Astha a42af89384 Addition of code for XCD remapping ...

This change adds in a function to remap block ids from their original round
robin assignment to a contiguous layout across XCDs. This function is added
to the StreamKTilePartitioner and called in the operator() functions. There
are also unit tests to verify the correctness of the function on minimal
arrays. These changes should improve locality and the cache hit rate, therefore
improving performance overall.

2026-01-26 04:47:01 -05:00

..

ck

Add new instances for merging multiple fwd conv groups into a single GEMM batch. Allow group merging for C > 1 when vector load/store size is 1 for the output tensor. (#3639)

2026-01-25 13:42:23 +01:00

ck_tile

Addition of code for XCD remapping

2026-01-26 04:47:01 -05:00

rapidjson

Update pre-commit to fixed versions, run remod for ck_tile (#2895)

2025-10-16 15:29:17 -07:00