The search functionality is under construction.

Author Search Result

[Author] Fumiyasu HIROSE(2hit)

1-2hit
  • Integrated Design and Test Assistance for Pipeline Controllers

    Hiroaki IWASHITA  Tsuneo NAKATA  Fumiyasu HIROSE  

     
    PAPER

      Vol:
    E76-D No:7
      Page(s):
    747-754

    We Propose an integrated design and test assistance method for pipelined processors. Our approach generates behavioral-level test environments for pipeline control mechanisms from a machine-readable specification. It includes automatic generation of test programs and behavioral descriptions. Verification can be done by applying logic simulation to both the designers' descriptions and the behavioral descriptions, and then comparing the results. We have implemented an experimental system that enumerates all hazard patterns--instruction patterns that cause pipeline hazards--from the specifications, and generates the test programs and the behavioral descriptions for the pipeline controllers. The test programs cover all of the hazard patterns. The behavioral descriptions can manipulate any instruction stream. Experimental results for several RISC processors show that actual hazard patterns are too numerous to be easily enumerated by hand. Using workstations, our system can generate the test programs that cover all of the patterns, taking a few minutes. Results suggest that the system can be used to evaluate pipeline design.

  • Mincut Partitioning Acceleration Using Hardware CAD Accelerator TP5000

    Masahiro SANO  Shintaro SHIMOGORI  Fumiyasu HIROSE  

     
    PAPER

      Vol:
    E78-A No:12
      Page(s):
    1785-1792

    This paper presents a new approach of data pipelining for mincut partitioning acceleration using a parallel computer. When using a parallel computer, it is important to have many processors always active, also the quality of the partitioning must not be sacrificed. Out approach covers both speed and quality. We choose the hardware CAD accelerator TP5000 to implement our approach, which consists of dedicated Very Long Instruction Word (VLIW) processors with high-speed interconnections. The TP5000 allows its connections to be reconfigured to optimize the data pipelines. We estimate that the speed of our approach using 10 processors on the TP5000 is 30 times faster than a SPARCStation-10.