ayo
22026ede12
Preset K, byte-identical to the previous output in all captured buffers and the harness image, measured on an RX 7700 XT: - LDS row padding so WMMA operand loads hit distinct banks - WGP mode for enc1/enc2/dec1/dec2, ds_bpermute half-wave exchange for enc0/dec0/dec3, more waves per window in enc4/dec4/dec5 - cheaper operand plumbing in common/wmma_layout.h (packed f16 conversions, no dynamic vector indexing), raw buffer weight loads, accumulator-init vectors prepared once - enc1 as a single-buffer recompute layer, dec1 Q/K/V chains split - dec5 as four dependent phase kernels
16 lines
503 B
Plaintext
16 lines
503 B
Plaintext
// fast numerics: intermediate activations f16 (no FP8 re-quantisation), f32 accumulation through GEMM chains
|
|
#define SWIN_NO_Q8
|
|
#define SWIN_F32ACC
|
|
#ifndef SWIN_MAX_VGPR
|
|
#define SWIN_MAX_VGPR 192
|
|
#endif
|
|
#ifndef SWIN_MLP_R
|
|
#define SWIN_MLP_R(NW) 1
|
|
#endif
|
|
#define SWIN_SCALAR_BLOAD // scalar-base weight loads: faster here (no spills), slower in the 8-wave layers
|
|
#include "swin_block.h"
|
|
#ifndef SWIN_NW
|
|
#define SWIN_NW 4
|
|
#endif
|
|
SWIN_MODULE_W(rrlite_enc1_4x4, 64, 2, SWIN_NW, 96, false, CommonParams, 0)
|