supply;

[0024] The second end of resistor R4 is connected to the second end of resistor R3 and the positive input end of op amp U2 respectively; the second end of resistor R5 is connected to the first end of resistor R8 and the negative input end of op amp U2 respectively; the output end of op amp U2 is connected to the second end of resistor R8, the second end of capacitor C2, the cathode of diode D1 and the output end of the signal processing circuit respectively;

[0025] The second end of the resistor R2 , the first end of the resistor R3 , the second end of the resistor R7 , the second end of the capacitor C2 , and the anode of the diode D1 are grounded respectively.

[0026] It should be further explained that the monitoring module further includes: a display unit; and a data processing unit electrically connected to the display unit.

[0027] The present application also provides a server, comprising: a multi-triple-width GPU card chassis with temperature monitoring.

[0028] It can be seen from the above technical solutions that the present invention has the following advantages:

[0029] The multi-triple-wide GPU card chassis with temperature monitoring provided in this application realizes real-time monitoring of the operating temperature of multiple triple-wide GPU cards, and ensures the stability of the triple-wide GPU cards during high-load operation by controlling the heat dissipation device.

[0030] This application also achieves electrical connection with each connection card through an aggregation unit, allowing the data processing unit to simultaneously monitor the temperature information of multiple triple-width GPU cards. Furthermore, this application also includes a power module for powering each rack node unit, which is electrically connected to the data processing unit via a bidirectional buffer. This allows the data processing unit to obtain real-time voltage and current information at the output of the power module, thereby promptly detecting and addressing power supply anomalies and improving power supply stability.

[0031] This application provides multiple mounting slots on the GPU BOX module to support the installation of six triple-wide GPU cards, allowing the entire chassis to support up to 60 triple-wide GPU cards, improving the server's computing power and scalability. Furthermore, by using PCIe Gen4x16 or PCIe Gen5x16 communication interfaces, it enables compatible connections with triple-wide GPU cards of different models and specifications. This improves the chassis' versatility and flexibility, and also facilitates device upgrades and expansions based on actual needs. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the technical solution of the present invention, the following is a brief introduction to the drawings required for the description. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0033] Figure 1 A schematic diagram of a multi-triple-width GPU card chassis with temperature monitoring.

[0034] Figure 2 This is a schematic diagram of an embodiment of a triple-width GPU card chassis;

[0035] Figure 3 This is an example diagram of a triple-width GPU card chassis;

[0036] Figure 4 This is a schematic diagram of an embodiment of a Retimer card;

[0037] Figure 5 This is the signal processing circuit diagram. DETAILED DESCRIPTION

[0038] In order to make the purpose, features, and advantages of the present invention more obvious and easy to understand, the technical solutions of the present invention will be clearly and completely described below in conjunction with the drawings in the specific embodiments. Obviously, the embodiments described below are only part of the embodiments of the present invention, not all of them. Based on the embodiments in this patent, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this patent.

[0039] like Figure 1 Schematic diagram of a multi-triple-width GPU card chassis with temperature monitoring provided in this embodiment. A monitoring module 1 and multiple rack node units 2 are set inside the multi-triple-width GPU card chassis.

[0040] like Figure 2 As shown, the rack node unit 2 of this embodiment can be an NF5468M7 node 15. The monitoring module 1 of this embodiment includes: a data processing unit, a communication unit, and a second communication interface 14. The data processing unit can be a BMC module 11, the communication unit can be a PHY module 12, and the second communication interface 14 can be an RJ45 interface, although other network interfaces can also be used.

[0041] The NF5468M7 node 15 is equipped with multiple connection cards 3 and a heat sink 19. Connection cards 3 can be retimer cards. Retimer cards eliminate noise, distortion, and attenuation accumulated during transmission. Retimer cards retime, equalize, shape, and amplify incoming PCIe signals, ensuring that the output signals meet regulatory requirements and improving signal integrity.

[0042] like Figure 3 As shown in the figure, a sample diagram of a chassis with multiple triple-width GPU cards is given. Figure 4 In this embodiment, the retimer card is connected to an image processing module 4. The image processing module 4 can be a GPU BOX module. The GPU BOX module is provided with multiple mounting slots 6. Optionally, the multiple mounting slots 6 can accommodate six triple-wide GPU cards 5. The corresponding NF5468M7 node 15 is provided with ten retimer cards, corresponding to the installation of sixty triple-wide GPU cards 5. Furthermore, the chassis design supports multiple rack node units 2, each of which can accommodate multiple triple-wide GPU cards 5. This modular design facilitates flexible expansion based on actual needs, meeting computing requirements of varying scale and complexity.

[0043] In the embodiments of the present application, the triple-width GPU card 5 is a microprocessor (Graphics Processing Unit) that processes images and graphics-related operations. A triple-width GPU card is a three-slot GPU card, compared to a standard dual-slot or single-slot GPU card. It occupies a wider PCI slot space, potentially covering the width of three PCI slots. Because it can accommodate more transistors, memory, and other computing resources, the triple-width GPU card 5 offers higher computing performance. By using the triple-width GPU card 5, servers can improve their efficiency in processing large-scale data, complex calculations, and graphics rendering tasks.

[0044] This embodiment provides multiple mounting slots 6 on the GPU BOX module to support the installation of six triple-wide GPU cards, allowing the entire chassis to support up to 60 triple-wide GPU cards, thereby improving the server's computing power and scalability. Of course, the multi-triple-wide GPU chassis with temperature monitoring can also flexibly configure the number of GPUs based on actual needs to meet different computing requirements.

[0045] Because the GPU BOX is detachable, when a triple-wide GPU card fails, it can be removed and replaced without shutting down the entire system, thereby improving system availability and maintenance efficiency.

[0046] This embodiment also rationally arranges the positions of the mounting slots 6, providing stable support for the triple-wide GPU card while also enhancing the structural strength of the entire chassis. This also effectively improves the heat dissipation environment for the triple-wide GPU card, reducing heat accumulation, optimizing air flow within the chassis, improving heat dissipation efficiency, and enhancing the stability and reliability of the server.

[0047] The monitoring module 1 of this embodiment also includes an aggregation unit. This aggregation unit can be a PCIe SW module, specifically a PCI Express Switch module, which provides expansion or aggregation capabilities and allows more devices to be connected to a single PCIe SW module. This allows the BMC module 11 to be electrically connected to each connection card 3 via the PCIe SW module.

[0048] This embodiment also includes: a power supply module for supplying power to each rack node unit 2; the monitoring module 1 also includes: a bidirectional buffer; the data processing unit is electrically connected to the power supply module through the bidirectional buffer to obtain voltage and current information at the output end of the power supply module.

[0049] Optionally, the bidirectional buffer is a PCA9511 chip 16, and the power module is a PSU module 17. The BMC module 11 communicates with the PSU module 17 via the PCA9511 chip 16, obtains power module operating information, and displays it on the display unit 18. The BMC module 11 can monitor chassis parameters such as temperature, voltage, and fan speed, and perform remote management and control over the network.

[0050] The BMC module 11 of this embodiment is connected to the NF5468M7 node 15 via the PHY module 12 and the J45 interface. The NF5468M7 node 15 is also provided with a J45 interface compatible with the J45 interface to achieve a communication connection. The BMC module 11 detects the operating temperature information of the triple-width GPU card and, through electrical connection with the heat sink 19, controls the operation of the heat sink 19 based on the operating temperature information. The heat sink 19 of this embodiment can be a combination of a heat sink and a cooling fan. The monitoring module 1 also includes: a display unit 18; and a data processing unit electrically connected to the display unit 18.

[0051] The BMC module 11 of this embodiment is connected to the 2 The C bus is connected to the PCA9511 chip 16 and communicates with each PSU module 17, enabling the BMC module 11 to obtain information such as the operating status, voltage output, and current load of the PSU module 17 in real time.

[0052] After the BMC module 11 reads the operating information from each PSU module 17 through the PCA9511 chip 16, the information is displayed through the display unit 18 for the administrator to view and monitor.

[0053] This embodiment adopts a modular design in the overall structure of the chassis to facilitate expansion and maintenance. The interior of the chassis is divided into multiple rack node units 2, and each rack node unit 2 is equipped with multiple Retimer cards and a heat sink 19. The Retimer card is used to connect to the GPU BOX module, and the GPU BOX module is provided with multiple triple-width GPU cards. To ensure the stability of data transmission. Each Retimer card is electrically connected to the GPU BOX module via a PCIe Gen4x16 interface or a PCIe Gen5x16 interface. This ensures the stability and reliability of data transmission.

[0054] To accurately monitor the temperature of the three-wide GPU cards, the BMC module 11, as a core component, possesses multiple functions, including data processing, communication, and interface connectivity. It collects and processes temperature information from each GPU card in real time, intelligently determines cooling requirements through a built-in algorithm, and controls the operating status of the heat sink 19 accordingly.

[0055] This embodiment employs an intelligent heat dissipation method based on temperature information. The BMC module 11 monitors the operating temperature of the triple-width GPU card in real time. When the temperature exceeds a preset threshold, the monitoring module 1, through an electrical connection to the heat sink 19, controls the heat sink 19 to accelerate operation, thereby reducing the temperature of the triple-width GPU card. This ensures stable operation of the triple-width GPU card.

[0056] To improve the overall performance and stability of the chassis, the PCIe SW module and PCA9511 chip 16 enable efficient connectivity between the data processing unit and each link card 3. Electrical connections to each link card 3 are established via a PCIe Gen4x16 or PCIe Gen5x16 interface, ensuring high-speed data transmission. The PCA9511 chip 16 provides a stable connection between the data processing unit and the power module, acquiring voltage and current information from the power module's output, and providing a stable operating environment for the data processing unit.

[0057] In another implementation of this embodiment, the communication and data processing functions of the monitoring module are further refined, and the information interaction capability between the rack node unit and the monitoring module is enhanced.

[0058] In this embodiment, the monitoring module not only monitors the temperature of the triple-width GPU cards but also enables high-speed data transmission and command control with rack node units through advanced communication interfaces and protocols. Specifically, the monitoring module includes multiple functional components, including a data processing unit, a communication unit, a first communication interface, an aggregation unit, and a bidirectional buffer.

[0059] The data processing unit is the core of the monitoring module, responsible for receiving, processing, and analyzing temperature data, voltage and current information, and other status parameters from each rack node unit. Using a high-performance microprocessor or FPGA chip, the data processing unit can process large amounts of data in real time and determine the normal operating status of the triple-width GPU cards based on preset algorithms and thresholds.

[0060] The communication unit is responsible for enabling communication between the data processing unit and external devices. In this embodiment, the communication unit supports multiple communication protocols and standards, such as I²C, SPI, and UART, and can be flexibly configured according to the needs of actual application scenarios. Through the communication unit, the data processing unit can connect to the secondary communication interface on the rack node unit, thereby establishing a stable and reliable data transmission channel.

[0061] The primary communication interface is the physical connection point between the monitoring module and the rack node unit. To accommodate the design requirements of different rack node units, the primary communication interface utilizes a standardized design that can be paired with and connected to various types of secondary communication interfaces. Furthermore, the primary communication interface boasts high-speed data transmission capabilities, ensuring real-time information exchange between the monitoring module and the rack node unit.

[0062] In this embodiment, the rack node units also incorporate temperature sensors for detecting the temperature of the triple-width GPU cards. These sensors can be attached directly to the surface of the triple-width GPU cards or mounted nearby. They collect real-time data on the GPU card's operating temperature and convert it into electrical signals for output. These signals are then amplified, filtered, and digitized by signal processing circuitry before being transmitted via a communication interface to the monitoring module's data processing unit for analysis and assessment.

[0063] This embodiment also optimizes the connection between the link card and the image processing module. Specifically, the link card is electrically connected to the image processing module via a high-speed PCIe Gen4x16 or PCIe Gen5x16 interface. This connection method not only provides high data transfer rates and bandwidth, but also ensures that the image processing module can receive data and instructions from the link card in real time, thereby achieving efficient management and control of the triple-width GPU card.

[0064] It can be seen that this embodiment further improves the performance and reliability of the multi-triple-width GPU card chassis with temperature monitoring by refining the communication and data processing functions of the monitoring module, enhancing the information interaction capability between the rack node unit and the monitoring module, and optimizing the connection method between the connection card and the image processing module.

[0065] As a further embodiment of the present application, in addition to the above-mentioned method, there is also an implementation method of temperature monitoring a multi-triple-width GPU card chassis.

[0066] In this embodiment, the rack node unit is also provided with a temperature sensor for detecting the temperature of the triple-wide GPU card; the monitoring module also includes: a signal processing circuit; the data processing unit is connected to the temperature sensor in sequence through the signal processing circuit, the communication unit, and the communication interface.

[0067] Specifically, if Figure 5 As shown, the signal processing circuit includes: a resistor R1, a resistor R2, a resistor R3, a resistor R4, a resistor R5, a resistor R6, a resistor R7, a resistor R8, a capacitor C1, a capacitor C2, a diode D1, an op amp U2 and a connection terminal U1; the input end of the connection terminal U1 is connected to the temperature sensor; the output end of the connection terminal U1 is connected to the first end of the resistor R1; the second end of the resistor R1 is respectively connected to the first end of the resistor R2, the first end of the capacitor C1 and the first end of the resistor R4; the second end of the capacitor C1 is respectively connected to the first end of the resistor R5, the first end of the resistor R6 The second end of the resistor R4 is connected to the second end of the resistor R3 and the positive input terminal of the operational amplifier U2; the second end of the resistor R5 is connected to the first end of the resistor R8 and the negative input terminal of the operational amplifier U2; the output terminal of the operational amplifier U2 is connected to the second end of the resistor R8, the second end of the capacitor C2, the cathode of the diode D1 and the output terminal of the signal processing circuit; the second end of the resistor R2, the first end of the resistor R3, the second end of the resistor R7, the second end of the capacitor C2 and the anode of the diode D1 are grounded respectively.

[0068] In this embodiment, connection terminal U1 serves as the input interface of the signal processing circuit, connected to the temperature sensor for receiving the temperature signal. Its specific model can adopt JST connection terminal or Molex connection terminal. Resistor R1 serves as the voltage divider resistor for signal input, and together with the subsequent circuit, it forms a voltage divider network. Resistors R2, R3, R1, and R4 together form a voltage divider and feedback network. Resistor R5, R6, and R7 form another set of voltage divider networks, which affects the negative input voltage of the op amp. Resistor R6 serves as a current limiting resistor, protecting the circuit from the influence of excessive current. Resistor R8 serves as the output resistor of the op amp, and together with the op amp, it forms a voltage follower or amplifier to regulate the voltage and current of the output signal. Capacitor C1 serves as a filter capacitor, and capacitor C2 serves as a decoupling capacitor. Together with resistors R1, R2, R4, etc., it forms a low-pass filter to filter out high-frequency noise in the input signal. Diode D1 serves as a protection diode to prevent excessive reverse voltage from occurring at the output of the op amp, protecting the op amp from damage. As the core component of the signal processing circuit, the op amp U2 amplifies and filters the input signal and outputs the processed signal. The op amp U2 uses the LM741 op amp.

[0069] This embodiment has temperature monitoring, and in terms of temperature monitoring, through temperature sensors and data processing units, it can monitor the operating temperature of each triple-width GPU card in real time, and adjust the working status of the heat dissipation device according to actual conditions to ensure that the GPU card always operates within the optimal temperature range.

[0070] The terms "first," "second," "third," "fourth," and so on (if any) in the description and claims of the present invention and the accompanying drawings are used to distinguish similar items and are not necessarily used to describe a particular order or sequential sequence. It should be understood that the terms used in this manner are interchangeable where appropriate, such that the embodiments of the present invention described herein can be implemented in sequences other than those illustrated or described herein. Furthermore, the terms "including," "having," and any variations thereof are intended to cover non-exclusive inclusions.

[0071] The above description of the disclosed embodiments will enable one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not limited to the embodiments shown herein but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A multi-triple-width GPU card chassis with temperature monitoring, characterized in that: include: Monitoring module and multiple rack node units; The rack node unit is provided with multiple connection cards and a heat dissipation device, the connection card is connected to the image processing module, the image processing module is provided with multiple installation slots, and the installation slots are connected to the triple-width GPU card; The monitoring module is electrically connected to the triple-wide GPU card through the connection card, the image processing module and the installation slot in sequence, detects the operating temperature information of the triple-wide GPU card, and is electrically connected to the heat dissipation device to control the operation of the heat dissipation device based on the operating temperature information.

2. The multi-triple-width GPU card chassis with temperature monitoring according to claim 1, characterized in that: The monitoring module includes: a data processing unit, a communication unit and a first communication interface; The data processing unit is connected to the rack node unit through the communication unit and the communication interface in turn.

3. The multi-triple-width GPU card chassis with temperature monitoring according to claim 2, characterized in that: The rack node unit is also provided with a second communication interface adapted to the first communication interface.

4. The multi-triple-width GPU card chassis with temperature monitoring according to claim 2, characterized in that: The monitoring module also includes: an aggregation unit; The data processing unit is electrically connected to each connection card through the aggregation unit.

5. The multi-triple-width GPU card chassis with temperature monitoring according to claim 2, characterized in that: Also includes: A power supply module for supplying power to each rack node unit; The monitoring module also includes: a bidirectional buffer; The data processing unit is electrically connected to the power module through a bidirectional buffer to obtain voltage and current information of the output end of the power module.

6. The multi-triple-width GPU card chassis with temperature monitoring according to claim 1 or 2, characterized in that: The connection card is electrically connected to the image processing module through a PCIe Gen4 x16 interface or a PCIe Gen5 x16 interface.

7. The multi-triple-width GPU card chassis with temperature monitoring according to claim 4, characterized in that: The data processing unit and the communication unit are electrically connected via an I²C bus; The aggregation unit is electrically connected to each connection card through a PCIe Gen4 x16 interface or a PCIe Gen5 x16 interface.

8. The multi-triple-width GPU card chassis with temperature monitoring according to claim 2, characterized in that: The rack node unit is also provided with a temperature sensor for detecting the temperature of the triple-width GPU card; the monitoring module further comprises: a signal processing circuit; The data processing unit is connected to the temperature sensor through the signal processing circuit, the communication unit, and the communication interface in sequence; The signal processing circuit includes: a resistor R1, a resistor R2, a resistor R3, a resistor R4, a resistor R5, a resistor R6, a resistor R7, a resistor R8, a capacitor C1, a capacitor C2, a diode D1, an operational amplifier U2, and a connection terminal U1; The input end of the connection terminal U1 is connected to the temperature sensor; the output end of the connection terminal U1 is connected to the first end of the resistor R1; the second end of the resistor R1 is respectively connected to the first end of the resistor R2, the first end of the capacitor C1, and the first end of the resistor R4; the second end of the capacitor C1 is respectively connected to the first end of the resistor R5, the second end of the resistor R6, and the first end of the resistor R7; the first end of the resistor R6 is connected to the power supply; The second end of resistor R4 is connected to the second end of resistor R3 and the positive input end of op amp U2 respectively; the second end of resistor R5 is connected to the first end of resistor R8 and the negative input end of op amp U2 respectively; the output end of op amp U2 is connected to the second end of resistor R8, the second end of capacitor C2, the cathode of diode D1 and the output end of the signal processing circuit respectively; The second end of the resistor R2 , the first end of the resistor R3 , the second end of the resistor R7 , the second end of the capacitor C2 , and the anode of the diode D1 are grounded respectively.

9. The multi-triple-width GPU card chassis with temperature monitoring according to claim 2, characterized in that: The monitoring module further includes: a display unit; and a data processing unit electrically connected to the display unit.

10. A server, characterized in that: include: A multi-triple-width GPU card chassis with temperature monitoring as claimed in any one of claims 1 to 9.

Images