A fault-tolerant PE array based matrix multiplier design

Jan, B.-Y.; Huang, Jiun-Lang

doi:10.1109/vlsi-dat.2012.6212624

Cited by 1 publication

(28 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this paper, we propose a highly fault-tolerant approach for a systolic array executing Cannon's algorithm for matrix multiplication. We observe that the main limitation of the previous work [30] lies in the necessity to find a fault-free column (even in the form of a twisted column). Consequently, the previous work is only suitable for situations where the number of faulty PEs is relatively low.…”

Section: Introductionmentioning

confidence: 82%

“…Jan and Huang [30] discuss the fault-tolerant design of a systolic array executing Cannon's algorithm [1] for matrix multiplications. Their basic idea involves assigning the task of a faulty PE to a fault-free PE for execution.…”

Section: Introductionmentioning

confidence: 99%

“…Therefore, their approach does not require redundant PEs. In [30], they propose a mapping algorithm to implement both a twisted column scheme and a column replacement scheme. If the number of faulty PEs is small, their method can achieve very high fault tolerance.…”

Section: Introductionmentioning

confidence: 99%

“…However, the approach proposed by Jan and Huang [30] cannot be applied when there is a large number of faulty PEs. The main reason is that their method requires the existence of at least one fault-free column (even in the form of a twisted column).…”

Section: Introductionmentioning

confidence: 99%

“…Consequently, the previous work is only suitable for situations where the number of faulty PEs is relatively low. In contrast to the previous approach [30], our method does not require the existence of a fault-free column, thus enabling its application in scenarios with a higher number of faulty PEs.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Highly Fault-Tolerant Systolic-Array-Based Matrix Multiplication

Lu,

Su,

Huang

2024

Electronics

View full text Add to dashboard Cite

Matrix multiplication plays a crucial role in various engineering and scientific applications. Cannon’s algorithm, executed within two-dimensional systolic arrays, significantly enhances computational efficiency through parallel processing. However, as the matrix size increases, reliability issues become more prominent. Although the previous work has proposed a fault-tolerant mechanism, it is only suitable for scenarios with a limited number of faulty processing elements (PEs). This paper introduces a pair-matching mechanism, assigning a fault-free PE as a proxy for each faulty PE to execute its tasks. Our fault-tolerant mechanism comprises two stages: in the first stage, each fault-free PE completes its designated computations; in the second stage, computations intended for each faulty PE are executed by its assigned fault-free PE proxy. The experimental results demonstrate that compared to the previous work, our approach not only significantly improves the fault tolerance of systolic arrays (applicable to scenarios with a higher number of faulty PEs) but also reduces circuit areas. Therefore, the proposed approach proves effective in practical applications.

show abstract

Section: Introductionmentioning

confidence: 82%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations