Efficient reconfiguration scheme for mesh-connected network: the recursive shift approach

Numata, I.; Horiguchi, Susumu

doi:10.1109/ispan.1996.508985

Cited by 6 publications

(3 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Kung proposed an algorithm based on graph theory [6] and Numata proposed a heuristic algorithm (HS method) which gives highly system yield [3]. in each wafer and there are no shifting in inter-wafer.…”

Section: Shift Directionmentioning

confidence: 99%

“…There are many algorithms to determine the shift direction for mesh network. The HS algorithm [3] can obtain good system yields. This algorithm determines the shift direction randomly so that there are no biases for the shift direction.…”

Section: Biased Shiftingmentioning

confidence: 99%

“…The shift direction much effects on the system yield. Kung et.a1 [6] proposed a reconfiguration algorithm based on graph theory and Numata et.a1 [3] proposed a self-reconfiguration algorithm that uses only local information to determine the shift direction.…”

mentioning

confidence: 99%

See 2 more Smart Citations

An optimal replacement policy for cooling of 3D stacked mesh array

Inoguchi

Matsuzawa

Horiguchi

2000

Proceedings Fourth International Conference/Exhibition on High Performance Computing in the Asia-Pacific Region

Self Cite

View full text Add to dashboard Cite

This paper addresses reconjgurations and cooling of mesh array in 3 0 stacked integration by considering thermo-radiation.We propose heuristic replacement policy to avoid defects on a wafer by moving hot processing elements (PES) toward edge of a wafer to cool eflciently. Introducing a thermoradiation model in 3 0 stacked integration, reconfiguration performance arid thermo-radiation of 3 0 stacked mesh array are evaluated.Three cooling schemes are proposed and evaluated. Placement of spare PES, moving direction of hot PES and heuristic replacenlent are discussed from viewpoint of system yield and cooling performance. The heuristic replacement policy gathers a various replacement by repeating recotlfiguration procedure, and estimates cooling eficiency based on thermo-radiation model for each gathered replacement, then the policy adopts the best replacement.Comparing system yields and marimiini temperatures by thermo-radiation simulations on Cray-T3E super computer, these cooling approaches maintain the 3 0 stacked mesh array to be lower temperature keeping high system yield.

show abstract

Section: Shift Directionmentioning

confidence: 99%

Section: Biased Shiftingmentioning

confidence: 99%