指令预取

  • 网络Prefetch;Instruction prefetch;Instruction Prefetching
指令预取指令预取
  1. 基于控制流的混合指令预取

    Hybrid Instruction Prefetch Based on Control Flow

  2. 指令预取技术能够有效地降低指令Cache的访问失效率,提高微处理器的取指令能力,进而提高微处理器的性能。

    Instruction prefetching is an effective mechanism to reduce the instruction cache miss rate .

  3. 由于现代微处理器使用了基于Cache存储和指令预取技术,增加了准确确定这一上界的难度,为此提出了一种基于指令Cache和指令预取联合模型的嵌入式软件性能评估新方法。

    Modern microprocessors which use Cache memory system and instruction pre-fetching increase the difficulty to compute the upper bound accurately .

  4. 该算法的实现代价合理,具有良好的可扩展性。(3)扩展了片上trace技术的应用领域,提出一种将代码排布技术与指令预取技术结合使用的方法。

    The algorithm has good scalability with reasonable hardware costs . ( 3 ) Expanding the application of on-chip trace debug , a method combining code layout and instruction prefetching is proposed .

  5. 以基本块为单位的非顺序指令预取

    The Non - Sequential Instruction Prefetching Based on Basic Blocks

  6. 复位置位同步触发器同步数据触发体系结构中指令预取技术研究

    Research on the Instruction Prefetching Technology for the Synchronous Data Triggered Architecture

  7. 实验结果表明,值预测技术、指令预取技术和数据预取技术混和使用可以获得较大的性能提升。

    Experimental results show that RSVP has better cost-performance than other data value prediction schemes .

  8. 逆指令流与指令队列预取反跟踪技术

    Anti tracking technology of the converse instructions and the prefetch instruction queue

  9. 本文分别研究了指令访问和数据访问技术,提出了基于控制流的混合指令预取方法,并探讨了在动态VLIW结构中实现Load前瞻所需的体系结构和编译支持。

    This paper puts forward hybrid instruction prefetching based on control flow and discusses how to perform load speculation in dynamic VLIW architectures .

  10. 阐述了一种高性能低功耗MCU的设计。该MCU和标准8051具有相同的指令系统和功能,通过体系结构上采用哈佛结构、1时钟机器周期和指令预取,提高MCU的工作效率;

    A high-performance and low-power 8-bit microcontroller is described , which shares the same instruction set with the standard Intel 8051 . In system architecture , one-clock period per machine cycle architecture and pre-fetching instruction method are adopted to improve MCU 's power efficiency .

  11. 利用SSE指令集,对原有60Co集装箱CT系统图像重建算法进行优化,包括程序的并行化处理及指令预取优化。

    SSE is used in evolution of cobalt-60 container CT image reconstruction , including the parallel computing of program and the prefetching of instruction .