Hansung Kim 986d507223 flash: Fix single-tile GEMM for warp-specialized ...

With 4 warps, we can only do 32x64 GEMM; serialize 64x64 into 2 32x64
GEMM calls by split by the row.

2024-08-30 17:12:46 -07:00

..

bad_apple

dma and demo kernels

2024-06-07 18:11:19 -07:00

basic

adding tensor regression test.

2023-11-14 05:37:46 -08:00

demo

adding opencl convolution benchmark

2023-11-14 22:31:30 -08:00

diverge

adding tensor regression test.

2023-11-14 05:37:46 -08:00

dogfood

Vortex 2.0 changes:

2023-11-10 02:47:05 -08:00

fence

adding tensor regression test.

2023-11-14 05:37:46 -08:00

flash_attention

flash: Fix single-tile GEMM for warp-specialized

2024-08-30 17:12:46 -07:00

flops

Add regression flops

2024-04-24 21:10:21 -07:00

idle

idle: Use barriers instead to hang the core

2024-06-22 01:37:00 -07:00

io_addr

adding tensor regression test.

2023-11-14 05:37:46 -08:00

mstress

adding tensor regression test.

2023-11-14 05:37:46 -08:00

no_mf_ext

adding tensor regression test.

2023-11-14 05:37:46 -08:00

no_smem

adding tensor regression test.

2023-11-14 05:37:46 -08:00

printf

adding tensor regression test.

2023-11-14 05:37:46 -08:00

rickroll

dma and demo kernels

2024-06-07 18:11:19 -07:00

sgemm_gemmini

make non dma gemmini use 64x64 tile size

2024-06-19 17:45:01 -07:00

sgemm_gemmini_dma

fp16 kernel

2024-08-06 02:43:44 -07:00

sgemm_gemmini_duo

sgemm_gemmini_duo: Check in serialized kernel as separate file

2024-06-12 22:44:14 -07:00

sgemm_tcore

sgemm_impl: Accept threads_per_threadblock in load_tile_to_smem

2024-08-29 21:43:57 -07:00

sgemm_wg

Add args.bin to ELF

2024-06-06 15:19:39 -07:00

sgemmx

minor udpate

2023-11-27 02:21:47 -08:00

sort

adding tensor regression test.

2023-11-14 05:37:46 -08:00

vecaddx

vecaddx: Hardcode args/input device address to match chipyard

2024-03-27 15:15:52 -07:00

common.mk

tensor: Change B in-memory layout to column-major

2024-08-12 15:22:07 -07:00

Makefile

minor udpate

2023-11-27 02:21:47 -08:00