Standard computing node design system and method based on multi-core processor
By designing a standard computing node system based on multi-core processors, combining redundant fault tolerance and on-orbit reconstruction technology, the problem of insufficient performance of the satellite-on-board processing system is solved, and the requirements of high-performance real-time processing and data processing are realized to adapt to the space environment. It has good market and economic benefits.
Patent Information
- Application Number
- CN202410975716.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-19
- Publication Date
- 2024-11-19
AI Technical Summary
The main frequency of the existing mainstream on-site processing system does not exceed 200MHz, which cannot meet the needs of data computing and comprehensive data processing of massive data resources. The down-passing data analysis of traditional digital transmission channels cannot meet the real-time requirements.
Design a standard computing node system based on multi-core processors, including processing units, power supply management units and health management units. Through multi-core processors, FPGAs and clock management units, program loading, health monitoring and data processing are realized, and redundant fault tolerance is adopted. Technology and on-orbit reconstruction technology meet the needs of high-performance real-time processing.
Realize the real-time processing and rapid fusion capabilities of high-performance image data on satellites, meet the hardware needs of high-performance computers on satellites, improve the redundancy and reliability of the system, adapt to the space environment, have a wide application foundation and market benefits. .
Smart Images

Figure CN118981448A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of space embedded computer applications, and in particular relates to a standard computing node design system and method based on a multi-core processor. Background Art
[0002] my country's space application technology is constantly improving. With the continuous improvement of image resolution and measurement accuracy, the traditional approach of relying on digital transmission channels to transmit data for analysis can no longer meet the requirements of massive data transmission and real-time performance. Space products represented by space reconnaissance and situational awareness satellites have put forward application requirements for high-performance real-time processing computers.
[0003] The main frequency of current mainstream satellite-borne processing systems generally does not exceed 200 MHz, which is unable to realize data calculation and comprehensive data processing of massive data resources. Summary of the invention
[0004] The technical problem to be solved by the present invention is to provide a standard computing node design system and method based on a multi-core processor in view of the deficiencies in the above-mentioned prior art, so as to solve the technical problem of small data transmission volume of the existing mainstream satellite-borne processing system.
[0005] The present invention adopts the following technical solutions:
[0006] A standard computing node design system based on a multi-core processor, comprising:
[0007] Processing unit: used for program loading, image data receiving and calculation result sending;
[0008] Power supply management unit: used to connect with the processing unit and the health management unit to supply power to the processing unit and the health management unit respectively;
[0009] Health management unit: used to connect with the processing unit to monitor the health of the processing unit.
[0010] Furthermore, it also includes a clock management unit, which is connected to the processing unit and the health management unit respectively to provide a clock signal.
[0011] Furthermore, the processing unit includes a CPU chip, and the CPU chip is connected to a data storage device and a PROM.
[0012] Furthermore, the CPU chip adopts a multi-core processor.
[0013] Furthermore, the power supply management unit includes a first power supply management unit and a second power supply management unit, the first power supply management unit is connected to the health management unit, and the second power supply management unit is connected to the processing unit and the health management unit respectively.
[0014] Furthermore, the health management unit is used to control the power supply timing of the second power supply management unit.
[0015] Furthermore, the health management unit includes a management FPGA, and the management FPGA is connected to a FLASH and a reset circuit.
[0016] Furthermore, the FLASH includes a main FLASH and a backup FLASH, and both the main FLASH and the backup FLASH are connected to the management FPGA.
[0017] Furthermore, the management FPGA is also connected to an LVDS transceiver for program annotation.
[0018] In a second aspect, an embodiment of the present invention provides a method for designing a standard computing node based on a multi-core processor, comprising the following steps:
[0019] The processing unit and the power supply management unit are unidirectionally connected, and the power supply management unit supplies power to the processing unit;
[0020] The health management unit and the power management unit are bidirectionally connected, the power management unit supplies power to the health management unit, and the health management unit controls the power supply timing of the power management unit supplying power to the processing unit;
[0021] The health management unit and the processing unit are bidirectionally connected. The processing unit reads the output data of the health management unit for program loading; the health management unit reads the health status signal output by the processing unit to monitor the health of the processing unit.
[0022] Compared with the prior art, the present invention has at least the following beneficial effects:
[0023] The present invention provides a standard computing node design system based on a multi-core processor, including a power supply management unit, which is connected to a processing unit and a health management unit to supply power to the processing unit and the health management unit respectively, and the processing unit and the health management unit are bidirectionally connected, wherein the processing unit is used to read program information output by the health management unit to realize program loading, and the health management unit is used to read health data output by the processing unit to perform status monitoring on the health data. The present invention aims to realize a standard computing node design system based on a multi-core processor, which, as a standard computing node of a satellite-borne high-performance real-time processing computer, can meet the hardware requirements of current satellite-borne high-performance image real-time processing, and realize image data classification extraction, real-time processing and rapid fusion capabilities on board.
[0024] Furthermore, the processing unit and the health management unit are both connected to a clock management unit, and the clock management unit is used to provide a clock signal.
[0025] Furthermore, the multi-core processor is connected to a data memory and a PROM. The data memory can store protocol stack applications and intermediate process variables, etc.; the PROM is used to store a boot program.
[0026] Furthermore, the second power supply management unit in the power supply management unit is connected to the processing unit and the health management unit respectively, and the health management unit controls the power supply timing of the second power supply management unit so that the power supply timing output by the second power supply management unit to the processing unit meets the requirements.
[0027] Furthermore, the health management unit includes a management FPGA, and the FLASH is connected to the management FPGA. The FLASH includes a main FLASH and a backup FLASH, thereby improving the redundancy of the system.
[0028] Furthermore, the management FPGA is connected to an LVDS transceiver, which is used for program injection to achieve on-orbit reconstruction and meet the requirements of high product reliability and adaptability to space environment.
[0029] The present invention also provides a method for designing a standard computing node based on a multi-core processor, in which the power supply is connected to the processing unit and the health management unit through the power supply management unit, the health management unit and the processing unit are bidirectionally connected, the processing unit realizes program loading, and the health management unit realizes monitoring of the health status of the processing unit. According to actual applications, the design integrates redundant fault-tolerant technology and on-orbit reconstruction technology to meet the requirements of high reliability and adaptability to the space environment of the product. The high-performance real-time processing computer developed on this basis has a broad application base in reconnaissance and early warning satellites, and has begun to be promoted and applied, with good market and economic benefits.
[0030] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS
[0031] Figure 1 Design system block diagrams for standard computing nodes based on multi-core processors;
[0032] Figure 2 It is the circuit principle block diagram of the processor system;
[0033] Figure 3 This is the power-on reset timing diagram for a multi-core processor;
[0034] Figure 4 This is the circuit principle block diagram of the power supply management system;
[0035] Figure 5 It is the circuit principle frame of the health management system;
[0036] Figure 6 Design schematics for a standard 6U size compute node. DETAILED DESCRIPTION
[0037] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0038] In the description of the present invention, it should be understood that the terms “include” and “comprises” indicate the presence of described features, wholes, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or collections thereof.
[0039] It should be understood that, although the terms first, second, third, etc. may be used to describe preset ranges, etc. in the embodiments of the present invention, these preset ranges should not be limited to these terms. These terms are only used to distinguish preset ranges from each other. For example, without departing from the scope of the embodiments of the present invention, the first preset range may also be referred to as the second preset range, and similarly, the second preset range may also be referred to as the first preset range.
[0040] Various structural schematic diagrams of the embodiments disclosed in the present invention are shown in the accompanying drawings. These figures are not drawn to scale, and some details are magnified and some details may be omitted for the purpose of clear expression. The shapes of various regions and layers shown in the figures and the relative sizes and positional relationships therebetween are only exemplary, and may deviate in practice due to manufacturing tolerances or technical limitations, and those skilled in the art may additionally design regions / layers with different shapes, sizes, and relative positions according to actual needs.
[0041] The present invention provides a standard computing node design system based on a multi-core processor, such as Figure 1 As shown, it includes a processing unit, a power supply management unit and a health management unit, wherein the power supply management unit is unidirectionally connected to the processing unit, and the power supply management unit supplies power to the processing unit. The power supply management unit is bidirectionally connected to the health management unit, and the power supply management unit supplies power to the health management unit. At the same time, the health management unit controls the timing of the power supply management unit supplying power to the processing unit. The processing unit and the health management unit are bidirectionally connected, and the processing unit is used to receive the program of the health management unit to realize program loading. The health management unit is used to receive the health data output by the processing unit and perform health monitoring on the processing unit.
[0042] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. The components of the embodiments of the present invention described and shown in the drawings here can usually be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the present invention. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.
[0043] The present invention provides a standard computing node design system based on a multi-core processor, which includes, in addition to a processing unit, a power supply management unit and a health management unit, a clock management unit, wherein the clock management unit is respectively connected to the processing unit and the health management unit for providing a clock signal.
[0044] Specifically, Figure 2 As shown, the processing unit includes a CPU chip. The CPU chip adopts a multi-core processor, model FT-M6678. The multi-core processor is also connected to a data storage device and a PROM. The data storage device adopts a DDR3 chip. The DDR3 chip is used for external data storage, storing protocol stack applications and intermediate process variables, etc. The PROM chip is used to store the boot program. When the CPU chip is powered on, it first starts running from the boot program of the PROM, and then loads the operating system and application programs through the health management unit. The CPU software is loaded using a 16-bit data bus.
[0045] After the CPU chip is initialized, it runs the application according to the task cycle on the one hand, and outputs health status information such as heartbeat signals and dog feeding signals to the management FPGA on the other hand. If the heartbeat signal is abnormal or the dog feeding signal is not timely, it is defined as entering an abnormal state. The management FPGA triggers the CPU chip reset and reloads the program.
[0046] When the CPU software is reconstructed on-orbit, the management FPGA first takes over the read and write control of FLASH and writes the uploaded program segment into the backup FLASH storage area. When the entire software is uploaded, the ground sends instructions to trigger the CPU chip to reload the program in the backup FLASH storage area. If the telemetry information determines that the CPU is working normally, the uploaded software will be written to the main FLASH storage area.
[0047] like Figure 5As shown, the health management unit includes a management FPGA, a FLASH chip and a reset circuit. There are two FLASH chips, which are used as the main FLASH and the backup FLASH respectively to improve the redundancy of the system. The reset circuit is a power-on reset circuit to realize the reset function of the management FPGA. After power-on, it outputs a 200ms low pulse to effectively reset the FPGA logic. At the same time, the reset circuit can receive an external reset instruction to realize the reset signal output.
[0048] The management FPGA is bidirectionally connected to the multi-core processor of the processing unit. The management FPGA completes health information collection functions such as watchdog monitoring and heartbeat monitoring of the multi-core processor. The spatial environment adaptability of the multi-core processor is relatively weak, and health monitoring is an effective measure to manage the multi-core processor. When the multi-core processor enters an abnormal state, it cannot output the heartbeat signal and the dog feeding signal normally. The management FPGA can perform a reset process and reload the program to restore normal operation; it can also complete the power-off and power-on operation of the multi-core processor according to external instruction analysis, effectively eliminating the fault phenomenon.
[0049] The management FPGA can also realize the power supply timing control of the power supply module that supplies power to the processing unit in the power supply management unit, output three power change enable controls according to the power-on reset timing requirements of the multi-core processor, and read the power status signal. Under abnormal circumstances, it can autonomously complete the power-off processing to prevent abnormal power supply from damaging the processor.
[0050] The management FPGA completes the erasing, burning, and reading back of the multi-core processor program memory FLASH; in addition, the management FPGA is also connected to an LVDS transceiver, and the LVDS interface circuit can realize the reception function of soft instructions and injection data. When the software is reconstructed on-orbit, after the LVDS interface receives the injection instruction, the management FPGA writes the application program to the corresponding FLASH storage area. After the injection is successful, the reset processing can be carried out according to the instruction and the new program can be loaded to realize the on-orbit function reconstruction.
[0051] The multi-core processor and management FPGA are connected using the EMIF bus, and the multi-core processor uses the SRIO high-speed serial bus externally for high-speed image data reception and calculation result transmission.
[0052] The power supply management unit includes a first power supply management unit and a second power supply management unit, wherein the first power supply management unit is connected to the health management unit, and the second power supply management unit is connected to the processing unit and the health management unit respectively.
[0053] The power management unit is one of the cores of the standard computing node and is the source of three power conversions for the entire node. The power management unit consists of two parts: the first power management unit directly supplies power to the health management unit; the second power management unit performs three power conversions on the processing unit. The timing of the three power conversions must meet the power-on timing requirements of the multi-core processor, such as Figure 3 As shown, the enable end of the power conversion is controlled by the management FPGA. Figure 3 The requirements for each time point are shown in Table 1:
[0054] Table 1 Power-on reset timing description
[0055] time illustrate T0 IO power supply powered on T1 IO power supply is powered on T2 The core power supply and DDRIO power supply are powered on with a delay of about 20ms relative to T1. T3 Core power supply power-up completed T4 DDRIO power supply power-on completed T5 The delay relative to T4 is about 10ms, and the SERDES core and SERDES IO power supplies are powered on. T6 SERDES IO power supply power-on completed T7 The SERDES core power supply is powered on, and the POR is pulled low to start the power-on reset T8 Delay 40ms relative to T7, pull up POR, and complete power-on reset
[0056] like Figure 4 As shown, the first power supply management unit implements the three-power conversion of the management FPGA, which is always enabled. After the secondary power supply is input, it is converted into the three-power supply to power the management FPGA. The power supply of the multi-core processor is converted by the second power supply management unit, and its power supply timing is controlled by the management FPGA. According to the timing diagram, the EN enable is output, and the status indication signal PG of the conversion circuit is received. When the PG signal meets the requirements, the next one is enabled as required, otherwise the enable is turned off.
[0057] The management FPGA enables the power-on timing of the second power supply management unit, outputs three power change enable controls according to the power-on reset timing requirements of the multi-core processor, and reads the power status signal. Under abnormal circumstances, it can independently complete the power-off processing to prevent abnormal power supply from damaging the processor.
[0058] The working process of the present invention is as follows:
[0059] 1) After power-on, the power management unit first powers on the management FPGA, and the reset circuit outputs a 200ms width reset signal to the management FPGA;
[0060] 2) After the reset is completed, the management FPGA implements the logic to enable the power management unit to power on the multi-core processor circuit system and output a 200ms width reset signal to the multi-core processor;
[0061] 3) The multi-core processor loads the boot program and sequentially loads the application program, and performs the dog feeding operation during the loading process;
[0062] 4) After the CPU program is loaded successfully, the standard computing node is in working state;
[0063] 5) The standard computing node completes the calculation and processing functions of the payload image data according to the instructions, interfaces and software functions.
[0064] Another embodiment of the present invention provides a standard computing node design system based on a multi-core processor. The node is implemented using a standard 3U. A standard 6U single board can implement two nodes. The design block diagram is shown in the attached figure. Figure 6 shown.
[0065] The computing nodes are fully functional and can be used individually or in multiples depending on actual computing needs. The nodes form a computing resource pool, which serves as processing resources for the onboard high-performance computer to achieve real-time processing on orbit.
[0066] The present invention also provides a method for designing a standard computing node based on a multi-core processor, comprising the following steps:
[0067] The processing unit and the power supply management unit are unidirectionally connected, and the power supply management unit supplies power to the processing unit;
[0068] The health management unit and the power management unit are bidirectionally connected, the power management unit supplies power to the health management unit, and the health management unit controls the power supply timing of the power management unit supplying power to the processing unit;
[0069] The health management unit and the processing unit are bidirectionally connected. The processing unit reads the output data of the health management unit for program loading; the health management unit reads the health status signal output by the processing unit to perform health monitoring on the processing unit.
[0070] In summary, the present invention is a standard computing node design system and method based on a multi-core processor. According to actual applications, redundant fault-tolerant technology and on-orbit reconstruction technology are integrated in the design to meet the requirements of high reliability and adaptability to the space environment. The high-performance real-time processing computer developed on this basis has a wide range of applications in reconnaissance and early warning satellites, and has begun to be promoted and applied, with good market benefits and economic benefits.
[0071] The above contents are only for explaining the technical idea of the present invention and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution in accordance with the technical idea proposed by the present invention shall fall within the protection scope of the claims of the present invention.
Claims
1. A standard computing node design system based on multi-core processors, It is characterized in that include: Processing unit: The processing unit includes a CPU chip, which uses a multi-core processor and is used for program loading, image data reception, and calculation result transmission; Power supply management unit: used to connect with the processing unit and the health management unit to supply power to the processing unit and the health management unit respectively; Health management unit: The health management unit includes a management FPGA, which is connected to the multi-core processor of the processing unit to monitor the health of the processing unit.
2. The standard computing node design system based on multi-core processor according to claim 1, It is characterized in that It also includes a clock management unit, which is connected to the processing unit and the health management unit respectively to provide a clock signal.
3. The standard computing node design system based on multi-core processor according to claim 1, It is characterized in that The multi-core processor is connected with a data storage and a PROM.
4. The standard computing node design system based on multi-core processor according to claim 3, It is characterized in that The multi-core processor model is FT-M6678, and the data storage device uses a DDR3 chip.
5. The standard computing node design system based on multi-core processor according to claim 1, It is characterized in that The power supply management unit includes a first power supply management unit and a second power supply management unit, the first power supply management unit is connected to the health management unit, and the second power supply management unit is connected to the processing unit and the health management unit respectively.
6. The standard computing node design system based on multi-core processor according to claim 5, It is characterized in that The health management unit is used to control the power supply timing of the second power supply management unit.
7. The standard computing node design system based on multi-core processor according to claim 1, It is characterized in that The health management unit includes a management FPGA, and the management FPGA is connected to a FLASH and a reset circuit.
8. The multi-core processor-based standard computing node design system according to claim 7, It is characterized in that The FLASH includes a main FLASH and a backup FLASH, and both the main FLASH and the backup FLASH are connected to the management FPGA.
9. The standard computing node design system based on multi-core processor according to claim 7, It is characterized in that The management FPGA is also connected to an LVDS transceiver, and the LVDS transceiver is used for program injection.
10. A standard computing node design method based on multi-core processors. It is characterized in that The following steps are involved: The processing unit and the power supply management unit are unidirectionally connected, and the power supply management unit supplies power to the processing unit; The health management unit and the power management unit are bidirectionally connected, the power management unit supplies power to the health management unit, and the health management unit controls the power supply timing of the power management unit supplying power to the processing unit; The health management unit and the processing unit are bidirectionally connected. The processing unit reads the output data of the health management unit for program loading; the health management unit reads the health status signal output by the processing unit to perform health monitoring on the processing unit.
