Abstract:With the ever-increasing integration and parallel computing capabilities of 3D processor arrays, the occurrence of processor elements (PEs) failures caused by various factors has become more prevalent. Therefore, the implementation of a fault-tolerant mechanism that uses the remaining fault-free PEs to reconfigure sub-array becomes critical. In this paper, we study the problem of reconfiguring a 3D subarray with as many fault-free PEs as possible, which has been shown to be NP-complete in previous work. Although prior algorithms have been effective under low fault densities, they are severely limited when faced with high fault densities. To address this, we first define the bottleneck of the 3D processor array, proposed a novel method to identify the physical bottleneck plane that restricts the reconfigurable size of the logical sub-array and prove its correctness. Then, we propose an effective compensation strategy that can fully utilize the fault-free PEs in the bottleneck plane. Under this strategy, a sliding-window weight calculation method is proposed to determine the priority of compensation. Finally, we proposed a heuristic algorithm, which can construct the maximum target array from different dimensions in polynomial time. Experimental results demonstrate that the proposed algorithm exhibits favorable performance in terms of harvest and degradation. For the random-failure model, the improvement in the harvest for fault-free PEs is up to 32.03% on a 32x32x3232x32x32 host array with a 20% fault density. And for the clustered fault model, the improvement in harvest is up to 70.63% on a 32x32x3232x32x32 host array distributed with 12 cluster failures of size 6x6x66x6x6.

Neural Network Algorithm of VLSI Fault-Tolerant Array Using Single-Track Switches

Reconfiguration Algorithm for Redundancy VLSI/WSI Arrays Using Single-Track Switches Based on Neural Networks

Reconfiguration Algorithm for Degradable VLSI/WSI Arrays Based on Neural Networks

Interconnection Network Reconstruction for Fault-Tolerance of Torus-Connected VLSI Array.

Satisfiability-based Method for Reconfiguring Power Efficient VLSI Array

A High-Performance VLSI Array Reconfiguration Scheme Based on Network Flow under Row and Column Rerouting

Optimal Reconfiguration of High-Performance VLSI Subarrays with Network Flow

An Efficient Method for Reconfiguring Power‐efficient VLSI Array with Maximum Satisfiability

Efficient Reconfiguration Algorithm for Three-dimensional VLSI Arrays.

An efficient multiple shortest augmenting paths algorithm for constructing high performance VLSI subarray

A Mathematical Model for Reconfiguring VLSI Subarrays under Row and Column Rerouting

An Efficient Bottleneck Planes Exclusion Method for Reconfiguring 3D VLSI Arrays

Efficient Reconfiguration Algorithm with Flexible Rerouting Schemes for Constructing 3-D VLSI Subarrays

Non-Backtracking Reconfiguration Algorithm for Three-dimensional VLSI Arrays

A Neural Network Method For Multiple Fault Detection And Diagnosis Of Digital Circuits

Efficient Abstraction Algorithms for Accelerating Reconfiguration of Vlsi Arrays

Divide and conquer algorithm for parallel reconfiguration of VLSI array with faults

A Mathematical Programming Method for Constructing the Shortest Interconnection VLSI Arrays.

Neural Network Approach for Multiple Fault Test of Digital Circuit

Flexible Scheme for Reconfiguring 2D Mesh-Connected VLSI Subarrays under Row and Column Rerouting.

A Deterministic-Path Routing Algorithm for Tolerating Many Faults on Very-Large-Scale Network-on-Chip