flash: Fix DMA addr stride, stop at S=Q*K
This commit is contained in:
@@ -6,7 +6,7 @@
|
||||
#include "include/gemmini.h"
|
||||
#include "gemmini_mmio.h"
|
||||
|
||||
#define FP_SIZE 16
|
||||
#define FP_SIZE 32
|
||||
|
||||
// "fake" fp16 type that only has the correct data width.
|
||||
using float16_t = uint16_t;
|
||||
@@ -29,7 +29,7 @@ using float_type = float16_t;
|
||||
// (BM*BN) / (TM*TN) == threadblock size >= NT * CORES_PER_CLUSTER
|
||||
// * Combining BM * BK >= (BM*BN) / (TM*TN) == threadblock yields
|
||||
// BM <= BK*TM*TN
|
||||
#define BM 128
|
||||
#define BM 64
|
||||
#define BN 64
|
||||
#if (FP_SIZE == 32)
|
||||
#define BK 64
|
||||
@@ -62,18 +62,18 @@ static_assert(WMITER * WNITER * TCM * TCN * NUM_WARPS * CORES_PER_CLUSTER ==
|
||||
#define BK_LOOP 1
|
||||
// Whether to transpose smem A tile at GMEM->SMEM (produce), or SMEM->RF
|
||||
// (consume). This is because the tensor core expects the A tile to be stored
|
||||
// in column-major order in SMEM, whereas it will be ultimately stored in
|
||||
// row-major in the RF.
|
||||
// in column-major order in SMEM, so a transpose is necessary if A was stored
|
||||
// row-major in GMEM.
|
||||
//
|
||||
// For correctness, only one of either should be 1. E.g., PRODUCE 1 CONSUME 0
|
||||
// generates the NN kernel where both A and B are stored row-major in GMEM.
|
||||
// To model the case where the A matrix is already stored column-major in GMEM,
|
||||
// set both to 0.
|
||||
#define TRANSPOSE_AT_PRODUCE 0
|
||||
#define TRANSPOSE_AT_CONSUME 0
|
||||
#define TRANSPOSE_AT_CONSUME 1
|
||||
|
||||
#define GEMMINI_DMA 0
|
||||
#define GEMMINI_DMA_MN_MAJOR 1
|
||||
#define GEMMINI_DMA 1
|
||||
#define GEMMINI_DMA_MN_MAJOR 0
|
||||
#if SMEM_SIZE == 0x4000
|
||||
#define SMEM_ADDR_Q0 ((float * const) 0xff000000)
|
||||
#define SMEM_ADDR_Q1 ((float * const) 0xff001000)
|
||||
|
||||
Reference in New Issue
Block a user