Efficient machine room global optimization AI computing power box

Abstract

The utility model discloses a kind of high-efficiency computer room global optimization AI computing power box, and it is related to computer room intelligent control technical field.The computing power box includes base box, CPU and external interface mainboard, core computing power GPU board, active heat dissipation temperature control unit, shell and extension module.CPU mainboard is integrated with network port, USB port, video interface, serial port, digital quantity and analog quantity interface;GPU board is arranged side by side with CPU mainboard, responsible for AI calculation;Active heat dissipation unit is installed in the top of GPU board;Shell is connected with base box and encapsulates internal component;Extension module provides additional interface, realizes plug and play.The present application is solved by highly integrated hardware architecture and edge side AI computing power, the problem that current AI deployment scheme cost is high, implementation is complex, can conveniently upgrade central air conditioning refrigeration computer room to global optimization AI high-efficiency computer room, significantly improve system energy efficiency and operating stability.

Description

Technical Field

[0001] This utility model belongs to the field of intelligent control technology for computer rooms, and in particular relates to a high-efficiency AI computing power box for global optimization in computer rooms. Background Technology

[0002] With the increasing scale of data centers and central air conditioning systems in large buildings, energy consumption optimization and intelligent control of data center equipment have become key concerns in the industry. Traditional data center control often adopts a distributed data acquisition and local control model, which has the following technical shortcomings:

[0003] The data acquisition interface is fixed, making it difficult to adapt to different brands and models of data center equipment, resulting in poor scalability;

[0004] Insufficient computing power prevents the real-time execution of complex global optimization AI algorithms, resulting in control lag.

[0005] Poor heat dissipation performance can easily lead to unstable equipment operation in the high-temperature environment of the computer room;

[0006] The equipment has low integration, is complex to install and maintain, and occupies a large space.

[0007] In existing technologies, some edge computing devices attempt to address the aforementioned issues, but they suffer from poor interface compatibility, mismatch between computing power and energy consumption, and low heat dissipation efficiency, failing to meet the high requirements of real-time performance, stability, and scalability for central air conditioning chiller rooms. Therefore, there is an urgent need for an AI computing device with high integration, strong scalability, sufficient computing power, and adaptability to the data center environment. Utility Model Content

[0008] The purpose of this invention is to provide a high-efficiency AI computing power box for global optimization in data centers, thereby solving existing problems.

[0009] To solve the above-mentioned technical problems, this utility model is achieved through the following technical solution:

[0010] This utility model is a high-efficiency AI computing power box for global optimization in data centers, comprising:

[0011] Base box;

[0012] The CPU and external interface motherboard are mounted on the base box and integrate network port, USB port, video interface, serial port, digital interface and analog interface.

[0013] A core computing power GPU board is mounted on the CPU and external interface motherboard and arranged side by side with it;

[0014] An active heat dissipation temperature control unit is installed on the top of the core computing power GPU board;

[0015] The outer shell is fixedly connected to the top of the base box at its bottom, and the active heat dissipation temperature control unit, the core computing power GPU board, the CPU and external interface motherboard are all located inside the outer shell.

[0016] An expansion module, located on one side of the housing, is adapted to connect with the CPU and external interface motherboard to realize interface expansion functions. The expansion module is a serial port, digital interface, or analog interface, used for plug-and-play acquisition of operating data of the central air conditioning refrigeration room and sending optimized control parameters to the underlying devices.

[0017] As a preferred technical solution of this utility model, the CPU and external interface motherboard are further provided with a memory card mounting position and operation buttons. The memory card mounting position is used to install a storage medium to store data, and the operation buttons are used for device function operation control.

[0018] As a preferred embodiment of this utility model, the digital interface is used to acquire the switch status signal of the computer room equipment and output the switch control signal, and the analog interface is used to acquire the analog operating parameters of the computer room equipment and output the analog control signal.

[0019] As a preferred technical solution of this utility model, the active heat dissipation temperature control unit includes an active heat dissipation fan, which dissipates heat inside the device through the operation of the fan.

[0020] As a preferred technical solution of this utility model, the active heat dissipation temperature control unit further includes a temperature sensor and a temperature control chip. The temperature sensor collects internal temperature data of the device and transmits it to the temperature control chip. The temperature control chip adjusts the speed of the active heat dissipation fan according to the temperature data to achieve adaptive heat dissipation.

[0021] As a preferred embodiment of this utility model, the USB port on the CPU and external interface motherboard includes a USB 2.0 interface and a USB 3.0 interface. The USB 2.0 interface is used to connect low-speed peripherals, and the USB 3.0 interface is used to connect high-speed data transmission peripherals.

[0022] This utility model has the following beneficial effects:

[0023] High integration and plug-and-play: Integrating AI computing power and multiple industry standard interfaces into a compact device greatly simplifies on-site wiring and technical integration, enabling rapid deployment of AI systems.

[0024] Advantages of edge computing: It moves AI computing tasks from the cloud to the edge of the device, reducing network dependence and latency, improving the real-time performance of control and the reliability of the system, while ensuring data security.

[0025] Strong adaptability to different scenarios: Through its scalable modular design, it can flexibly adapt to project requirements with different numbers and types of interfaces, and can be well applied to both new data centers and the renovation of old data centers.

[0026] Industrial-grade reliability: The PE material shell, active heat dissipation and intelligent temperature control design ensure that the equipment can operate stably for a long time in high temperature, high humidity and dusty industrial environments.

[0027] Significant cost-effectiveness: Replacing expensive server clusters with a low-cost dedicated device significantly reduces the initial investment and maintenance costs for users implementing AI energy-saving technologies.

[0028] Of course, any product implementing this utility model does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0029] To more clearly illustrate the technical solutions of the embodiments of this utility model, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this utility model. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0030] Figure 1 This is an exploded view of the present invention;

[0031] Figure 2 This is a three-dimensional structural diagram of the present invention;

[0032] Figure 3 A three-dimensional structural diagram of the expansion module in this utility model when a digital interface is selected.

[0033] Figure 4 This refers to the analog interface in the expansion module mentioned in this utility model;

[0034] Figure 5 This refers to the serial port in the expansion module mentioned in this utility model;

[0035] The attached diagram lists the components represented by each number as follows:

[0036] 1-Dock box; 2-CPU and external interface motherboard; 3-Core computing power GPU board; 4-Active heat dissipation temperature control unit; 41-Active heat dissipation fan; 5-Casing; 6-Expansion module; 7-Serial port; 8-USB port; 9-Operation buttons; 10-Memory card mounting slot; 11-Integrated network port; 12-Power interface; 13-Video interface. Detailed Implementation

[0037] The technical solutions of the present utility model will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of the present utility model, and not all embodiments. Based on the embodiments of the present utility model, all other embodiments obtained by those skilled in the art without creative effort are within the protection scope of the present utility model.

[0038] In the description of this utility model, it should be understood that the terms "opening", "upper", "lower", "thickness", "top", "middle", "length", "inner", "around" and other terms indicating orientation or positional relationship are only for the convenience of describing this utility model and simplifying the description, and do not indicate or imply that the components or elements referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this utility model.

[0039] See Figure 1 , Figure 2 , Figure 3 and Figure 4 , Figure 5 This utility model provides a high-efficiency AI computing power box for global optimization in computer rooms.

[0040] The physical structure of the computing box, from bottom to top, consists of: a base box 1, typically made of metal or high-strength engineering plastic, providing structural strength and mounting support; and a CPU and external interface motherboard 2, fixed to the base box 1 with screws. This motherboard is the control and communication hub of the device, integrating at least one integrated network port 11, such as an RJ45 interface, for connecting to the computer room LAN or the Internet; multiple USB ports 8 (including USB 2.0 and USB 3.0); a video interface 13 (such as HDMI); multiple serial ports 7 (such as RS485); and several digital and analog interfaces. The motherboard also includes a memory card slot 10 (such as a TF card slot), operation buttons 9 (such as a power button, reset button, and function keys), and a power interface 12.

[0041] The core computing power GPU board 3 is mounted vertically / parallel to the CPU and external interface motherboard 2 via board-to-board connectors or slots, and is arranged side by side with it. This GPU board can use the core module of a high-performance, low-power AI chip to provide sufficient computing power (e.g., computing power of more than 20 TOPS) for running deep learning inference models.

[0042] The active cooling temperature control unit 4 is securely mounted on the top of the core computing GPU board 3 using screws and thermal paste. This unit primarily consists of one or more active cooling fans, a temperature sensor mounted near the GPU, and a temperature control chip. When the GPU operates, it generates heat, and the temperature sensor monitors its temperature in real time. Based on the monitored temperature value, the temperature control chip dynamically adjusts the fan speed. For example, when the temperature is below 50°C, the fan operates at a low speed; when the temperature rises to 50°C-70°C, the fan speed increases linearly; and when the temperature exceeds 70°C, the fan operates at its highest speed, thus achieving efficient and quiet adaptive cooling.

[0043] The outer casing 5 is made of flame-retardant and high-temperature resistant PE (polyethylene) or ABS engineering plastic. Its bottom is fixedly connected to the top of the base box 1 by screws, encapsulating all the core electronic components inside to form a complete device.

[0044] The expansion module 6 is located on the side of the casing 5 and is connected to the CPU and external interface motherboard 2 via a board-to-board connector or ribbon cable. For example... Figure 3 and Figure 4 As shown, the expansion module can be a dedicated analog input / output expansion module (providing an additional 8 AI / AO channels), a digital input / output expansion module (providing an additional 16 DI / DO channels), or a dual serial port expansion module (providing an additional 2 RS485 interfaces). Users can select the appropriate expansion module according to the actual number of IO points required on-site, achieving a high degree of flexibility and customizability.

[0045] Working principle:

[0046] During on-site deployment, technicians install the computing power box inside the control cabinet in the computer room or near the equipment. Through its integrated and expanded interfaces, it connects to various field devices such as chillers, variable frequency pumps, cooling towers, temperature sensors, pressure sensors, and flow meters. After power-on, the computing power box begins collecting system operating data and performs real-time calculations using built-in AI algorithm models (such as trained deep neural network models) to optimize the set parameters of each device (such as chiller outlet water temperature and pump frequency) for maximum system energy efficiency. These optimized parameters are then sent to the field actuators through the corresponding interfaces. The entire process is completed automatically at the edge, forming a closed-loop intelligent optimization control system.

[0047] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0048] The preferred embodiments of this utility model disclosed above are merely illustrative of the present utility model. These preferred embodiments do not exhaustively describe all details, nor do they limit the utility model to the specific implementations described. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of this utility model, thereby enabling those skilled in the art to better understand and utilize it. This utility model is limited only by the claims and their full scope and equivalents.

Claims

1. A high-efficiency AI computing power box for global optimization in data centers, characterized in that, include: Base box; The CPU and external interface motherboard are mounted on the base box and integrate network port, USB port, video interface, serial port, digital interface and analog interface. A core computing power GPU board is mounted on the CPU and external interface motherboard and arranged side by side with it; An active heat dissipation temperature control unit is installed on the top of the core computing power GPU board; The outer shell is fixedly connected to the top of the base box at its bottom, and the active heat dissipation temperature control unit, the core computing power GPU board, the CPU and external interface motherboard are all located inside the outer shell. An expansion module, located on one side of the housing, is adapted to connect with the CPU and external interface motherboard to realize interface expansion functions. The expansion module is a serial port, digital interface, or analog interface, used for plug-and-play acquisition of operating data of the central air conditioning refrigeration room and sending optimized control parameters to the underlying devices.

2. The high-efficiency data center global optimization AI computing power box according to claim 1, characterized in that, The CPU and external interface motherboard are also provided with a memory card mounting slot and operation buttons. The memory card mounting slot is used to install storage media to store data, and the operation buttons are used for device function operation control.

3. The high-efficiency data center global optimization AI computing power box according to claim 1, characterized in that, The digital interface is used to acquire the switch status signals of the equipment in the computer room and output switch control signals, while the analog interface is used to acquire the analog operating parameters of the equipment in the computer room and output analog control signals.

4. The high-efficiency data center global optimization AI computing power box according to claim 1, characterized in that, The active heat dissipation temperature control unit includes an active heat dissipation fan, which dissipates heat from inside the device through fan operation.

5. The high-efficiency data center global optimization AI computing power box according to claim 4, characterized in that, The active cooling temperature control unit also includes a temperature sensor and a temperature control chip. The temperature sensor collects internal temperature data of the device and transmits it to the temperature control chip. The temperature control chip adjusts the speed of the active cooling fan according to the temperature data to achieve adaptive cooling.

6. The high-efficiency data center global optimization AI computing power box according to claim 1, characterized in that, The USB ports on the CPU and external interface motherboard include USB 2.0 and USB 3.0 interfaces. The USB 2.0 interface is used to connect low-speed peripherals, and the USB 3.0 interface is used to connect high-speed data transmission peripherals.

Images