WO2012119533A1 - 一种高端容错计算机系统及实现方法 - Google Patents
一种高端容错计算机系统及实现方法 Download PDFInfo
- Publication number
- WO2012119533A1 WO2012119533A1 PCT/CN2012/071956 CN2012071956W WO2012119533A1 WO 2012119533 A1 WO2012119533 A1 WO 2012119533A1 CN 2012071956 W CN2012071956 W CN 2012071956W WO 2012119533 A1 WO2012119533 A1 WO 2012119533A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- node
- board
- chipsets
- chipset
- router
- Prior art date
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/30—Circuit design
- G06F30/39—Circuit design at the physical level
- G06F30/398—Design verification or optimisation, e.g. using design rule check [DRC], layout versus schematics [LVS] or finite element methods [FEM]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F15/00—Digital computers in general; Data processing equipment in general
- G06F15/16—Combinations of two or more digital computers each having at least an arithmetic unit, a program unit and a register, e.g. for a simultaneous processing of several programs
- G06F15/163—Interprocessor communication
- G06F15/173—Interprocessor communication using an interconnection network, e.g. matrix, shuffle, pyramid, star, snowflake
- G06F15/17356—Indirect interconnection networks
Definitions
- the invention relates to the field of high-end computer design, and particularly relates to a high-end fault-tolerant computer system and an implementation method thereof.
- the technical problem to be solved by the present invention is to provide a high-end fault-tolerant computer system and an implementation method thereof, which can effectively realize global memory sharing, balance system transmission bandwidth and delay, and effectively solve the problem of integration reliability of multi-channel CPU systems, which is highly Technical value.
- the present invention provides a high-end fault-tolerant computer system including N single-node prototype verification systems and M cross-switch interconnect router chipsets, each of which is used to implement a cross-switch interconnect router chipset.
- the N single-node prototype verification systems are interconnected, and each of the cross-switch interconnect router chips sets is not switched.
- M and N are positive integers greater than or equal to 2, where:
- the single node prototype verification system includes:
- the computing board is a 4-way tightly coupled computing board
- the chip verification board comprises two node controller chipsets, wherein: each node controller chipset comprises two field programmable gate array (FPGA) chips, which jointly carry logic of one node controller;
- the interconnecting board includes two FPGA chips, wherein: each of the FPGA chips provides a high-speed interconnecting port, and is configured to: implement protocol interconnection between two paths in the computing board and one of the node controller chipsets.
- the above system may also have the following features:
- the 4-way tightly coupled computing board includes 4 CPUs, and the 4 CPUs are internally interconnected to share memory with each other;
- the CPUs in the N single-node prototype verification systems are interconnected by the node controller chipset and the crossbar interconnect router chipset to share memory.
- the above system may also have the following features:
- the logic of the node controller includes: Cache - Consistency Control and Interconnect Network Interface Control.
- the above system may also have the following features:
- the chip verification board has a network interface (NI);
- the plurality of single-node prototype verification systems are coupled to the crossbar interconnect router chipset via NI interfaces on respective chip verification boards.
- the above system may also have the following features:
- N 8;
- the present invention also provides an implementation method of a high-end fault-tolerant computer system, including:
- the computing board is a 4-way tightly coupled computing board
- the chip verification board includes two node controller chipsets, wherein: each node controller chipset includes two field programmable gate array (FPGA) chips, which jointly carry one node control Logic of the device;
- FPGA field programmable gate array
- the interconnection board includes two FPGA chips, wherein: each FPGA chip provides a high-speed interconnection port for implementing between two paths and one of the node controller chipsets in the computing board Protocol interconnection; Connecting two of the computing boards to one of the computing boards by one FPGA chip of the interconnecting board, and connecting the other two of the computing boards to the interconnect Another FPGA chip in the board is connected to another node controller chipset in the computing board to form a single node prototype verification system;
- Each of the single node prototype verification systems in the N single-node prototype verification systems is respectively associated with
- Each of the crossbar switch router chipsets in the M crossbar interconnect router chipset is connected, and the crossbar switch router chipsets are not switched between, and any one of the crossbar switch router chipsets implements the N connected thereto.
- the single-node prototype verification system interconnects to form an N-node 4*N-way system; M and N are positive integers greater than or equal to 2.
- the above method may also have the following features:
- the 4-way tightly coupled computing board includes 4 CPUs, and the 4 CPUs are internally interconnected to share memory with each other;
- the CPUs in the N single-node prototype verification systems are interconnected by the node controller chipset and the crossbar interconnect router chipset to share memory.
- the above method may also have the following features:
- the logic of the node controller includes: Cache - Consistency Control and Interconnect Network Interface Control.
- the above method may also have the following features:
- the chip verification board has a network interface (NI);
- the N single-node prototype verification systems are interconnected, the N single-node prototype verification systems are connected to the cross-switch interconnect router chip set via the NI interface on the respective chip verification board.
- the above method may also have the following features:
- N 8;
- a high-end fault-tolerant computer system including: N single nodes and M cross-switch interconnect router chipsets (NR ), each of the cross-switch interconnect router chipsets are used to implement the N said single node internal interconnections, and the cross switches are interconnected
- the router chipset does not perform handover, and M and N are positive integers greater than or equal to 2, where: the single node includes:
- the computing board is a 4-way tightly coupled computing board
- the node controller implements control of the 2-way CPU on the computing board.
- the 4-way tightly coupled computing board includes 4 CPUs, and the 4 CPUs are internally interconnected to share memory with each other.
- each of the N single nodes is interconnected with each other via the node controller and the crossbar interconnect router chipset to share a memory.
- the value of N is 8; the value of M is 4.
- the implementation method of the high-end fault-tolerant computer system as described above includes:
- the computing board is a 4-way tightly coupled computing board
- Interconnecting a plurality of said single nodes using an interconnect router chipset wherein N single nodes and M crossbar interconnect router chipsets (NR), each of said crossbar interconnect router chipsets are used to implement
- N single-node internal interconnections are described, and no switching is performed between the cross-switching interconnect router chip sets, and M and N are positive integers greater than or equal to 2.
- the value of N is 8; the value of M is 4.
- the invention provides a high-end fault-tolerant computer system and an implementation method thereof, which can effectively realize global memory sharing, balance system transmission bandwidth and delay, effectively solve the problem of multi-channel CPU system integration reliability, and has high technical value.
- BRIEF abstract 1 is a block diagram of a high-end fault tolerant computer system according to an embodiment of the present invention.
- FIG. 2 is a flow chart of a method for implementing a high-end fault-tolerant computer system according to an embodiment of the present invention. Preferred embodiment of the invention
- FIG. 1 there is shown a schematic diagram of a high-end fault-tolerant computer system according to an embodiment of the present invention, including N single-node prototype verification systems and M cross-switch interconnect router chipsets (NR), each of which is interconnected.
- the router chipset is configured to implement internal interconnection of the N single-node prototype verification systems to form an N-node 4*N-way system, and no switchover is performed between each of the cross-switch interconnect router chipsets. Therefore, the M sets of parallel networks formed by the M crossbar interconnect router chipsets are independent of each other, and M and N are positive integers greater than or equal to 2, where:
- the single node prototype verification system includes:
- the computing board is a 4-way tightly coupled computing board
- the chip verification board includes two node controller chipsets, wherein: each node controller chipset includes two field programmable gate array (FPGA) chips, which collectively carry the logic of one node controller;
- FPGA field programmable gate array
- the interconnect board includes two FPGA chips, wherein: each FPGA chip provides a high-speed interconnect port for implementing protocol interconnection between two channels and one of the node controller chipsets in the computing board.
- the interconnection board and the chip verification board constitute two node controllers (NC), and each node controller includes a node controller chipset on the chip verification board and a site on the interconnection board.
- a Field Programmable Gate Array (FPGA) chip implements control of two CPUs on the computing board.
- the 4-way tightly coupled computing board includes 4 CPUs, and the 4 CPUs are internally interconnected to share a memory with each other; and the CPUs in the N single-node prototype verification systems pass through the node controller
- the chipset and the crossbar interconnect router chipset are interconnected with each other, and the shared memory, that is, the 4*N CPUs inside the entire N-node 4*N system share the memory with each other.
- the logic of the node controller includes: a cache (Cache)-based control System and internet interface control.
- the two FPGA chips of the interconnect board not only realize the physical layer logic, but also ensure the initialization of the interconnect link and the signal transmission quality, and also provide various debugging methods and test methods for the verification work.
- the crossbar interconnect router chipset can pass, for example, a PCIe interface, NI
- High-speed interfaces such as (Network Interface) interface, optical interface, AMD HT interface, Intel QPI interface, and other self-developed protocol interfaces implement internal interconnection of multiple single-node prototype verification systems, realize multi-CPU system integration, and effectively realize global Memory sharing effectively solves the problem of complex verification of VLSI design in multi-channel CPU systems.
- the board-level multi-level interconnection provides a large number of debugging interfaces and verification methods for debugging verification, which greatly reduces the difficulty and complexity of verification, saves project development costs, and shortens the project development cycle.
- the invention intelligently implements large-scale node controller chipset FPGA verification by means of multi-cell interconnection, and after thorough research and trial and error, finally designs the single-node prototype verification system described in the above embodiment, which is a 4 Road single node system, where:
- the smallest calculation unit of the system When selecting the smallest calculation unit of the system, from the perspective of achieving the most optimization, it is not conducive to the plate making to increase the scale of the calculation board by 4 or more channels.
- the number of calculation boards will increase the number of calculation boards below 4 channels, which is not conducive to system-wide integration. Therefore, comprehensive After considering the 4-way tightly coupled computing board, the smallest computing unit of the system is selected.
- the logic verification unit that is, the node controller chipset, realize the system Cache-based control and the interconnection network interface control: realize the logic of one node controller chipset by using two large-capacity high-end FPGA chips, thereby effectively Ensure that the node controller chipset FPGA verifies the coverage, ensures full verification of the node controller logic, and lays the foundation for the chip ASIC implementation.
- the computing unit based on the design specifications and interfaces of the selected computing unit (ie, the 4-way tightly coupled computing board), the computing unit is provided with two node controller chipsets, which are respectively responsible for interconnecting the two CPUs.
- the 4-port interconnecting board is selected to complete the 4-port protocol interconnection between the computing unit and the logic verification unit, and 2 high-end FPGA chips are provided respectively.
- High-speed interconnect ports to ensure high-speed interconnect protocols for the entire system, And can provide a rich debugging interface and verification means for logic verification.
- the single-node prototype verification system also has good scalability, and can easily cascade multiple single-node prototype verification systems to realize multi-CPU system integration, effectively realize global memory sharing, and effectively solve multi-channel CPU system.
- the problem of verifying the complexity of large-scale integrated circuit design has a high technical value.
- the present invention is directed to a structural feature of a multi-channel computer system with high integration density, in order to improve system performance, reduce design difficulty and design complexity, and improve design reusability characteristics, based on the above-described single-node prototype verification system.
- the system interconnect router chipset is used to implement N-isomorphic single-node prototype verification system internal interconnection to form an N-node 4*N-channel system, thereby implementing the design requirements of tightly coupled shared memory.
- the symmetric structure of the N single-node prototype verification system realizes the long-distance mutual access between the system processors, guarantees the performance of the tightly coupled shared memory system, and the symmetric isomorphism also greatly reduces the multi-channel system.
- the design complexity increases the reusability of the design and saves a lot of project development time for design verification and board level debugging, shortening the project development cycle.
- M cross-switch interconnect router chip sets are parallel isomorphic extension.
- M isomorphic N-node 4*N-channel system because each of the cross-switch interconnect router chipsets does not make a transition, thereby ensuring M N-nodes formed by M cross-switch interconnect router chipsets *N-channel systems are independent of each other.
- This fault-tolerant mechanism greatly improves the reliability of high-end computer systems and improves the fault-tolerant mechanism of their application in special fields.
- the present invention is also directed to the characteristics of the 32-way high-end system, fully considering system efficiency, design complexity, and cost. From the perspective of achieving optimization, after trial and error and comparison, it provides a most reasonable 32.
- the high-end fault-tolerant computer system that is, selects the eight single-node prototype verification systems, and uses the above-mentioned manner to form an 8-node 32-channel prototype verification system, and uses four cross-switches to interconnect the router chipset parallel isomorphism to expand 4 An 8-node 32-way system, as shown in Figure 1. This is based on the requirements of system reliability.
- each 8-node 32-channel system based on the single-node 4-way prototype verification system, analyze the structural characteristics of the interconnected chipset, as well as the protocol processing capability and processing mechanism, and extend the implementation of the entire 32-channel system.
- the 16 node controller chipsets in the system implement Cache-induced control and interconnected network interface control of the entire system.
- the embodiment of the present invention further provides a method for implementing a high-end fault-tolerant computer system. As shown in FIG. 2, the method includes:
- Step S201 Select a computing board, where the computing board is a 4-way tightly coupled computing board;
- the 4-way tightly coupled computing board includes four CPUs, and the four CPUs are internally interconnected to share memory with each other.
- Step S202 Select a chip verification board, where the chip verification board includes two node controller chip sets, wherein: each node controller chipset includes two field programmable gate array (FPGA) chips, and jointly carries one The logic of the node controller;
- each node controller chipset includes two field programmable gate array (FPGA) chips, and jointly carries one The logic of the node controller;
- FPGA field programmable gate array
- the logic of the node controller includes: a cache (Cache)-induced control and an interconnected network interface control.
- cache cache-induced control
- interconnected network interface control interconnected network interface control
- Step S203 Select an interconnection board, where the interconnection board includes two FPGA chips, wherein: each FPGA chip provides a high-speed interconnection port for implementing two paths and one of the node controller chips in the computing board. Protocol interconnection between groups;
- Step S204 connecting two paths of the computing board to one node controller chipset in the computing board via one of the interconnecting boards, and inserting another two paths in the computing board Another FPGA chip in the interconnection board is connected to another node controller chipset in the computing board to form a single node prototype verification system;
- Step S205 Connect each single node prototype verification system of the N single-node prototype verification systems to each cross-switch interconnect router chipset of the M cross-switch interconnect router chipsets, and interconnect the cross-switches. No switching between the router chipsets, any one of the crossbar interconnect router chipsets realizes the internal interconnection of the N single-node prototype verification systems connected thereto Connected to form an N-node 4*N-way system; M, N are positive integers greater than or equal to 2.
- the CPUs in the N single-node prototype verification systems are interconnected and shared with each other via the node controller chipset and the crossbar interconnect router chipset.
- the memory that is, the 4*N CPUs inside the entire N-node 4*N system share the memory with each other.
- the crossbar interconnect router chipset can pass a high-speed interface such as a PCIe interface, a NI (Network Interface) interface, an optical interface, an AMD HT interface, an Intel QPI interface, and other self-developed protocol interfaces.
- a high-speed interface such as a PCIe interface, a NI (Network Interface) interface, an optical interface, an AMD HT interface, an Intel QPI interface, and other self-developed protocol interfaces.
- a high-speed interface such as a PCIe interface, a NI (Network Interface) interface, an optical interface, an AMD HT interface, an Intel QPI interface, and other self-developed protocol interfaces.
- step S205 eight single-node prototype verification systems can be selected to form an 8-node 32-channel prototype verification system, and four cross-switches are used to interconnect the router chipset.
- Parallel isomorphism extends four 8-node 32-way systems to provide reliability and fault tolerance for 32-channel systems in the most reasonable manner.
- the structure similar to that shown in FIG. 1 is still used, and the parallel network structure design of the high-end fault-tolerant computer system mainly includes: a single 32-channel system structure design (1), and 4 sets of parallel network structure fault-tolerant structure design (2) ).
- the 4-way computing board is used as the minimum computing module, and the 32-channel system uses 8 sets.
- the computing unit of the structure, the internal interconnection of the processor is realized inside the computing board, and the design requirement of the tightly coupled shared memory is realized.
- the single computing board is equipped with two node controller chipset structures, and each node is connected with a node controller to realize interconnection control with the entire system, and realize Its Cache-induced control, a total of 16 node controller chipsets of the 32-channel system control the interconnection of the system interconnect cross-switching interconnect router chipset through their respective interconnected network interfaces, thereby realizing the 32-channel tightly coupled shared memory high-end fault-tolerant computer system.
- high-end fault-tolerant computer systems include:
- the switch interconnect router chipset is configured to implement the internal interconnection of the N single nodes to form an N-node 4*N system, and the cross-switch interconnect router chip sets are not switched between, Ensure that the M sets of parallel networks formed by the M crossbar switch router chips are independent of each other, and M and N are positive integers greater than or equal to 2, wherein: the single nodes include:
- the computing board is a 4-way tightly coupled computing board
- the node controller implements control of the 2-way CPU on the computing board.
- the 4-way tightly coupled computing board includes 4 CPUs, and the 4 CPUs are internally interconnected to share a memory with each other; and the node controller and the crossbar switch between the CPUs of the N single nodes
- the interconnect router chipsets are interconnected with each other, and the shared memory, that is, the 4*N CPUs inside the entire N-node 4*N system share the memory with each other.
- the implementation method of the foregoing high-end fault-tolerant computer system includes:
- the computing board is selected as a 4-way tightly coupled computing board; wherein the 4-way tightly coupled computing board includes four CPUs, and the four CPUs are internally interconnected to share memory with each other.
- Interconnecting a plurality of said single nodes using an interconnect router chipset wherein N single nodes and M crossbar interconnect router chipsets (NR), each of said crossbar interconnect router chipsets are used to implement
- N single-node internal interconnections are described, and no switching is performed between the cross-switching interconnect router chip sets, and M and N are positive integers greater than or equal to 2.
- the CPUs of the N single-nodes are interconnected by the cross-switch interconnect router chipset, and the shared memory, that is, the entire N-node 4*N
- the 4*N CPUs inside the road system share the memory with each other.
- N 8; the value of M is 4.
- the beneficial effects of the above embodiments are as follows:
- the characteristics of the single-layer 32-channel system structure design mainly refer to the system structure design stage. According to the system scale and performance analysis, the whole computer system is designed based on the 4-way computing board system, and the calculation board internal processing is realized.
- the internal interconnection of the device meets the requirements of its shared memory design.
- the 2-way processor realizes the processing and control of the remote processor access message through a node controller.
- the entire 32-channel system uses 16 node controllers to implement the entire system.
- the 16 node controller chipset in the system is connected to the crossbar interconnection router chipset through an interconnection network port, and the whole system is interconnected by the crossbar interconnection router chipset; 4 sets of parallel network structures
- the characteristics of the fault-tolerant structure design mainly refer to the parallel expansion of four cross-switch interconnect router chipsets to four sets of network structures based on the design of a single 32-channel system structure.
- the four cross-switch interconnect router chips do not do. Transfer, to ensure that the system can still use the unique non-faulty network to communicate when three sets of networks fail, greatly improving the system reliability and fault tolerance mechanism.
- the design of the computer system architecture realizes the system integration of multiple CPUs, effectively realizes the global memory sharing, equalizes the transmission bandwidth and delay of the system, and effectively solves the problem of integration reliability of the multi-channel CPU system.
- a high-end fault-tolerant computer system and an implementation method provided by the embodiments of the present invention can effectively implement global memory sharing, balance system transmission bandwidth and delay, effectively solve the problem of integrated reliability of multi-channel CPU systems, and have high technical value.
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Hardware Design (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- Hardware Redundancy (AREA)
- Multi Processors (AREA)
- Memory System Of A Hierarchy Structure (AREA)
Description
一种高端容错计算机系统及实现方法
技术领域
本发明涉及高端计算机设计领域, 具体涉及一种高端容错计算机系统及 实现方法。
背景技术
随着计算机技术的飞速发展, 为了满足经济社会发展的需要, 高可靠的 计算机系统成为制约社会发展关键领域的瓶颈之一。 金融、 电信等关键领域 对计算机系统的可靠性要求极高, 因此需要构建高可靠的庞大的多路计算机 系统, 以便更好适应当今各领域的应用需求, 但是另一方面也陷入了多路计 算机系统互联带来的系统可靠性的技术难题中, 计算机系统的处理器数量不 断增多, 越来越高的集成密度使系统可靠性设计难度加大。
发明内容
本发明要解决的技术问题是,提供一种高端容错计算机系统及实现方法, 能够有效实现全局存储器共享, 均衡系统传输带宽和延迟, 有效解决了多路 CPU系统集成可靠性的问题, 具有很高的技术价值。
为了解决上述技术问题, 本发明提出一种高端容错计算机系统, 包括 N 个单结点原型验证系统和 M个交叉开关互联路由器芯片组,每个所述交叉开 关互联路由器芯片组均用于实现所述 N个单结点原型验证系统之间互联, 各 所述交叉开关互联路由器芯片组之间不做转接, M, N均为大于等于 2的正 整数, 其中:
所述单结点原型验证系统包括:
计算板, 为一 4路紧耦合计算板;
芯片验证板, 包括 2个结点控制器芯片组, 其中: 每一结点控制器芯片 组包括 2个现场可编程门阵列(FPGA )芯片, 共同承载 1个结点控制器的逻 辑;
互联板, 包括 2个 FPGA芯片, 其中: 每个 FPGA芯片提供一个高速互 联端口, 设置为: 实现所述计算板中的 2路与 1个所述结点控制器芯片组之 间的协议互联。
优选地, 上述系统还可具有以下特点:
所述 4路紧耦合计算板包括 4个 CPU, 所述 4个 CPU内部互联,彼此共 享存储器;
所述 N个单结点原型验证系统中的各 CPU之间经所述结点控制器芯片组 和交叉开关互联路由器芯片组彼此互联, 共享存储器。
优选地, 上述系统还可具有以下特点:
所述结点控制器的逻辑包括: 高速緩冲存储器(Cache ) —致性控制和互 联网络接口控制。
优选地, 上述系统还可具有以下特点:
所述芯片验证板具有网络接口 (NI ) ;
所述多个单结点原型验证系统经各自的芯片验证板上的 NI接口与所述 交叉开关互联路由器芯片组相连。
优选地, 上述系统还可具有以下特点:
N的取值为 8;
M的取值为 4。
为了解决上述技术问题, 本发明还提出一种高端容错计算机系统的实现 方法, 包括:
选择计算板, 所述计算板为一 4路紧耦合计算板;
选择芯片验证板, 所述芯片验证板包括 2个结点控制器芯片组, 其中: 每一结点控制器芯片组包括 2个现场可编程门阵列(FPGA )芯片, 共同承载 1个结点控制器的逻辑;
选择互联板, 所述互联板包括 2个 FPGA芯片, 其中: 每个 FPGA芯片 提供一个高速互联端口, 用于实现所述计算板中的 2路与 1个所述结点控制 器芯片组之间的协议互联;
将所述计算板中的 2路经所述互联板中的 1个 FPGA芯片连接至所述计 算板中的一个结点控制器芯片组, 将所述计算板中的另外 2路经所述互联板 中的另外 1个 FPGA芯片连接至所述计算板中的另外一个结点控制器芯片组, 从而组成一个单结点原型验证系统;
将 N个所述单结点原型验证系统中的每一个单结点原型验证系统分别与
M个交叉开关互联路由器芯片组中的每一个交叉开关互联路由器芯片组相 连, 各交叉开关互联路由器芯片组之间不做转接, 任意一个交叉开关互联路 由器芯片组实现与其连接的 N个所述单结点原型验证系统内部互联, 以构成 一个 N结点 4*N路系统; M, N均为大于等于 2的正整数。
优选地, 上述方法还可具有以下特点:
所述 4路紧耦合计算板包括 4个 CPU, 所述 4个 CPU内部互联,彼此共 享存储器;
所述 N个单结点原型验证系统中的各 CPU之间经所述结点控制器芯片组 和交叉开关互联路由器芯片组彼此互联, 共享存储器。
优选地, 上述方法还可具有以下特点:
所述结点控制器的逻辑包括: 高速緩冲存储器(Cache ) —致性控制和互 联网络接口控制。
优选地, 上述方法还可具有以下特点:
所述芯片验证板具有网络接口 (NI ) ;
在将所述 N个所述单结点原型验证系统互联时,是将所述 N个单结点原 型验证系统经各自的芯片验证板上的 NI接口与所述交叉开关互联路由器芯 片组相连。
优选地, 上述方法还可具有以下特点:
N的取值为 8;
M的取值为 4。
一种高端容错计算机系统, 包括:
N个单结点和 M个交叉开关互联路由器芯片组(NR ) , 每个所述交叉 开关互联路由器芯片组均用于实现所述 N个所述单结点内部互联, 各所述交 叉开关互联路由器芯片组之间不做转接, M, N均为大于等于 2的正整数, 其中: 所述单结点包括:
计算板, 为一 4路紧耦合计算板;
结点控制器, 实现对计算板上的 2路 CPU的控制。
优选地,所述 4路紧耦合计算板包括 4个 CPU,所述 4个 CPU内部互联, 彼此共享存储器。
优选地,所述 N个单结点中的各 CPU之间经所述结点控制器和交叉开关 互联路由器芯片组彼此互联, 共享存储器。
优选地, N的取值为 8; M的取值为 4。
如上所述的高端容错计算机系统的实现方法, 包括:
选择计算板, 所述计算板为一 4路紧耦合计算板;
选择结点控制器, 实现对计算板上的 2路 CPU的控制, 组成一个单结点 互联;
使用互联路由器芯片组将多个所述单结点互联, 其中, N个单结点和 M 个交叉开关互联路由器芯片组(NR ) , 每个所述交叉开关互联路由器芯片组 均用于实现所述 N个所述单结点内部互联, 各所述交叉开关互联路由器芯片 组之间不做转接, M, N均为大于等于 2的正整数。
优选地, N的取值为 8; M的取值为 4。
本发明实施例提供的一种高端容错计算机系统及实现方法, 能够有效实 现全局存储器共享, 均衡系统传输带宽和延迟, 有效解决了多路 CPU系统集 成可靠性的问题, 具有很高的技术价值。 附图概述
图 1是本发明实施例一种高端容错计算机系统方框图;
图 2是本发明实施例一种高端容错计算机系统的实现方法流程图。 本发明的较佳实施方式
下面将结合附图来详细说明本发明实施方案。
参见图 1 , 该图示出了本发明实施例一种高端容错计算机系统示意图, 包括 N个单结点原型验证系统和 M个交叉开关互联路由器芯片组(NR ) , 每个所述交叉开关互联路由器芯片组均用于实现所述 N个所述单结点原型验 证系统内部互联, 以构成一个 N结点 4*N路系统, 各所述交叉开关互联路由 器芯片组之间不做转接,从而保证 M个交叉开关互联路由器芯片组构成的 M 套平行网络彼此独立, M, N均为大于等于 2的正整数, 其中:
所述单结点原型验证系统包括:
计算板, 为一 4路紧耦合计算板;
芯片验证板, 包括 2个结点控制器芯片组, 其中: 每一结点控制器芯片 组包括 2个现场可编程门阵列(FPGA )芯片, 共同承载 1个结点控制器的逻 辑;
互联板, 包括 2个 FPGA芯片, 其中: 每个 FPGA芯片提供一个高速互 联端口, 用于实现所述计算板中的 2路与 1个所述结点控制器芯片组之间的 协议互联。
其中, 在图 1中, 互联板和芯片验证板构成 2个结点控制器(NC ) , 每 个结点控制器包括芯片验证板上的一个结点控制器芯片组和互联板上的一个 现场可编程门阵列 ( Field Programmable Gate Array, FPGA ) 芯片, 实现对计 算板上的 2路 CPU的控制。
其中, 所述 4路紧耦合计算板包括 4个 CPU , 所述 4个 CPU内部互联, 彼此共享存储器;所述 N个单结点原型验证系统中的各 CPU之间经所述结点 控制器芯片组和交叉开关互联路由器芯片组彼此互联, 共享存储器, 即整个 N结点 4*N路系统内部的 4*N个 CPU彼此共享存储器。
其中, 所述结点控制器的逻辑包括: 高速緩冲存储器( Cache )—致性控
制和互联网络接口控制。
其中, 互联板的 2个 FPGA芯片除了实现物理层逻辑, 保证互联链路握 手初始化以及信号传输质量外, 还可以为验证工作提供了多种调试手段和测 试方法。
其中, 所述交叉开关互联路由器芯片组可以通过例如 PCIe接口、 NI
( Network Interface )接口、 光接口、 AMD HT接口、 Intel QPI接口、 以及自 主开发的其他协议接口等高速接口实现多个所述单结点原型验证系统内部互 联, 实现多 CPU系统集成, 有效实现全局存储器共享, 有效解决了多路 CPU 系统中超大规模集成电路设计验证复杂度的问题, 在保证计算机系统验证和 高端服务器关键芯片组验证的基础上, 实现了高端计算机系统性能的要求, 具有很高的技术价值。 同时板级多级互联为调试验证提供了大量的调试接口 和验证手段, 大大减少了验证难度和复杂度, 节约了项目研制开销, 缩短了 项目研制周期。
本发明巧妙地釆用多单元板互联方式实现大规模结点控制器芯片组 FPGA验证, 经过深入研究和反复试验, 最终设计出上述实施例所述的单结 点原型验证系统, 其为一 4路单结点系统, 其中:
在选择系统最小的计算单元时, 从实现最的优化角度出发, 充分考虑到 4路以上会增加计算板规模不利于制版, 4路以下会增加计算板数量, 不利于 全系统集成, 因此, 综合考虑后选择 4路紧耦合计算板作为系统最小的计算 单元。
自主设计逻辑验证单元, 即结点控制器芯片组, 实现系统 Cache—致性 控制和互联网络接口控制: 通过釆用 2片大容量高端 FPGA芯片实现 1个结 点控制器芯片组逻辑, 从而有效保证结点控制器芯片组 FPGA验证覆盖率, 确保对结点控制器逻辑进行全面验证, 为芯片 ASIC 实现打下基础。 其中, 基于选定的计算单元(即 4路紧耦合计算板) 的设计规格和接口, 需为所述 计算单元配备 2个结点控制器芯片组, 分别负责两路 CPU的互联。
基于选定的计算单元(即 4路紧耦合计算板) 的设计规格和接口, 选择 4端口互联单板完成计算单元与逻辑验证单元的 4端口协议互联, 以 2片高 端 FPGA芯片分别提供 2个高速互联端口, 保证整个系统的高速互联协议,
且能为逻辑验证提供丰富的调试接口和验证手段。
上述单结点原型验证系统还具有很好的扩展性, 能够方便地将多个上述 单结点原型验证系统级联, 从而实现多 CPU系统集成, 有效实现全局存储器 共享, 有效解决多路 CPU系统中超大规模集成电路设计验证复杂度的问题, 具有很高的技术价值。
进一步地, 本发明针对多路计算机系统集成密度高的结构特点, 为提高 系统性能, 减少设计难度和设计复杂度, 提高设计的可复用特性, 在上述单 结点原型验证系统的基础上, 釆用系统互联路由器芯片组实现 N个同构的单 结点原型验证系统内部互联, 以构成一个 N结点 4*N路系统, 从而实现紧耦 合共享存储器的设计要求。 同时, N个单结点原型验证系统对称的同构结构 实现了系统处理器间互访同步长, 保证了紧耦合共享存储器系统的性能, 并 且对称的同构结构也大大减少了多路系统的设计复杂度, 提高了设计的可重 用性, 并且为设计验证和板级调试工作节省了大量项目研制时间, 缩短了项 目研制周期。
进一步地, 本发明充分考虑多路处理器系统集成的特点和系统高可靠的 设计要求, 在上述 N结点 4*N路系统的基础上, 使用 M个交叉开关互联路 由器芯片组平行同构扩展出 M个同构的 N结点 4*N路系统, 由于各所述交 叉开关互联路由器芯片组之间不做转接,从而保证 M个交叉开关互联路由器 芯片组构成的 M个 N结点 4*N路系统彼此独立, 这种容错机制大大提高了 高端计算机系统的可靠性, 提高了其特殊领域应用的容错机制。
较佳地, 本发明还针对 32路的高端系统的特点, 充分考虑系统效率、 设 计复杂性, 以及成本, 从实现最优化的角度出发, 经反复试验及比较后, 提 供一种最合理的 32路高端容错计算机系统, 即, 选择 8个所述单结点原型验 证系统, 釆用上述方式组成 8结点 32路原型验证系统, 并使用 4个交叉开关 互联路由器芯片组平行同构扩展出 4个 8结点 32路系统, 如图 1所示。 这是 根据系统可靠性的要求提出来的, 经反复试验比较后, 发现平行同构扩展出 的 8结点 32路原型验证系统的数量小于 4套时, 系统可靠性较低, 大于 4套 可靠性会越高, 但是系统规模太大, 对机箱、 散热等架构设计带来难度, 综 合考虑可以选择 4套为最佳选择。 图 1中, 包括 4套平行同构的 8结点 32路
系统, 结合系统自适应路由机制, 保证系统在 3套网络发生故障时, 仍能够 使用唯一的非故障网络进行通信,从而大大提高了高端计算机系统的可靠性, 提高了其特殊领域应用的容错机制; 对于每一个 8结点 32路系统而言, 则是 以单结点 4路原型验证系统为基础, 分析互联芯片组结构特点, 以及协议处 理能力和处理机制, 扩展实现的, 整个 32路系统中的 16个结点控制器芯片 组实现整个系统的 Cache—致性控制和互联网络接口控制。
本发明实施例还提供了一种高端容错计算机系统的实现方法, 如图 2所 示, 包括:
步骤 S201 : 选择计算板, 所述计算板为一 4路紧耦合计算板;
所述 4路紧耦合计算板包括 4个 CPU, 所述 4个 CPU内部互联,彼此共 享存储器。
步骤 S202: 选择芯片验证板, 所述芯片验证板包括 2个结点控制器芯片 组,其中: 每一结点控制器芯片组包括 2个现场可编程门阵列(FPGA )芯片, 共同承载 1个结点控制器的逻辑;
所述结点控制器的逻辑包括: 高速緩冲存储器(Cache )—致性控制和互 联网络接口控制。
步骤 S203: 选择互联板, 所述互联板包括 2个 FPGA芯片, 其中: 每个 FPGA芯片提供一个高速互联端口, 用于实现所述计算板中的 2路与 1个所 述结点控制器芯片组之间的协议互联;
步骤 S204: 将所述计算板中的 2路经所述互联板中的 1个 FPGA芯片连 接至所述计算板中的一个结点控制器芯片组, 将所述计算板中的另外 2路经 所述互联板中的另外 1个 FPGA芯片连接至所述计算板中的另外一个结点控 制器芯片组, 从而组成一个单结点原型验证系统;
步骤 S205:将 N个所述单结点原型验证系统中的每一个单结点原型验证 系统分别与 M个交叉开关互联路由器芯片组中的每一个交叉开关互联路由器 芯片组相连, 各交叉开关互联路由器芯片组之间不做转接, 任意一个交叉开 关互联路由器芯片组实现与其连接的 N个所述单结点原型验证系统内部互
联, 以构成一个 N结点 4*N路系统; M, N均为大于等于 2的正整数。
在所述构成的一个 N结点 4*N路系统中, N个单结点原型验证系统中的 各 CPU之间经所述结点控制器芯片组和交叉开关互联路由器芯片组彼此互 联, 共享存储器, 即整个 N结点 4*N路系统内部的 4*N个 CPU彼此共享存 储器。
其中,在执行步骤 S205时, 所述交叉开关互联路由器芯片组可以通过例 如 PCIe接口、 NI ( Network Interface )接口、光接口、 AMD HT接口、 Intel QPI 接口、 以及自主开发的其他协议接口等高速接口实现多个所述单结点原型验 证系统内部互联, 实现多 CPU系统集成, 有效实现全局存储器共享。
其中, 针对 32路的高端系统的特点, 在执行步骤 S205时, 可以选择 8 个所述单结点原型验证系统, 组成 8结点 32路原型验证系统, 并使用 4个交 叉开关互联路由器芯片组平行同构扩展出 4个 8结点 32路系统,以最合理的 方式提供 32路系统的可靠性和容错性。
在另一实施例中, 仍然釆用类似图 1所示的结构, 高端容错计算机系统 平行网络结构设计主要包括: 单套 32路系统结构设计(1 ) 、 4套平行网络 结构容错结构设计(2 ) 。 根据多路计算机系统集成密度高的结构特点, 为实现系统容错机制, 提 高系统可靠性, 以及设计的可复用, 釆用 4路计算板为系统最少计算模块, 32路系统釆用 8套同构的计算单元, 计算板内部实现处理器内部互联, 实现 紧耦合共享存储器的设计要求。 根据处理器(即 CPU ) 配置结构的特点, 釆 用单计算板配备两个结点控制器芯片组的结构, 每两个处理器连接一个结点 控制器实现与整个系统的互联控制, 并且实现其 Cache—致性控制, 32路系 统总共 16 个结点控制器芯片组通过各自的互联网络接口控制与系统互联交 叉开关互联路由器芯片组的互联,从而实现 32路紧耦合共享存储器高端容错 计算机系统。 具体地, 高端容错计算机系统包括:
N个单结点和 M个交叉开关互联路由器芯片组(NR ) , 每个所述交叉
开关互联路由器芯片组均用于实现所述 N个所述单结点内部互联, 以构成一 个 N结点 4*N路系统, 各所述交叉开关互联路由器芯片组之间不做转接, 从 而保证 M个交叉开关互联路由器芯片组构成的 M套平行网络彼此独立, M , N均为大于等于 2的正整数, 其中: 所述单结点包括:
计算板, 为一 4路紧耦合计算板;
结点控制器, 实现对计算板上的 2路 CPU的控制。
其中, 所述 4路紧耦合计算板包括 4个 CPU , 所述 4个 CPU内部互联, 彼此共享存储器;所述 N个单结点中的各 CPU之间经所述结点控制器和交叉 开关互联路由器芯片组彼此互联, 共享存储器, 即整个 N结点 4*N路系统内 部的 4*N个 CPU彼此共享存储器。
相应地, 上述高端容错计算机系统的实现方法, 包括:
选择计算板, 所述计算板为一 4路紧耦合计算板; 其中, 所述 4路紧耦 合计算板包括 4个 CPU, 所述 4个 CPU内部互联, 彼此共享存储器。
选择结点控制器, 实现对计算板上的 2路 CPU的控制。
使用互联路由器芯片组将多个所述单结点互联, 其中, N个单结点和 M 个交叉开关互联路由器芯片组(NR ) , 每个所述交叉开关互联路由器芯片组 均用于实现所述 N个所述单结点内部互联, 各所述交叉开关互联路由器芯片 组之间不做转接, M, N均为大于等于 2的正整数。
在所述构成的一个 N结点 4*N路系统中, N个单结点中的各 CPU之间 经所述交叉开关互联路由器芯片组彼此互联,共享存储器,即整个 N结点 4*N 路系统内部的 4*N个 CPU彼此共享存储器。
N的取值为 8; M的取值为 4。
上述实施例的有益效果是: 单套 32路系统结构设计的特性, 主要是指系 统结构设计阶段根据系统规模以及性能分析, 设计整个计算机系统以 4路计 算板系统为基础, 计算板内部实现处理器内部互联, 满足其共享存储器设计 要求, 2路处理器通过 1个结点控制器实现对远程处理器访问报文的处理和 控制, 整个 32路系统釆用 16个结点控制器实现整个系统的 Cache—致性控
制和互联网络接口控制, 系统中 16个结点控制器芯片组各通过一个互联网络 端口连接到交叉开关互联路由器芯片组, 由交叉开关互联路由器芯片组实现 整个系统的互联; 4 套平行网络结构容错结构设计的特性, 主要是指在单套 32路系统结构设计的基础上釆用 4个交叉开关互联路由器芯片组平行扩展为 4套网络结构, 4个交叉开关互联路由器芯片组之间不做转接, 保证系统在 3 套网络发生故障时, 仍能够使用唯一的非故障网络进行通信, 大大提高了系 统的可靠性和容错机制。这种计算机系统体系结构的设计实现了多路 CPU的 系统集成, 有效实现了全局存储器共享, 均衡系统传输带宽和延迟, 有效解 决了多路 CPU系统集成可靠性的问题。
当然, 本发明还可有其他多种实施例, 在不背离本发明精神及其实质的 些相应的改变和变形都应属于本发明所附的权利要求的保护范围。
工业实用性
本发明实施例提供的一种高端容错计算机系统及实现方法, 能够有效实 现全局存储器共享, 均衡系统传输带宽和延迟, 有效解决了多路 CPU系统集 成可靠性的问题, 具有很高的技术价值。
Claims
1、 一种高端容错计算机系统, 包括 N个单结点原型验证系统和 M个交 叉开关互联路由器芯片组, 每个所述交叉开关互联路由器芯片组均用于实现 所述 N个单结点原型验证系统之间互联, 各所述交叉开关互联路由器芯片组 之间不做转接, M, N均为大于等于 2的正整数, 其中:
所述单结点原型验证系统包括:
计算板, 为一 4路紧耦合计算板;
芯片验证板, 包括 2个结点控制器芯片组, 其中: 每一结点控制器芯片 组包括 2个现场可编程门阵列(FPGA )芯片, 共同承载 1个结点控制器的逻 辑;
互联板, 包括 2个 FPGA芯片, 其中: 每个 FPGA芯片提供一个高速互 联端口, 设置为: 实现所述计算板中的 2路与 1个所述结点控制器芯片组之 间的协议互联。
2、 如权利要求 1所述的系统, 其中,
所述 4路紧耦合计算板包括 4个 CPU , 所述 4个 CPU内部互联,彼此共 享存储器;
所述 N个单结点原型验证系统中的各 CPU之间经所述结点控制器芯片组 和交叉开关互联路由器芯片组彼此互联, 共享存储器。
3、 如权利要求 1所述的系统, 其中,
所述结点控制器的逻辑包括: 高速緩冲存储器(Cache )—致性控制和互 联网络接口控制。
4、 如权利要求 1所述的系统, 其中,
所述芯片验证板具有网络接口 (NI ) ;
所述多个单结点原型验证系统经各自的芯片验证板上的 NI接口与所述 交叉开关互联路由器芯片组相连。
5、 如权利要求 1 -4中任何一项所述的系统, 其中,
N的取值为 8;
M的取值为 4。
6、 一种高端容错计算机系统的实现方法, 包括:
选择计算板, 所述计算板为一 4路紧耦合计算板;
选择芯片验证板, 所述芯片验证板包括 2个结点控制器芯片组, 其中: 每一结点控制器芯片组包括 2个现场可编程门阵列(FPGA )芯片, 共同承载 1个结点控制器的逻辑;
选择互联板, 所述互联板包括 2个 FPGA芯片, 其中: 每个 FPGA芯片 提供一个高速互联端口, 用于实现所述计算板中的 2路与 1个所述结点控制 器芯片组之间的协议互联;
将所述计算板中的 2路经所述互联板中的 1个 FPGA芯片连接至所述计 算板中的一个结点控制器芯片组, 将所述计算板中的另外 2路经所述互联板 中的另外 1个 FPGA芯片连接至所述计算板中的另外一个结点控制器芯片组, 从而组成一个单结点原型验证系统;
将 N个所述单结点原型验证系统中的每一个单结点原型验证系统分别与 M个交叉开关互联路由器芯片组中的每一个交叉开关互联路由器芯片组相 连, 各交叉开关互联路由器芯片组之间不做转接, 任意一个交叉开关互联路 由器芯片组实现与其连接的 N个所述单结点原型验证系统内部互联, 以构成 一个 N结点 4*N路系统; M, N均为大于等于 2的正整数。
7、 如权利要求 6所述的方法, 其中,
所述 4路紧耦合计算板包括 4个 CPU, 所述 4个 CPU内部互联,彼此共 享存储器;
所述 N个单结点原型验证系统中的各 CPU之间经所述结点控制器芯片组 和交叉开关互联路由器芯片组彼此互联, 共享存储器。
8、 如权利要求 6所述的方法, 其中, 所述结点控制器的逻辑包括: 高速緩冲存储器(Cache )—致性控制和互 联网络接口控制。
9、 如权利要求 6所述的方法, 其中,
所述芯片验证板具有网络接口 (NI ) ;
在将所述 N个所述单结点原型验证系统互联时,是将所述 N个单结点原 型验证系统经各自的芯片验证板上的 NI接口与所述交叉开关互联路由器芯 片组相连。
10、 如权利要求 6-9中任何一项所述的方法, 其中,
N的取值为 8;
M的取值为 4。
11、 一种高端容错计算机系统, 包括:
N个单结点和 M个交叉开关互联路由器芯片组(NR ) , 每个所述交叉 开关互联路由器芯片组均用于实现所述 N个所述单结点内部互联, 各所述交 叉开关互联路由器芯片组之间不做转接, M, N均为大于等于 2的正整数, 其中: 所述单结点包括:
计算板, 为一 4路紧耦合计算板;
结点控制器, 实现对计算板上的 2路 CPU的控制。
12、 如权利要求 11所述的系统, 其中, 所述 4路紧耦合计算板包括 4 个 CPU, 所述 4个 CPU内部互联, 彼此共享存储器。
13、 如权利要求 11所述的系统, 其中, 所述 N个单结点中的各 CPU 之间经所述结点控制器和交叉开关互联路由器芯片组彼此互联,共享存储器。
14、 如权利要求 11所述的系统, 其中, N的取值为 8; M的取值为 4。
15、 一种如权利要求 11-14任一所述的高端容错计算机系统的实现方 法, 包括:
选择计算板, 所述计算板为一 4路紧耦合计算板;
选择结点控制器, 实现对计算板上的 2路 CPU的控制, 组成一个单结点 互联;
使用互联路由器芯片组将多个所述单结点互联, 其中, N个单结点和 M 个交叉开关互联路由器芯片组(NR ) , 每个所述交叉开关互联路由器芯片组 均用于实现所述 N个所述单结点内部互联, 各所述交叉开关互联路由器芯片 组之间不做转接, M, N均为大于等于 2的正整数。
16、 如权利要求 15所述的方法, 其中, N的取值为 8; M的取值为 4。
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US14/002,827 US8769459B2 (en) | 2011-03-07 | 2012-03-06 | High-end fault-tolerant computer system and method for same |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201110053727.2 | 2011-03-07 | ||
CN2011100537272A CN102129418B (zh) | 2011-03-07 | 2011-03-07 | 一种高端容错计算机系统及实现方法 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2012119533A1 true WO2012119533A1 (zh) | 2012-09-13 |
Family
ID=44267504
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2012/071956 WO2012119533A1 (zh) | 2011-03-07 | 2012-03-06 | 一种高端容错计算机系统及实现方法 |
Country Status (3)
Country | Link |
---|---|
US (1) | US8769459B2 (zh) |
CN (1) | CN102129418B (zh) |
WO (1) | WO2012119533A1 (zh) |
Families Citing this family (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102122259A (zh) | 2011-03-03 | 2011-07-13 | 浪潮(北京)电子信息产业有限公司 | 一种高端容错计算机原型验证系统及验证方法 |
CN102129418B (zh) | 2011-03-07 | 2013-04-17 | 浪潮(北京)电子信息产业有限公司 | 一种高端容错计算机系统及实现方法 |
CN103092807B (zh) * | 2012-12-24 | 2015-09-09 | 杭州华为数字技术有限公司 | 节点控制器、并行计算服务器系统以及路由方法 |
CN103902501A (zh) * | 2014-04-02 | 2014-07-02 | 浙江大学 | 基于fpga开发板的板间光互连网络拓扑结构性能的检测方法 |
CN104750581A (zh) * | 2015-04-01 | 2015-07-01 | 浪潮电子信息产业股份有限公司 | 一种冗余互连的内存共享的服务器系统 |
CN104965797A (zh) * | 2015-05-22 | 2015-10-07 | 浪潮电子信息产业股份有限公司 | 一种高端容错计算机目录体系结构实现方法 |
CN105808499A (zh) * | 2016-04-01 | 2016-07-27 | 浪潮电子信息产业股份有限公司 | 一种cpu互联装置以及多路服务器cpu互联拓扑结构 |
CN105933286B (zh) * | 2016-04-05 | 2019-08-02 | 浪潮电子信息产业股份有限公司 | 一种验证协议的方法及装置 |
CN107766599A (zh) * | 2016-08-22 | 2018-03-06 | 深圳市中兴微电子技术有限公司 | 集成电路芯片的原型验证装置 |
CN106776186B (zh) * | 2016-12-29 | 2020-04-07 | 湖南国科微电子股份有限公司 | 一种多cpu架构下的cpu运行状态调试方法和系统 |
CN107182069B (zh) * | 2017-05-23 | 2020-06-26 | 深圳市微网力合信息技术有限公司 | 无线路由器传输速率切换方法、控制方法及无线路由器 |
CN108509371A (zh) * | 2018-04-09 | 2018-09-07 | 郑州云海信息技术有限公司 | 一种高端容错计算机节点互联系统及实现方法 |
CN109190276A (zh) * | 2018-09-14 | 2019-01-11 | 天津市滨海新区信息技术创新中心 | Fpga原型验证系统 |
CN114860519B (zh) * | 2022-04-08 | 2022-12-23 | 中国人民解放军国防科技大学 | 一种面向大规模asic芯片的多芯片联合验证方法及装置 |
CN117610472B (zh) * | 2024-01-24 | 2024-03-29 | 上海合见工业软件集团有限公司 | 超大规模集群fpga原型验证系统 |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1742259A (zh) * | 2002-09-30 | 2006-03-01 | 英特尔公司 | 设备表示装置和方法 |
CN101216815A (zh) * | 2008-01-07 | 2008-07-09 | 浪潮电子信息产业股份有限公司 | 一种双翼可扩展多处理器紧耦合共享存储器体系结构 |
CN101330413A (zh) * | 2007-06-22 | 2008-12-24 | 上海红神信息技术有限公司 | 基于环绕网络与超立方网络架构的混合多阶张量扩展方法 |
CN101354694A (zh) * | 2007-07-26 | 2009-01-28 | 上海红神信息技术有限公司 | 基于mpu架构的超高扩展超级计算系统 |
CN102122259A (zh) * | 2011-03-03 | 2011-07-13 | 浪潮(北京)电子信息产业有限公司 | 一种高端容错计算机原型验证系统及验证方法 |
CN102129418A (zh) * | 2011-03-07 | 2011-07-20 | 浪潮(北京)电子信息产业有限公司 | 一种高端容错计算机系统及实现方法 |
CN102142050A (zh) * | 2011-03-01 | 2011-08-03 | 浪潮(北京)电子信息产业有限公司 | 一种高端容错计算机单结点原型验证系统及验证方法 |
Family Cites Families (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US4968977A (en) * | 1989-02-03 | 1990-11-06 | Digital Equipment Corporation | Modular crossbar interconnection metwork for data transactions between system units in a multi-processor system |
US7191380B2 (en) * | 2003-09-10 | 2007-03-13 | Hewlett-Packard Development Company, L.P. | Defect-tolerant and fault-tolerant circuit interconnections |
US7185138B1 (en) * | 2004-05-14 | 2007-02-27 | Peter Galicki | Multi-dimensional data routing fabric |
US7310004B2 (en) * | 2004-05-28 | 2007-12-18 | California Institute Of Technology | Apparatus and method of interconnecting nanoscale programmable logic array clusters |
US7495942B2 (en) * | 2004-08-13 | 2009-02-24 | University Of Florida Research Foundation, Inc. | Nanoscale content-addressable memory |
CN101833491B (zh) * | 2010-04-26 | 2012-10-24 | 浪潮电子信息产业股份有限公司 | 一种节点互连系统链路检测电路的设计与fpga实现方法 |
US8510239B2 (en) * | 2010-10-29 | 2013-08-13 | International Business Machines Corporation | Compact cognitive synaptic computing circuits with crossbar arrays spatially in a staggered pattern |
-
2011
- 2011-03-07 CN CN2011100537272A patent/CN102129418B/zh active Active
-
2012
- 2012-03-06 US US14/002,827 patent/US8769459B2/en active Active
- 2012-03-06 WO PCT/CN2012/071956 patent/WO2012119533A1/zh active Application Filing
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1742259A (zh) * | 2002-09-30 | 2006-03-01 | 英特尔公司 | 设备表示装置和方法 |
CN101330413A (zh) * | 2007-06-22 | 2008-12-24 | 上海红神信息技术有限公司 | 基于环绕网络与超立方网络架构的混合多阶张量扩展方法 |
CN101354694A (zh) * | 2007-07-26 | 2009-01-28 | 上海红神信息技术有限公司 | 基于mpu架构的超高扩展超级计算系统 |
CN101216815A (zh) * | 2008-01-07 | 2008-07-09 | 浪潮电子信息产业股份有限公司 | 一种双翼可扩展多处理器紧耦合共享存储器体系结构 |
CN102142050A (zh) * | 2011-03-01 | 2011-08-03 | 浪潮(北京)电子信息产业有限公司 | 一种高端容错计算机单结点原型验证系统及验证方法 |
CN102122259A (zh) * | 2011-03-03 | 2011-07-13 | 浪潮(北京)电子信息产业有限公司 | 一种高端容错计算机原型验证系统及验证方法 |
CN102129418A (zh) * | 2011-03-07 | 2011-07-20 | 浪潮(北京)电子信息产业有限公司 | 一种高端容错计算机系统及实现方法 |
Also Published As
Publication number | Publication date |
---|---|
CN102129418A (zh) | 2011-07-20 |
US20130346934A1 (en) | 2013-12-26 |
US8769459B2 (en) | 2014-07-01 |
CN102129418B (zh) | 2013-04-17 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2012119533A1 (zh) | 一种高端容错计算机系统及实现方法 | |
CN101710314B (zh) | 高速外围部件互连交换控制器及其实现方法 | |
WO2012116654A1 (zh) | 一种高端容错计算机原型验证系统及验证方法 | |
US9674114B2 (en) | Modular decoupled crossbar for on-chip router | |
Zhang et al. | A reconfigurable routing algorithm for a fault-tolerant 2D-mesh network-on-chip | |
Tsai et al. | A fault-tolerant NoC scheme using bidirectional channel | |
CN104579951A (zh) | 片上网络中新颖的故障与拥塞模型下的容错方法 | |
WO2022179105A1 (zh) | 一种多路服务器及多路服务器信号互联系统 | |
Rezaei et al. | Fault-tolerant 3-D network-on-chip design using dynamic link sharing | |
Bistouni et al. | Rearranging links: a cost-effective approach to improve the reliability of multistage interconnection networks | |
Montano et al. | A low-latency reconfigurable multistage interconnection network | |
Yin et al. | Comparison of mesh and honeycomb network-on-chip architectures | |
Wang et al. | Design and implementation of fault-tolerant and cost effective crossbar switches for multiprocessor systems | |
Zhou et al. | A fault-tolerant deflection routing for network-on-chip | |
Rantala et al. | Fault-tolerant routing approach for reconfigurable Networks-on-Chip | |
KR102691170B1 (ko) | 실리콘 포토닉스 기반 인터커넥션 네트워크에서 (SiPh-HCA) 컴퓨팅 호스트 간 고성능 컴퓨팅을 실현하는 유연한 토폴로지 구성 (FlexibleX) 기법과 패킷 제어 기술 | |
Chen et al. | Router-shared-pair mesh: a reconfigurable fault-tolerant network-on-chip architecture | |
Sudhina et al. | FAULT-TOLERANT ROUTING METHODS IN NETWORK ON CHIP. | |
Gu et al. | Research on network fault tolerance method on chip | |
Kamal et al. | Network on chip: topologies, routing, implementation | |
Agyeman | A low overhead fault reporting scheme for resilient 3D network-on-chip applications | |
Kumar et al. | Design of NoC-Based High-Speed Processor to Suit IoT for a Hyper-Connected Smart World | |
Zakaria | A Comparative Review of Adaptive Routing Approach for Network-on-Chip Router Architecture | |
Malladhi et al. | Design and Development of Novel AXI Interconnect based NoC Architecture for SoC with Reduced Latency and Improved Throughput | |
Fu et al. | Low-cost router microarchitecture based on dimension-switch for fault-tolerance in 2D-mesh NoC |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 12755121 Country of ref document: EP Kind code of ref document: A1 |
|
WWE | Wipo information: entry into national phase |
Ref document number: 14002827 Country of ref document: US |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 12755121 Country of ref document: EP Kind code of ref document: A1 |