A multi-purpose universal data processing and control SiP microsystem
By designing a multi-purpose universal data processing and control SiP microsystem in satellite-based and missile-based computer systems, combining high-performance CPUs and reconfigurable FPGAs, the problems of poor universality and low performance of existing systems are solved, and high Integrated and low-cost microsystems for multitasking applications and high reliability requirements.
Patent Information
- Application Number
- CN202410995609.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-24
- Publication Date
- 2024-11-12
- Estimated Expiration
- 2044-07-24
AI Technical Summary
There are problems such as poor versatility, low performance, low integration and high cost in existing satellite and missile-based computer systems.
A multi-purpose universal data processing and control SiP micro system is designed, using SiP technology to integrate CPU, FPGA, high-reliable program storage unit within the CPU and FPGA internal configuration storage unit, interconnect communication through the eLBC bus, combining DDR3 memory and FLASH memory realizes a heterogeneous architecture of high-performance CPU computing and reconfigurable FPGA expansion unit, supports multi-task applications and high-reliability startup mode, and adopts plastic-packaged three-dimensional packaging for high-density integration.
Achieves multi-purpose reconstructible data processing and control of SiP microsystems, meets the needs of low-orbit satellites and missile-borne platforms, improves the system's universality and flexibility, reduces development costs, reduces area, and has high integration and low Cost, multi-purpose, high-performance miniaturization features.
Smart Images

Figure CN118519959B_ABST
Abstract
Description
Technical Field
[0001] The invention discloses a multi-purpose general-purpose data processing and control SiP microsystem, which belongs to the technical field of satellite-borne and missile-borne computer systems, and is particularly suitable for satellite-borne and missile-borne computer systems that have miniaturized, high-reliability and high-performance requirements for electronic systems. Background Art
[0002] As low-orbit constellation satellites, mass-produced missiles and other weapon equipment place increasing demands on high performance, miniaturization, high reliability, reconfigurability and low cost of computer systems, the adaptive transformation and upgrading of existing systems based on traditional information processing system architecture and finished electronic components can no longer meet application needs.
[0003] Therefore, the application of System in a Package (SiP) technology in the above fields is becoming more and more popular; as an important advanced packaging technology under the trend of system miniaturization, System in a Package technology is an important solution to achieve the miniaturization of electronic systems. SiP microsystem refers to the use of System in a Package (SiP) technology to integrate multiple integrated circuit chips into a single package, and its functions form a complete electronic system with the characteristics of high integration, small size, light weight, high reliability and high efficiency.
[0004] The Chinese authorized patent with publication number CN104731744B discloses a reconfigurable embedded computer module based on SiP, which aims to solve the requirements of missile-borne weapon systems for miniaturization, lightness, high performance and high reliability of electronic systems. The SiP microsystem adopts a minimum system SiP module based on DSP. The SiP module includes DSP, FPGA, bootstrap configuration unit and configuration data interface unit. To solve the FPGA reliability problem, the microsystem uses DSP as the configuration control unit of FPGA to realize FPGA multiple mode switching. This solution uses DSP for FPGA mode switching. The DSP system itself has no reliability reinforcement design. When the DSP is abnormal, the entire system is at risk of collapse. Moreover, this solution is only applied to missile-borne electronic systems, with strong specialization and poor versatility, and cannot meet the application in fields such as satellite-borne.
[0005] The Chinese authorized patent with publication number CN113126878B discloses a programmable signal processing module based on SiP, which integrates a large number of resources such as aerospace-grade FPGA, ADC, DAC, and refresh chip bare core into a relatively complete independent module / system based on ceramic packaging SiP technology, with the characteristics of high integration and high performance. However, this solution uses ceramic packaging to integrate expensive aerospace-grade devices as a whole through mixed processes such as high-density bonding and flip-chip welding. There are many process links and poor high-speed signal conduction performance. In addition, the microsystem almost completely relies on expensive internal radiation-resistant devices to achieve the purpose of resisting space radiation effects, which ultimately results in high overall prices, low performance, heavy weight, and rigid architecture. It cannot meet the requirements of low-orbit constellation satellites for high integration, high reliability, low cost, and reconfigurability of electronic systems, and cannot be used interchangeably with SiP microsystem chips used in missile-borne and rocket-borne fields.
[0006] In addition, it is also necessary to emphasize that low-orbit constellation satellites need to operate for a long time, and because they are constantly affected by cosmic radiation when operating in outer space, they have very high requirements for the reliability of computer systems; and missile-borne weapon systems are basically disposable products, which pay more attention to instantaneous performance bursts. However, based on considerations of volume and cost, existing technologies usually adopt a special development method to meet these two needs, resulting in poor versatility of existing technologies, wasting scientific research resources, and pushing up R&D costs. Summary of the invention
[0007] The present invention provides a multi-purpose general-purpose data processing and control SiP microsystem, which aims to solve the problems of poor versatility, low performance, low integration and high cost of existing data processing and control SiP microsystem modules used in satellite-borne, missile-borne and other fields.
[0008] To achieve the above object, the technical solution of the present invention is:
[0009] The present invention provides a multi-purpose general-purpose data processing and control SiP microsystem, comprising a CPU, a CPU data storage unit, an FPGA, a CPU internal high-reliability program storage unit and an FPGA internal configuration storage unit, a CPU external large-capacity program storage unit, an FPGA external configuration control unit, an FPGA external configuration storage unit, and an FPGA configuration switching switch, which are integrated and packaged based on the SiP technology;
[0010] The CPU is connected to the CPU data storage unit, and the CPU is connected to the FPGA;
[0011] The FPGA is provided with a CPU address / data latch unit, a CPU program reinforcement unit and a CPU startup mode selection unit implemented by IP logic;
[0012] The CPU address / data latch unit is connected to the CPU program reinforcement unit and the CPU external large-capacity program storage unit respectively, and the CPU program reinforcement unit is connected to the CPU internal high-reliability program storage unit; the CPU startup mode selection unit is connected to the CPU program reinforcement unit and the CPU external large-capacity program storage unit respectively;
[0013] The FPGA is respectively connected to the FPGA internal configuration storage unit, the FPGA external configuration control unit and the FPGA configuration switching switch, and the FPGA external configuration control unit is connected to the FPGA external configuration storage unit.
[0014] Furthermore, the CPU and the FPGA are internally interconnected and communicated via an eLBC bus.
[0015] Furthermore, the CPU data storage unit includes five 16-bit DDR3 memories, four of which are used in parallel, and the fifth DDR3 memory uses 8 bits for ECC checking.
[0016] Furthermore, the five DDR3 memories are powered separately by an external DDR3 power supply unit.
[0017] Furthermore, the high-reliability program storage unit inside the CPU is a NOR-flash memory.
[0018] Furthermore, the CPU external large-capacity program storage unit is a NOR-FLASH memory, an MRAM memory or a NAND-FLASH memory connected via an external eLBC bus.
[0019] Furthermore, the FPGA internal configuration storage unit is a SPI-FLASH storage.
[0020] Furthermore, the FPGA external configuration storage unit is a NOR-FLASH or PROM memory.
[0021] Furthermore, the SiP microsystem is also provided with a CPU data reconstruction channel and an FPGA data reconstruction channel, the CPU data reconstruction channel includes a main reconstruction channel and an emergency reconstruction channel, and the FPGA data reconstruction channel includes an internal fast start reconstruction channel and an external high reliability start reconstruction channel.
[0022] The beneficial effects achieved by the present invention are:
[0023] (1) A multi-purpose reconfigurable data processing and control SiP microsystem adopts a heterogeneous architecture of high-performance CPU computing unit + reconfigurable FPGA expansion unit. The microsystem can reconfigure system functions by loading various control IP cores or function configuration programs into the FPGA according to application needs, meeting the needs of low-orbit satellites, missile-borne platforms, etc., and solving the problem of poor versatility and flexibility of existing satellite-borne, missile-borne and other microsystem products.
[0024] (2) A multi-purpose reconfigurable data processing and control SiP microsystem is developed using high-performance COTS devices and is designed with two working modes: high reliability and high performance. When it is used for reliable applications of onboard computers, it uses the CPU and FPGA high-reliability startup method; when it is used for high-performance applications such as missile-borne computers, it uses the CPU high-performance startup and FPGA fast startup method. One microsystem chip can meet the application requirements of more than two different fields and meet the requirements of multi-task applications.
[0025] (3) A multi-purpose reconfigurable data processing and control SiP microsystem designed with main reconstruction and emergency reconstruction channels to ensure reliable reconstruction of the system.
[0026] (4) In order to improve the load drive and expansion capability of the CPU eLBC bus interface, FPGA is used as the CPU bus expansion unit. The address / data latch logic is designed inside the FPGA to latch the data line and address line and then lead them out, without the need for address data latches; the driving current of the FPGA pin is used to drive the eLBC bus pin, without the need for an external current driver, which greatly improves the CPU's eLBC bus driving capability. Using software to implement hardware circuit functions reduces costs and reduces area.
[0027] (5) A multi-purpose reconfigurable data processing and control SiP microsystem adopts comprehensive measures to improve the reliability of the microsystem in space environment. FPGA is used to implement the CPU program reinforcement function, and low-cost memory programs are reinforced by multi-chip, multi-address backup writing and "three-out-of-two" comparison and correction strategies, without the need for high-reliability space-grade memory; five non-space-grade DDR3 memories are used as data cache storage units to solve the SEU problem of the memory during space applications; among them, four DDR3s are used as data cache memories, and the fifth is used as ECC check data memory, using the CPU's own ECC function to reinforce the SEU effect during space applications; to solve the SEL problem of the memory during space applications, SiP designed an independent power supply network for the DDR3 combination, using power supply current limiting, current monitoring and other measures to solve its SEL effect, improving reliability and reducing development costs.
[0028] 6) A multi-purpose reconfigurable data processing and control SiP microsystem, which uses a plastic-encapsulated three-dimensional packaging method for high-density integration. All devices are soldered on the plastic-encapsulated substrate through a flip-chip welding process, with fewer process flows and a high yield rate. Compared with the minimum system of traditional finished devices, the area size is reduced by more than 70%. The microsystem has the characteristics of high integration and low cost. BRIEF DESCRIPTION OF THE DRAWINGS
[0029] In order to more clearly illustrate the embodiments of the present invention 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 present invention. For ordinary technicians in this field, other drawings can be obtained based on the structures shown in these drawings without paying creative work.
[0030] Figure 1 It is a module principle block diagram of the present invention.
[0031] Figure 2 It is an assembly schematic diagram of the present invention.
[0032] Figure 3 It is a schematic diagram of the metal layer of the CPU RDL microcomponent of the present invention.
[0033] Figure 4 It is a schematic diagram of the CPU RDL micro-component medium of the present invention.
[0034] Figure 5 It is a schematic diagram of the CPU chip RDL micro-component bump of the present invention.
[0035] Figure 6 It is a schematic diagram of the CPU chip RDL micro-component structure of the present invention.
[0036] Figure 7 It is a schematic diagram of the metal layer of the DDR3 chip RDL micro-component of the present invention.
[0037] Figure 8 It is a schematic diagram of the DDR3 chip RDL micro-component medium of the present invention.
[0038] Fig. 9 It is a schematic diagram of the DDR3 chip RDL micro-component bump of the present invention.
[0039] Fig.10 It is a schematic diagram of the DDR3 chip RDL micro-component structure of the present invention.
[0040] Fig.11 It is a schematic diagram of the metal layer of the NOR-FLASH chip RDL micro-component of the present invention.
[0041] Fig.12It is a schematic diagram of the NOR-FLASH chip RDL micro-component medium of the present invention.
[0042] Fig.13 It is a schematic diagram of the NOR-FLASH chip RDL micro-component bump of the present invention.
[0043] Fig.14 It is a schematic diagram of the NOR-FLASH chip RDL micro-component structure of the present invention. DETAILED DESCRIPTION
[0044] 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 only 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.
[0045] It should be noted that if the embodiments of the present invention involve directional indications (such as up, down, left, right, front, back, etc.), the directional indications are only used to explain the relative position relationship, movement status, etc. between the components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indication will also change accordingly.
[0046] In addition, if there are descriptions involving "first", "second", etc. in the embodiments of the present invention, the descriptions of "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, if the meaning of "and / or" appearing in the full text is to include three parallel schemes, taking "A and / or B" as an example, it includes scheme A, or scheme B, or a scheme that satisfies both A and B. In addition, the technical solutions between the various embodiments can be combined with each other, but it must be based on the ability of ordinary technicians in this field to implement. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.
[0047] This application proposes a multi-purpose general-purpose data processing and control SiP microsystem, which can simultaneously meet the needs of satellite-borne computer systems for high-reliability processing and missile-borne computer systems for high-performance processing through mode switching, and significantly reduce the cost of microsystem development through software IP fault-tolerant reinforcement + low-cost plastic packaging, greatly improving the versatility and practicality of the microsystem. Specifically, the present invention uses non-aerospace-grade devices with space flight experience to achieve system-level integration through low-cost and high-speed plastic packaging substrates. Through numerous reliability reinforcement logic designs and multi-working mode settings, it can be simultaneously applied to high-reliability satellite-borne computer systems and high-performance missile-borne computer systems. The microsystem module has the characteristics of multi-purpose, high integration, high performance, miniaturization, low cost, and scalability.
[0048] like Figures 1 to 14 As shown, the present invention provides a multi-purpose general-purpose data processing and control SiP microsystem, including a CPU, a CPU data storage unit, an FPGA, a CPU internal high-reliability program storage unit and an FPGA internal configuration storage unit, a CPU external large-capacity program storage unit, an FPGA external configuration control unit, an FPGA external configuration storage unit, and an FPGA configuration switching switch, which are integrated and packaged based on SiP technology;
[0049] The CPU is connected to the CPU data storage unit, and the CPU and the FPGA are interconnected and communicated via an eLBC bus;
[0050] The FPGA is provided with a CPU address / data latch unit, a CPU program reinforcement unit and a CPU startup mode selection unit implemented by IP logic;
[0051] The CPU address / data latch unit is connected to the CPU program reinforcement unit and the CPU external large-capacity program storage unit respectively, and the CPU program reinforcement unit is connected to the CPU internal high-reliability program storage unit; the CPU startup mode selection unit is connected to the CPU program reinforcement unit and the CPU external large-capacity program storage unit respectively;
[0052] The FPGA is respectively connected to the FPGA internal configuration storage unit, the FPGA external configuration control unit and the FPGA configuration switching switch, and the FPGA external configuration control unit is connected to the FPGA external configuration storage unit.
[0053] The CPU is a computing processing core, which is used to receive the application program stored in the high-reliability program storage unit inside the CPU or the large-capacity program storage unit outside the CPU, and put it into the CPU data storage unit for operation to realize various computing functions and control functions. The CPU and the FPGA are interconnected and communicated internally through the eLBC bus.
[0054] The CPU data storage unit includes five 16-bit automotive industrial-grade DDR3 memories; among them, four of the DDR3 memories are used in parallel to form a whole to increase the overall data transmission width (i.e., 16bit×4=64bit (i.e., 8 bytes) of data are processed in a single data transmission cycle), and the fifth DDR3 memory uses 8bit as ECC check (an advanced technology for memory error checking and correction) to reduce the probability of single-particle flips causing program operation abnormalities in aerospace applications. In response to the spatial single-particle latch problem of non-radiation-resistant DDR3 memories, the DDR3 memory array power supply pins are separately led out. When used, they are powered separately by an external DDR3 power supply unit and reinforced by current limiting and other measures.
[0055] Among them, the external DDR3 power supply unit uses an external independent power supply chip to provide 1.5V and 0.75V voltage power supply for the DDR3 memory. The external DDR3 power supply unit has functions such as temperature protection, current limitation, undervoltage protection, and enable shutdown. The model of this power supply chip can be an LDO chip such as TPS51200, which is an externally purchased part and adopts existing technology. The specific details are not repeated here.
[0056] The high-reliability program storage unit inside the CPU is mounted on the FPGA and connected to the CPU program reinforcement unit implemented by IP logic in the FPGA; further, the high-reliability program storage unit inside the CPU is a low-cost, radiation-resistant automotive-grade NOR-flash memory. When writing a program to the high-reliability program storage unit inside the CPU, the CPU program reinforcement unit performs automatic multi-address backup writing; when the CPU reads out a program, the CPU program reinforcement unit simultaneously reads the backup data stored in the multi-address space from the high-reliability program storage unit inside the CPU to perform automatic "three-out-of-two" verification and error correction, thereby reducing the probability of CPU system loading abnormalities caused by single-particle flips of the CPU program during aerospace applications. The program data stored in the high-reliability program storage unit inside the CPU are all reinforced by the CPU program reinforcement unit, and are also verified by the CPU program reinforcement unit when reading, which can solve the problem of program abnormalities caused by single-particle effects in aerospace, but the CPU program reinforcement unit will reduce the effective capacity of the program storage, slow down the reading and writing speeds, and reduce the overall performance; therefore, it is generally used in environments such as aerospace that require high reliability.
[0057] The CPU external large-capacity program storage unit is arranged outside the packaging structure, the CPU external large-capacity program storage unit is mounted on the FPGA, and is connected to the CPU address / data latch unit and the CPU startup mode selection unit implemented by the IP logic in the FPGA, that is, writing or reading program data does not need to pass through the CPU program reinforcement unit; specifically, the CPU external large-capacity program storage unit is implemented by a large-capacity memory such as NOR-FLASH, MRAM or NAND-FLASH externally connected through the external eLBC bus. Since the external large-capacity program storage unit of the CPU does not need to consider the volume excessively, a larger capacity can be used; and since it does not need to pass through the CPU program reinforcement unit, the effective capacity and read-write speed of the storage will not be reduced, meeting the large-scale program storage needs and achieving high-performance processing.
[0058] In order to meet the application requirements of different occasions, the present invention designs two CPU high-reliability startup modes and high-performance startup modes. The startup mode of the CPU is controlled by the CPU startup mode selection unit of the FPGA. Specifically, the FPGA logically combines the LCS startup chip select and the startup switch pin of the CPU to realize the startup mode selection. The CPU startup mode selection unit controls the CPU to read and write data from the high-reliability program storage unit inside the CPU or the large-capacity program storage unit outside the CPU according to the configuration data of the FPGA. When applied in a satellite-borne computer system, the internal high-reliability startup mode is adopted; when applied in a missile-borne computer system, the external high-performance startup mode is used.
[0059] Specifically, the two CPU startup modes are set by the high and low level states of the CPU's LCS0 chip select and other pins, which are controlled by the CPU startup mode selection unit implemented on the FPGA. During high-reliability startup, the CPU startup mode selection unit connects the CPU's LCS0 chip select signal to the nor-flash chip select signal attached to the internal FPGA, and the CPU will load the program in the internal nor-flash memory, which is reinforced by the FPGA's CPU program reinforcement unit. During high-performance startup, the startup selection unit connects the CPU's LCS0 chip select signal to the external memory and starts from the outside.
[0060] The FPGA can receive the configuration data converted by the FPGA internal configuration storage unit or the FPGA external configuration control unit, and the internal configuration and external configuration modes of the FPGA are switched by the FPGA configuration switch. The FPGA implements functions such as CPU address / data latching, CPU startup mode selection and CPU program reinforcement through logic code to solve the problem of poor load drive and expansion capability of the CPU eLBC bus interface (mainly to solve the problem of simultaneously transmitting address and data on a limited number of physical pins). The present invention uses FPGA as an expansion unit of the CPU's eLBC bus. The CPU address / data latch unit designed inside the FPGA latches and leads out the 32 data lines and 16 address lines of the CPU, without the need to use traditional address / data latch hardware circuits, which greatly improves the expansion capability of the CPU; the driving current of the FPGA pin is used to drive the eLBC bus pin, without the need to use an external current driving circuit, which greatly improves the bus driving capability of the CPU.
[0061] To meet the application requirements of different occasions, the FPGA has two configuration modes: internal fast startup and external high-reliability startup. The two modes are controlled by the FPGA configuration switching switch of the FPGA. The FPGA configuration switching switch realizes different startup modes by changing the combination of the three configuration pins M0-M2 of the FPGA.
[0062] The FPGA internal configuration storage unit is a large-capacity SPI-FLASH storage.
[0063] The FPGA external configuration storage unit is a configuration memory such as NOR-FLASH or PROM.
[0064] The FPGA external configuration control unit will monitor the running status of the FPGA. If any configuration area of the FPGA has a bit flip error, the FPGA external configuration control unit will rewrite the data of the FPGA external configuration storage unit into the configuration memory inside the FPGA to correct the error and implement read-back refresh; or the FPGA external configuration control unit will rewrite the data of the FPGA external configuration storage unit into the configuration memory inside the FPGA at regular intervals to correct the error and implement regular refresh. The FPGA external configuration control unit can be a refresh control logic ASIC chip, such as a BM501-003CBRH refresh control chip. The refresh control chip is an outsourced part, and the specific technical details are not repeated here.
[0065] The FPGA configuration switch is realized by externally pulling up the three configuration pins M0-M2 of the FPGA to 3.3V or pulling down to GND through resistors to achieve the setting of state 1 and state 0; the SARM type FPGA works in different modes according to the state combination of M0-M2 in Table 1, as shown in Table 1:
[0066] Table 1 SARM FPGA industrial model configuration description
[0067] Configuration Mode M[2:0] Data width CCLK direction Master Serial 000 x1 Output Master SPI 001 x1, x2, x4 Output Master BPI 010 x8, x16 Output Master SelectMAP 100 x8, x16 Output JTAG 101 x1 Not Applicable Slave SelectMAP 110 <![CDATA[x8,x16,x32 (1) ]]> Input <![CDATA[Slave Serial (2) ]]> 111 x1 Input
[0068] When the onboard computer system is used for high-reliability processing, the FPGA works in the Slave Select map mode, the logic combination of the M0~M2 pins is 011, and the external high-reliability startup method is used. The FPGA loads the configuration data in the FPGA external configuration storage unit through the FPGA external configuration control unit. The configuration data is reinforced by the FPGA external configuration control unit, so the data will not be too large, the system function will not be too complicated, and the loading is slow, but the reliability is high. In addition, the FPGA external configuration control unit performs a timed refresh reinforcement strategy on the FPGA to reduce the probability of program anomalies caused by single-particle flips in space during aerospace applications. When the onboard computer system is used for high-performance processing, the FPGA works in the Master-SPI mode, the logic combination of the M0~M2 pins is 100, and the internal fast startup method is used. The FPGA directly reads the data in the FPGA internal configuration storage unit, and the FPGA external configuration control unit does not need to be reinforced. The startup speed is very fast and more flexible.
[0069] The CPU startup mode selection unit is used to control whether the CPU reads data from an internal memory or an external memory, and the CPU address / data latch unit is used to solve the problem of simultaneously transmitting addresses and data on a limited number of physical pins.
[0070] In order to realize the reconstruction of SiP microsystem, data reconstruction channels are designed for both CPU and FPGA. The CPU data reconstruction channel includes a main reconstruction channel and an emergency reconstruction channel. When the CPU is operating normally, one UART interface of the CPU serves as the main reconstruction channel of the CPU. The CPU receives the reconstruction program data, which is written by the CPU to the high-reliability program storage unit inside the CPU or the large-capacity program storage unit outside the CPU where the FPGA is mounted through the eLBC bus to update the program; when the CPU is operating abnormally, one UART interface on the FPGA side serves as the emergency reconstruction channel of the CPU, and the FPGA writes the reconstruction data to the high-reliability program storage unit inside the CPU or the large-capacity program storage unit outside the CPU.
[0071] The FPGA is designed with an internal fast-start reconstruction channel and an external high-reliability startup reconstruction channel, wherein the internal fast-start reconstruction channel and the CPU emergency reconstruction channel share a UART interface. After receiving the reconstruction data, the FPGA writes the reconstruction data into the FPGA internal configuration storage unit through the SPI interface; the external high-reliability startup reconstruction channel is implemented by the UART interface of the FPGA external configuration control unit. After receiving the reconstruction data, the FPGA external configuration control unit writes the reconstruction data into the FPGA external configuration storage unit.
[0072] The entire microsystem adopts a three-dimensional flip-chip BGA welding solution, and all components are welded on the ABF substrate. Figures 3 to 14 As shown, the CPU, the CPU's internal high-reliability program storage unit, the CPU data storage unit, and the FPGA's internal configuration storage unit all use RDL rewiring technology to change the bonding fan-out to flip-chip FC fan-out, forming a series of RDL micro-components. The RDL micro-components and the FPGA bare core are welded on the front of the high-speed plastic-encapsulated substrate by flip-chip FC, and the resistors and capacitors are welded by labeling. All the front devices are attached to the heat dissipation cover plate through thermal conductive adhesive. The BGA pads on the back of the plastic-encapsulated substrate are fanned out by BGA solder balls, and the final package form is BGA1936. Finally, the size of the microsystem module is controlled within 42mm×42mm×8mm, which is more than 70% smaller than the minimum CPU and FPGA system composed of traditional finished devices.
[0073] Figure 6 The cross-sectional structure diagram of the CPU component is shown, and the specific parameters are shown in Table 2; Fig.10 The cross-sectional structure diagram of the DDR3 component is shown, and the specific parameters are shown in Table 3; Fig.14 The cross-sectional structure diagram of the NOR-FLASH component is shown, and the specific parameters are shown in Table 4.
[0074] Table 2 CPU component stacking design parameter description
[0075]
[0076] PI is the thickness of copper
[0077] Table 3 Description of stacking design parameters of DDR3 components
[0078]
[0079] Table 4 NOR-FLASH component stacking design parameter description
[0080]
[0081] PI is the thickness of copper
[0082] Specifically, the microsystem adopts the CPU+FPGA computing architecture, which integrates a high-performance processor CPU, an SRAM FPGA, a program memory NOR-FLASH, five data memories DDR3 and a number of resistors and capacitors. The CPU uses a high-performance dual-core processor with a main frequency of more than 1GHz, using SOI technology, and is naturally resistant to single-particle latch; the FPGA uses the domestic JFM7K325T SRAM FPGA; DDR3 uses a non-radiation-resistant large-capacity dynamic memory that has been verified by space flight, with a capacity of no less than 256M×16bit; NOR-FLASH uses a non-radiation-resistant large-capacity memory that has been verified by space flight, with a capacity of no less than 512Mbit.
[0083] The CPU and FPGA are interconnected through LAD<0~15>, LA<16~31>, LALE, LWE, LOE and IRQ0~1 of the CPU's eLBC bus, where LAD<0~15> is data / address multiplexing; generally, two address / data latch chips are needed in combination with the LALE latch signal to separate the data line and the address line to form data lines LD<0~15> and address lines LD<0~15>, and LA<0~15> and LA<16~31> are combined to form LA<0~31> to access the 4Gbit address space, and multiple driver chips are added to the data lines and address lines to improve the bus carrying capacity. However, due to the low switching rate of the address / data latch chip, the communication rate of the eLBC bus is severely limited. To solve this problem, the present invention uses high-performance FPGA logic code to implement CPU address / data latching, leads the latched address and data lines through FPGA pins, and uses the driving current of the FPGA pins to drive the eLBC bus pins. No external current driver is needed, which greatly improves the CPU's eLBC bus driving ability, increases the eLBC bus rate by more than 50%, and saves 2 address / data latch chips and multiple driver chips. Using software to implement hardware circuit functions reduces costs and reduces area.
[0084] The above descriptions are only optional embodiments of the present invention, and are not intended to limit the patent scope of the present invention. All equivalent structural changes made using the contents of the present invention's specification and drawings, or directly / indirectly applied in other related technical fields, are included in the patent protection scope of the present invention.
Claims
1. A multi-purpose general-purpose data processing and control SiP microsystem, characterized in that: It includes a CPU, a CPU data storage unit, an FPGA, a CPU internal high-reliability program storage unit and an FPGA internal configuration storage unit, a CPU external large-capacity program storage unit, an FPGA external configuration control unit, an FPGA external configuration storage unit, and an FPGA configuration switching switch, which are integrated and packaged based on SiP technology; The CPU is connected to the CPU data storage unit, and the CPU is connected to the FPGA; The FPGA is provided with a CPU address / data latch unit, a CPU program reinforcement unit and a CPU startup mode selection unit implemented by IP logic; The CPU address / data latch unit is connected to the CPU program reinforcement unit and the CPU external large-capacity program storage unit respectively, and the CPU program reinforcement unit is connected to the CPU internal high-reliability program storage unit; the CPU startup mode selection unit is connected to the CPU program reinforcement unit and the CPU external large-capacity program storage unit respectively; The FPGA is respectively connected to the FPGA internal configuration storage unit, the FPGA external configuration control unit and the FPGA configuration switching switch, and the FPGA external configuration control unit is connected to the FPGA external configuration storage unit.
2. The multi-purpose general-purpose data processing and control SiP microsystem according to claim 1, characterized in that: The CPU and the FPGA are interconnected and communicated internally via an eLBC bus.
3. The multi-purpose general-purpose data processing and control SiP microsystem according to claim 1, characterized in that: The CPU data storage unit includes five 16-bit DDR3 memories, four of which are used in parallel, and the fifth DDR3 memory uses 8 bits for ECC checking.
4. The multi-purpose general-purpose data processing and control SiP microsystem according to claim 3, characterized in that: The five DDR3 memories are powered individually by an external DDR3 power supply unit.
5. The multi-purpose general-purpose data processing and control SiP microsystem according to claim 1, characterized in that: The high-reliability program storage unit inside the CPU is a NOR-flash memory.
6. The multi-purpose general-purpose data processing and control SiP microsystem according to claim 1, characterized in that: The CPU external large-capacity program storage unit is a NOR-FLASH memory, an MRAM memory or a NAND-FLASH memory which is externally connected through an external eLBC bus.
7. The multi-purpose general-purpose data processing and control SiP microsystem according to claim 1, characterized in that: The FPGA internal configuration storage unit is a SPI-FLASH storage.
8. The multi-purpose general-purpose data processing and control SiP microsystem according to claim 1, characterized in that: The FPGA external configuration storage unit is a NOR-FLASH or PROM memory.
9. The multi-purpose general-purpose data processing and control SiP microsystem according to claim 1, characterized in that: The SiP microsystem is also provided with a CPU data reconstruction channel and an FPGA data reconstruction channel. The CPU data reconstruction channel includes a main reconstruction channel and an emergency reconstruction channel. The FPGA data reconstruction channel includes an internal fast-start reconstruction channel and an external high-reliability startup reconstruction channel.
