Evaluating Direct Compare for Double Error-Correction Codes

Direct compare of information coded with error-correction codes has been proposed to efficiently protect data comparisons against soft errors. So far, it has been evaluated only for single error correction-double error detection codes. As technology scales, multiple bit errors are more frequent and, thus, more advanced protection is needed. This letter evaluates direct compare when using double error-correction codes. The results show that it can significantly reduce the circuit overheads and that the savings are larger than when used for single error correction.