Chao Liu b491ebf384 FP16 data in-register transpose (#41) ...

* start fixing 16bit data packing

* adding StaticTensor

* adding StaticTensor

* adding StaticTensor

* add missing constexpr

* adding static tensor

* adding static tensor

* adding transpose

* add inline asm for transpose 2x2 of half_t

* add general transpose_vectors(), but have unnecessary register initialization using v_mov

* fix unnecessary register initialization in transpose_vector by using more pass-by-reference

* add hardcoded logic for NHWC wrw

* improve asm for v_pack

* make ThreadwiseTensorSliceTransfer_v3r2 support any tensor

* tweak

* reorganize file

2021-11-15 10:05:58 -06:00

..

debug.hpp

Tweak GEMM kernel (#38)

2021-10-06 11:12:36 -05:00

device_convolution_backward_data_implicit_gemm_v4r1_xdlops_nhwc_kyxc_nhwk.hpp

add nchw atomic , nhwc and nhwc atomic method for backward weight (#30)

2021-10-19 18:42:34 -05:00

device_convolution_backward_data_implicit_gemm_v4r1r2_xdlops_nhwc_kyxc_nhwk_1x1.hpp

FP16 data in-register transpose (#41)

2021-11-15 10:05:58 -06:00

device_convolution_backward_data_implicit_gemm_v4r1r2_xdlops_nhwc_kyxc_nhwk.hpp

Tweak GEMM kernel (#38)

2021-10-06 11:12:36 -05:00

device_convolution_backward_weight_implicit_gemm_v4r4r2_xdlops_atomic_nchw_kcyx_nkhw.hpp

[Bug Fix] GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_v2r4 loop issue (#44)

2021-10-27 09:39:18 -05:00

device_convolution_backward_weight_implicit_gemm_v4r4r2_xdlops_nchw_kcyx_nkhw.hpp

add nchw atomic , nhwc and nhwc atomic method for backward weight (#30)

2021-10-19 18:42:34 -05:00

device_convolution_backward_weight_implicit_gemm_v4r4r4_xdlops_atomic_nhwc_kyxc_nhwk.hpp

[Bug Fix] GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_v2r4 loop issue (#44)

2021-10-27 09:39:18 -05:00

device_convolution_backward_weight_implicit_gemm_v4r4r4_xdlops_nhwc_kyxc_nhwk.hpp

add nchw atomic , nhwc and nhwc atomic method for backward weight (#30)

2021-10-19 18:42:34 -05:00

device_convolution_backward_weight_implicit_gemm_v4r4r5_xdlops_atomic_nhwc_kyxc_nhwk.hpp

[Bug Fix] GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_v2r4 loop issue (#44)

2021-10-27 09:39:18 -05:00

device_convolution_forward_implicit_gemm_v4r4_dlops_nchw_kcyx_nkhw.hpp

rename

2021-08-11 00:08:42 +00:00

device_convolution_forward_implicit_gemm_v4r4r2_dlops_nhwc_kyxc_nhwk.hpp

ckProfiler and device-level XDL GEMM operator (#48)

2021-11-14 11:28:32 -06:00

device_convolution_forward_implicit_gemm_v4r4r2_xdlops_nchw_kcyx_nkhw.hpp

Backward weight v4r4r2 with xdlops (#18)

2021-08-30 22:49:17 -05:00

device_convolution_forward_implicit_gemm_v4r4r4_xdlops_nhwc_kyxc_nhwk.hpp

ckProfiler and device-level XDL GEMM operator (#48)

2021-11-14 11:28:32 -06:00

device_convolution_forward_implicit_gemm_v5r1_dlops_nchw_kcyx_nkhw.hpp

rename

2021-08-10 23:45:36 +00:00

device_convolution_forward_implicit_gemm_v6r1_dlops_nchw_kcyx_nkhw.hpp

make innner product compatiable on gfx900

2021-08-11 09:42:53 -05:00

device_gemm_xdlops_km_kn_mn.hpp

Tweak GEMM kernel (#38)

2021-10-06 11:12:36 -05:00

device_gemm_xdlops_km_kn_nm.hpp

Tweak GEMM kernel (#38)

2021-10-06 11:12:36 -05:00

device_gemm_xdlops_km_nk_mn.hpp

Tweak GEMM kernel (#38)

2021-10-06 11:12:36 -05:00

device_gemm_xdlops_km_nk_nm.hpp

Tweak GEMM kernel (#38)

2021-10-06 11:12:36 -05:00

device_gemm_xdlops_mk_kn_mn.hpp

Tweak GEMM kernel (#38)

2021-10-06 11:12:36 -05:00

device_gemm_xdlops_mk_kn_nm.hpp

Tweak GEMM kernel (#38)

2021-10-06 11:12:36 -05:00

device_gemm_xdlops_mk_nk_mn.hpp

Tweak GEMM kernel (#38)

2021-10-06 11:12:36 -05:00

device_gemm_xdlops_mk_nk_nm.hpp

Tweak GEMM kernel (#38)

2021-10-06 11:12:36 -05:00

driver_contraction_dlops_v1r2.hpp

rename

2021-08-11 00:08:42 +00:00

driver_convolution_forward_implicit_gemm_v5r1_dlops_nchw_kcyx_nkhw_outpad.hpp

rename

2021-08-11 00:08:42 +00:00

driver_convolution_forward_implicit_gemm_v5r1_dlops_nchw_kcyx_nkhw.hpp

rename

2021-08-11 00:08:42 +00:00

driver_gemm_dlops_v1r2.hpp

rename

2021-08-11 00:08:42 +00:00

driver_gemm_dlops_v1r3.hpp

rename

2021-08-11 00:08:42 +00:00

driver_gemm_xdlops_v2r3.hpp

ckProfiler and device-level XDL GEMM operator (#48)

2021-11-14 11:28:32 -06:00

driver_gemm_xdlops_v2r4.hpp

[Bug Fix] GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_v2r4 loop issue (#44)

2021-10-27 09:39:18 -05:00