sgemm_tcore: Fix address overlap for DMA
Enforce square shapes of tiles in smem. TODO need to configure loop bounds correctly.
This commit is contained in:
@@ -19,7 +19,7 @@
|
||||
// * Combining BM * BK >= (BM*BN) / (TM*TN) == threadblock yields
|
||||
// BM <= BK*TM*TN
|
||||
#define BM 64
|
||||
#define BN 64
|
||||
#define BN 32
|
||||
#define BK 64
|
||||
#define WM 16
|
||||
#define WN 8
|
||||
|
||||
Reference in New Issue
Block a user