| 摘要 |
In a memory system, a spare error correction bit is produced by processing data to be stored in sufficiently large chunks that the number of error correction bits required to protect each chunk are fewer than the available error correction signal lines on a memory bus and storage device. The spare bit is then used for an inversion bit in a parallel data bus inversion scheme, wherein data is selectively inverted to minimize bus switching. The transmission of data and error correction bits are spread over multiple phases, wherein parallel data bus inversion is applied to each phase. Alternatively, the transmission of data and error correction bits may be transmitted and stored in a single transaction. In either case, the spare bit is transmitted on a conventional memory bus and stored in a conventional memory module along with data and error correction bits. |
