SPRACC0A November 2017 – November 2020 TMS320F28075 , TMS320F28075-Q1 , TMS320F28076 , TMS320F28374D , TMS320F28374S , TMS320F28375D , TMS320F28375S , TMS320F28375S-Q1 , TMS320F28376D , TMS320F28376S , TMS320F28377D , TMS320F28377D-EP , TMS320F28377D-Q1 , TMS320F28377S , TMS320F28377S-Q1 , TMS320F28378D , TMS320F28378S , TMS320F28379D , TMS320F28379D-Q1
Error detection and correction (often referred to as ECC) goes beyond parity in that it corrects single bit errors. The EDAC circuitry can also detect a 2-bit uncorrectable error. It is possible to implement 2-bit correction, but per the discussions in Section 2 and Section 4.1, this does not provide a good return for the added cost and timing overhead.
EDAC addresses the system availability perspective for safety since the system will continue to run unabated in the presences of a single bit error.
However, EDAC:
For example consider the following:
Not all the SRAM necessarily requires EDAC protection so this can vary with device designs. Likewise there are ways to reduce access time to less than a wait state.