超大规模FPGA的单粒子效应脉冲激光测试方法

建立了一种28 nm HPL硅工艺超大规模SRAM型FPGA的单粒子效应测试方法。采用静态测试与动态测试相结合的方式,通过ps级脉冲激光模拟辐照实验,对超大规模FPGA进行单粒子效应测试。对实验所用FPGA的各敏感单元(包括块随机读取存储器、可配置逻辑单元、可配置存储器)的单粒子闩锁效应和单粒子翻转极性进行了研究。实验结果证明了测试方法的有效性,揭示了多种单粒子闩锁效应的电流变化模式,得出了各单元的单粒子效应敏感性区别。针对块随机读取存储器、可配置逻辑单元中单粒子效应翻转极性的差异问题,从电路结构方面进行了机理分析。     

基于FPGA技术的存储器设计及其应用

本文介绍了基于FPGA技术的存储器构造方法,重点介绍利用库函数构造双端口RAM的方法。进而结合高速数据采集的特点,详细介绍了双端口RAM,以及配合双端口RAM工作需要而涉及到的地址发生器、读写控制逻辑等外围电路在FPGA中的实现。     

FPGA双端口存储器映射优化算法

FPGA存储器映射算法负责将用户的逻辑存储需求映射到芯片中的分布式存储资源上实现。前人对双端口存储器的映射算法研究相对较少,成熟的商业EDA工具的映射结果仍有不少改进空间。该文分别针对面积、延时、功耗这3个常用指标,提出一种双端口存储器映射的优化算法,并给出了具体配置方案。实验表明,在面向简单存储需求时,与商用工具Vivado的映射结果一致;在面向复杂存储需求时,面积优化和功耗优化的映射结果对比商用工具改善了至少50%。     

反熔丝FPGA配置电路的研究

反熔丝FPGA的编程过程是对可编程逻辑模块进行配置的过程,具体过程是利用配置电路对连接在逻辑模块输入、输出端口的反熔丝单元施加高压,将其击穿形成连接通路,从而使得逻辑模块连接在信号线上.位流数据控制配置电路完成反熔丝单元的预充电、寻址、加压以及编程过程.通过控制晶体管的打开和关断及复用编程通道,可以实现不同类型反熔丝的编程,并简化配置电路的设计规模.测试结果表明,配置电路可靠地完成了反熔丝的编程功能.     

一种用于FPGA的可配置存储器设计

设计了一种用于FPGA中的同步、双端口、容量为4kbit、可配置的存储器模块(Block RAM,BRAM)。BRAM以阵列形式内嵌在FPGA内部,是FPGA的主要模块之一。该BRAM可实现1、2、4、8、16bit 5种不同的数据位宽,且具有数据初始化及回读验证的功能。本文分别对BRAM的逻辑层、配置层、布线层进行了描述,重点介绍了逻辑层中时序控制电路和配置层中配置电路的结构和实现方法。基于0.18μm 5层金属SOI CMOS工艺完成BRAM设计实现,并对BRAM进行了仿真,功能仿真结果符合时序控制电路和配置电路的预期设计目标,性能仿真表明其工作频率可达200 MHz。     

一种专用可重配置的FPGA嵌入式存储器模块的设计和实现

本文设计了一种满足FPGA芯片专用定制需求的嵌入式可重配置存储器模块.一共8块,每块容量为18Kbits的同步双口BRAM,可以配置成16K×1bit、8K×2bits、4K×4bits、2K×9bits、1K×18bits、512×36bits六种不同的位宽工作模式;write_first、no_change两种不同的写入模式.多个BRAM还可以通过FPGA中互连电路的级联来实现深度或宽度的扩展.本文重点介绍实现可重配置功能的电路及BRAM嵌入至FPGA中的互连电路.采用SMIC 0.13μm 8层金属CMOS工艺,产生FDP-II芯片的完整版图并成功流片,芯片面积约为4.5mm×4.4mm.运用基于March C+算法的MBIST测试方法,软硬件协同测试,结果表明FDP-II中的BRAM无任何故障,可重配置功能正确,证实了该存储器模块的设计思想.     

现场可编程门阵列

近年来,现场可编程门阵列FPGA(Field Programble GateArray)很风行,它属于编程逻辑器件。编程逻辑器件具有将软件编程和硬件制作相结合的特点。 FPGA是高集成度的半导体集成电路,其容量最多可达二万多门。它本身没有任何逻辑功能,通过编程,使之具有用户所要求的逻辑功能和外管脚分配,这些编程数据不是存储在FPGA芯片中的.而是存储在与FPGA配套使用的外部存储器中,如EPROM。     

采用AVR单片机对FPGA进行配置

使用AVR单片机对Altera公司的FPGA进行配置,讨论了配置的时序要求、配置文件的处理,给出了硬件设计和软件设计 Altera公司的ACEX、FLEX等系列的FPGA芯片应用广泛,但其FPGA基于SRAM结构,决定电路逻辑功能的编程数据存储于SRAM中.由于SRAM的易失性,每次上电时必须重新把编程数据装载到SRAM中,这一过程就是FPGA的配置过程.     

基于双FPGA的刀闸接口控制箱的设计

现有变电站改造成数字化变电站时需要增加过程层设备,其中对刀闸接口控制箱的动作可靠性提出了极高的要求。提出一种基于双FPGA实现多重逻辑闭锁的刀闸接口控制箱实现方案。设计了FPGA电源和时钟实现电路,两块FPGA的信息交换方式以及逻辑互锁方法。详细描述了出口电路自检方案。介绍FPGA配置和编程方法,给出了FPGA的时序仿真。测试了电源建立波形,实现并验证了装置运行的可靠性。该方案能够有效防止误动作发生,适用于有此需求的一般装置。     

一种抗辐射加固FPGA 编程电路的设计与实现

本文介绍了抗辐射加固SOI-SRAM基FPGA编程电路的设计与实现。该电路完成FPGA配置数据的下载与回读。该编程电路采用编程点直接寻址的方式,相对典型的移位寄存器链寻址方式不仅能够节约面积开销而且可以提供更为灵活的配置选择。通过对本电路提出的部分配置控制寄存器的配置,该编程电路可以实现的最小配置单元仅包含1位数据,FPGA更为灵活的部分重配置功能得以方便实现。层次化的仿真策略,对关键路径的优化及精密的版图布局保证了该电路的性能。此外对编程点进行了抗辐射加固设计。该电路在基于0.5μm部分耗尽SOI工艺SRAM基的FPGA中实现。功能测试结果表明, 该编程电路成功实现FPGA配置数据的下载与回读,且抗辐照实验结果表明,抗总剂量水平超过1x105Krad(Si), 抗瞬态剂量率水平超过1.5x1011 rad(Si)/s,抗中子注入量水平达到1x1014 n/cm2。     

利用FPGA和DSP结合实现雷达多目标实时检测

在高速并行流水信号处理中,ASIC(FPGA)+DSP+RAM是目前国际流行的一种方式,尤其是FPGA+DSP+RAM更适合中国的国情.本文利用FPGA的算术逻辑单元与外部存储器相结合,解决了线路板面积有限与雷达数据处理需要大量存储空间的矛盾;利用FPGA的并行流水特点解决了雷达数据的实时处理与有限的DSP处理速度之间的矛盾; 而FPGA处理结果的航迹相关、FPGA运行模式的控制和与上位机之间的通信与数据交换等工作利用DSP完成,从而达到系统的最佳配置.目前系统已通过验收,各项指标达到了设计的要求.     

基于FPGA的开关盒互连容错算法的研究

FPGA主要由两个基本部分组成,一是可配置逻辑部件,另一部分就是互联网络,负责对可配置逻辑块间的通信。FPGA内部大约80%的晶体管都是作为可编程开关和缓冲器来完成可编程路由网络工作的。文中主要对出现错误的开关盒阵列中可执行的路径数量进行评估,并且使用算法找到合适的路径,避开错误。     

Novel Test Approach for Interconnect Resources in Field Programmable Gate Arrays

A novel test approach for interconnect resources(IRs)in field programmable gate arrays (FPGA)has been proposed.In the test approach,SBs (switch boxes)of IRs in FPGA has been utilized to test IRs.Furthermore,configurable logic blocks(CLBs)in FPGA have also been employed to enhance driving capability and the position of fault IR can be determined by monitoring the IRs associated SBs.As a result,IRs can be scanned maximally with minimum configuration patterns.In the experiment,an in-house developed FPGA test system based on system-on-chip(SoC)hardware/software verification technology has been applied to test XC4000E family of Xilinx.The experiment results revealed that the IRs in FPGA can be tested by 6 test patterns.     

基于测试系统的FPGA逻辑资源的测试

FPGA在许多领域已经得到广泛应用,其测试问题也显得越来越突出。文章针对基于SRAM结构FPGA的特点,以Xilinx公司的XC4000系列芯片为例,利用检测可编程逻辑资源的多逻辑单元(CLB)混合故障的测试方法,阐述了如何在BC3192V50测试系统上实现FPGA的在线配置以及功能和参数测试。它是一种基于测试系统的通用的FPGA配置和测试方法。     

The flexibility of configurable computing

There has been growing recent interest in configurable computing, which can be viewed as a hybrid between ASICs and programmable processors. Configurable computing machines are implemented with programmable logic: flexible hardware that can be structured to fit the natural organization and data flow of a computation. The enabling device for configurable computing is the field-programmable array (FPGA). For applications characterized by deeply pipelined, highly parallel, and integer arithmetic processing, configurable computing machines can outperform alternative solutions by up to an order of magnitude. The combination in a single device of dedicated hardware and rapid, submillisecond-scale reprogrammability constitutes an exciting and promising development whose implications are only just beginning to be exploited. We begin with a brief tutorial on FPGAs that describes the most common FPGA architectures and how these architectures are used to support computation, memory access, and data flow. We then present FPGAs as computing machines and focus on devices that are reconfigured during run time. Ongoing research involving FPGAs and future directions are also discussed     

A survey of FPGA design for AI era

FPGA is an appealing platform to accelerate DNN.We survey a range of FPGA chip designs for AI.For DSP module,one type of design is to support low-precision operation,such as 9-bit or 4-bit multiplication.The other type of design of DSP is to support floating point multiply-accumulates(MACs),which guarantee high-accuracy of DNN.For ALM(adaptive logic module)module,one type of design is to support low-precision MACs,three modifications of ALM includes extra carry chain,or 4-bit adder,or shadow multipliers which increase the density of on-chip MAC operation.The other enhancement of ALM or CLB(configurable logic block)is to support BNN(binarized neural network)which is ultra-reduced precision version of DNN.For memory modules which can store weights and activations of DNN,three types of memory are proposed which are embedded memory,in-package HBM(high bandwidth memory)and off-chip memory interfaces,such as DDR4/5.Other designs are new architecture and specialized AI engine.Xilinx ACAP in 7 nm is the first industry adaptive compute acceleration platform.Its AI engine can provide up to 8X silicon compute density.Intel AgileX in 10 nm works coherently with Intel own CPU,which increase computation performance,reduced overhead and latency.     

大数模乘脉动阵列的FPGA细粒度映射实现

通过分析FPGA可配置逻辑块的细致结构,提出了一种基于FPGA的细粒度映射方法,并使用该方法高效实现了大数模乘脉动阵列.在保持高速计算特点的同时,将模乘脉动阵列的资源消耗降低为原来的三分之一.在低成本的20万门级FPGA器件中即可实现1024位模乘器.该实现每秒可进行20次RSA签名.如果换用高性能FPGA,签名速度更可提高至每秒40次.     

A novel FPGA architecture supporting wide, shallow memories

This paper investigates an architecture designed to implement wide, shallow memories on a field programmable gate array (FPGA). In the proposed architecture, existing configuration memory normally used to control the connectivity pattern of the FPGA is made user accessible. Typically, not all the switch blocks in an FPGA are used to transport signals. By adding only a modest amount of circuitry, the configuration memory in these unused switch blocks (or unused paths within used switch blocks) can be used to implement wide, shallow buffers and other similar memory structures. The size of FPGA required to implement a benchmark circuit that makes use of the wide, shallow memories, is 20% smaller than a standard memory architecture. In addition, the benchmark circuit is on average 40% faster using the proposed architecture.     

一种基于JTAG的CLB内建自测试方法

针对Virtex-4型FPGA中可编程逻辑块故障检测的需求,提出了一种基于JTAG的内建自测试方法,并基于DEV++平台自行开发了基于并口的专用边界扫描测试软件.该方法可以比较可靠的检测FPGA中存在故障的可编程逻辑块,并能比较高的分辨率实现故障的定位.与传统的单故障检测方法相比,提出的改进型测试方法可以检测和定位多个故障CLB,并可以对故障类型进行诊断.实验结果表明:提出的测试方法可以精确的检测和定位存在故障的多个CLB,对具有类似结构的SRAM型FPGA具有普遍适用性.     

A software/reconfigurable hardware SAT solver

This paper introduces a novel approach for solving the Boolean satisfiability (SAT) problem by combining software and configurable hardware. The suggested technique avoids instance-specific hardware compilation and, as a result, allows the total problem solving time to be reduced compared to other approaches that have been proposed. Moreover, the technique permits problems that exceed the resources of the available reconfigurable hardware to be solved. The paper presents the results obtained with some of the DIMACS benchmarks and a comparison of our implementation with other available SAT solvers based on reconfigurable hardware. The hardware part of the satisfier was realized on Virtex XCV812E FPGA, which has a large volume of embedded memory blocks that provide direct support for the proposed approach.