首页> 外文会议>OCEANS, 2001 >Concurrent error diagnosis in mesh array architectures based onoverlapping H-processes
【24h】

Concurrent error diagnosis in mesh array architectures based onoverlapping H-processes

机译:基于重叠H过程的网格阵列架构中的并发错误诊断

获取原文
获取原文并翻译 | 示例

摘要

Unlike other methods for concurrent error detection and locationn(CED), the one proposed is not application specific and does not requirenfault free comparators and custom VLSI design for the processing cells.nIt is suitable for any algorithm that can be decomposed in blocknoperations of the format [(a op1 b) op2 (cnop3 d)], where a, b, c, d are arbitrary operands and op1n, op2, op3 dyadic operators. The process ofncomputing such an operation in a distributed and redundant way on annH-tree shaped sub-array is called an H-process. Many H-processes cannoverlap providing a general purpose mechanism for run-time faultntolerance in data driven mesh array architectures. Errors can bendetected during normal operation. Suspected erroneous results can benmasked while location is attempted. There is no need for retries.nDiagnosis is achieved `on the fly' without graceful degradation, uponndetection
机译:与其他用于并发错误检测和定位的方法(CED)不同,本文提出的方法不是特定于应用的,不需要处理单元的无故障比较器和自定义VLSI设计.n适用于可在格式的块运算中分解的任何算法。 [(a op 1 b)op 2 (cnop 3 d)],其中a,b,c,d是任意操作数,而op 1n ,op 2 ,op 3 二元运算符。在nH树形子数组上以分布式和冗余方式对这种操作进行n计算的过程称为H过程。许多H进程可以重叠,从而为数据驱动的网格阵列体系结构中的运行时容错提供了一种通用机制。在正常操作期间可能无法检测到错误。尝试定位时,可能会掩盖可疑的错误结果。无需重试。n在检测到的情况下,可以“即时”实现诊断,而不会导致性能下降

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号