kernel instr estrechas ds vgpr thmax
kernel_add_id 168 0 0 183 1024
kernel_arange_f32 130 0 0 183 1024
kernel_argmax_f32 249 0 3 206 1024
kernel_argsort_f32_i32_asc 316 0 14 185 1024
kernel_argsort_f32_i32_desc 283 0 14 185 1024
kernel_argsort_merge_f32_i32_asc 748 0 0 186 1024
kernel_argsort_merge_f32_i32_desc 753 0 0 206 1024
kernel_bin_fuse_f16_f16_f16 FALLA
kernel_bin_fuse_f16_f16_f16_4 FALLA
kernel_bin_fuse_f32_f32_f32 FALLA
kernel_bin_fuse_f32_f32_f32_4 FALLA
kernel_col2im_1d_bf16 273 0 0 183 1024
kernel_col2im_1d_f16 270 0 0 182 1024
kernel_col2im_1d_f32 267 0 0 182 1024
kernel_concat_bf16 303 1 0 183 1024
kernel_concat_f16 278 1 0 206 1024
kernel_concat_f32 276 0 0 206 1024
kernel_concat_i16 278 1 0 206 1024
kernel_concat_i32 277 0 0 206 1024
kernel_concat_i64 278 0 0 206 1024
kernel_concat_i8 277 1 0 193 1024
kernel_conv_2d_dw_f16_f32 468 1 0 206 1024
kernel_conv_2d_dw_f32_f32 467 0 0 206 1024
kernel_conv_2d_dw_tiled_f16_f32 442 1 0 187 1024
kernel_conv_2d_dw_tiled_f32_f32 442 0 0 187 1024
kernel_conv_2d_f16_f32 1170 1 0 187 1024
kernel_conv_2d_f32_f32 1169 0 0 187 1024
kernel_conv_3d_f16_f32 907 1 0 187 1024
kernel_conv_3d_f32_f32 908 0 0 187 1024
kernel_conv_transpose_1d_f16_f32 266 0 0 187 1024
kernel_conv_transpose_1d_f32_f32 265 0 0 187 1024
kernel_conv_transpose_2d_f16_f32 779 0 2 187 1024
kernel_conv_transpose_2d_f32_f32 777 0 2 187 1024
kernel_count_equal_i32 FALLA
kernel_cpy_bf16_bf16 936 1 0 188 1024
kernel_cpy_bf16_f32 928 1 0 155 1024
kernel_cpy_contig_bf16_bf16 116 1 0 186 1024
kernel_cpy_contig_bf16_f32 154 0 0 186 1024
kernel_cpy_contig_f16_f16 124 1 0 186 1024
kernel_cpy_contig_f16_f32 124 0 0 186 1024
kernel_cpy_contig_f32_bf16 156 0 0 186 1024
kernel_cpy_contig_f32_f16 126 0 0 186 1024
kernel_cpy_contig_f32_f32 122 0 0 186 1024
kernel_cpy_contig_f32_i32 123 0 0 186 1024
kernel_cpy_contig_i32_f32 122 0 0 186 1024
kernel_cpy_contig_i32_i32 122 0 0 186 1024
kernel_cpy_f16_f16 966 1 0 188 1024
kernel_cpy_f16_f32 964 1 0 188 1024
kernel_cpy_f32_bf16 934 0 0 188 1024
kernel_cpy_f32_f16 966 0 0 188 1024
kernel_cpy_f32_f32 965 0 0 188 1024
kernel_cpy_f32_i32 964 0 0 188 1024
kernel_cpy_f32_iq4_nl 3307 0 0 188 1024
kernel_cpy_f32_q1_0 1343 8 0 188 1024
kernel_cpy_f32_q2_0 2397 64 0 188 1024
kernel_cpy_f32_q4_0 1416 0 0 188 1024
kernel_cpy_f32_q4_1 1457 0 0 188 1024
kernel_cpy_f32_q5_0 1609 0 0 188 1024
kernel_cpy_f32_q5_1 1555 0 0 188 1024
kernel_cpy_f32_q8_0 1431 0 0 188 1024
kernel_cpy_f32_turbo2 3652 8 0 232 448
kernel_cpy_f32_turbo3 3969 16 0 248 448
kernel_cpy_f32_turbo4 4253 8 0 252 448
kernel_cpy_i32_f32 964 0 0 188 1024
kernel_cpy_i32_i32 964 0 0 188 1024
kernel_cpy_q1_0_f16 1037 2 0 188 1024
kernel_cpy_q1_0_f32 1012 2 0 188 1024
kernel_cpy_q2_0_f16 1027 4 0 188 1024
kernel_cpy_q2_0_f32 1021 4 0 188 1024
kernel_cpy_q4_0_f16 1036 9 0 188 1024
kernel_cpy_q4_0_f32 1009 9 0 188 1024
kernel_cpy_q4_1_f16 1034 10 0 188 1024
kernel_cpy_q4_1_f32 1006 10 0 188 1024
kernel_cpy_q5_0_f16 1121 4 0 188 1024
kernel_cpy_q5_0_f32 1100 4 0 188 1024
kernel_cpy_q5_1_f16 1124 4 0 188 1024
kernel_cpy_q5_1_f32 1094 4 0 188 1024
kernel_cpy_q8_0_f16 1004 16 0 188 1024
kernel_cpy_q8_0_f32 1006 16 0 188 1024
kernel_cpy_turbo2_f32 2872 8 0 232 448
kernel_cpy_turbo3_f32 2951 16 0 248 448
kernel_cpy_turbo4_f32 2885 8 0 252 448
kernel_cumsum_add_f32 462 0 0 204 1024
kernel_cumsum_blk_f32 588 1 5 208 1024
kernel_diag_f32 159 0 0 179 1024
kernel_dsv4_hc_comb_f32 260 0 0 177 1024
kernel_dsv4_hc_post_f32 333 0 0 177 1024
kernel_dsv4_hc_pre_f32 233 0 0 183 1024
kernel_flash_attn_ext_bf16_dk112_dv112 FALLA
kernel_flash_attn_ext_bf16_dk128_dv128 FALLA
kernel_flash_attn_ext_bf16_dk192_dv128 FALLA
kernel_flash_attn_ext_bf16_dk192_dv192 FALLA
kernel_flash_attn_ext_bf16_dk256_dv256 FALLA
kernel_flash_attn_ext_bf16_dk320_dv256 FALLA
kernel_flash_attn_ext_bf16_dk32_dv32 FALLA
kernel_flash_attn_ext_bf16_dk40_dv40 FALLA
kernel_flash_attn_ext_bf16_dk48_dv48 FALLA
kernel_flash_attn_ext_bf16_dk512_dv512 FALLA
kernel_flash_attn_ext_bf16_dk576_dv512 FALLA
kernel_flash_attn_ext_bf16_dk64_dv64 FALLA
kernel_flash_attn_ext_bf16_dk72_dv72 FALLA
kernel_flash_attn_ext_bf16_dk80_dv80 FALLA
kernel_flash_attn_ext_bf16_dk96_dv96 FALLA
kernel_flash_attn_ext_blk FALLA
kernel_flash_attn_ext_f16_dk112_dv112 FALLA
kernel_flash_attn_ext_f16_dk128_dv128 FALLA
kernel_flash_attn_ext_f16_dk192_dv128 FALLA
kernel_flash_attn_ext_f16_dk192_dv192 FALLA
kernel_flash_attn_ext_f16_dk256_dv256 FALLA
kernel_flash_attn_ext_f16_dk320_dv256 FALLA
kernel_flash_attn_ext_f16_dk32_dv32 FALLA
kernel_flash_attn_ext_f16_dk40_dv40 FALLA
kernel_flash_attn_ext_f16_dk48_dv48 FALLA
kernel_flash_attn_ext_f16_dk512_dv512 FALLA
kernel_flash_attn_ext_f16_dk576_dv512 FALLA
kernel_flash_attn_ext_f16_dk64_dv64 FALLA
kernel_flash_attn_ext_f16_dk72_dv72 FALLA
kernel_flash_attn_ext_f16_dk80_dv80 FALLA
kernel_flash_attn_ext_f16_dk96_dv96 FALLA
kernel_flash_attn_ext_f32_dk112_dv112 FALLA
kernel_flash_attn_ext_f32_dk128_dv128 FALLA
kernel_flash_attn_ext_f32_dk192_dv128 FALLA
kernel_flash_attn_ext_f32_dk192_dv192 FALLA
kernel_flash_attn_ext_f32_dk256_dv256 FALLA
kernel_flash_attn_ext_f32_dk320_dv256 FALLA
kernel_flash_attn_ext_f32_dk32_dv32 FALLA
kernel_flash_attn_ext_f32_dk40_dv40 FALLA
kernel_flash_attn_ext_f32_dk48_dv48 FALLA
kernel_flash_attn_ext_f32_dk512_dv512 FALLA
kernel_flash_attn_ext_f32_dk576_dv512 FALLA
kernel_flash_attn_ext_f32_dk64_dv64 FALLA
kernel_flash_attn_ext_f32_dk72_dv72 FALLA
kernel_flash_attn_ext_f32_dk80_dv80 FALLA
kernel_flash_attn_ext_f32_dk96_dv96 FALLA
kernel_flash_attn_ext_pad FALLA
kernel_flash_attn_ext_pf_amd_dk128_kf16_vf16 4655 16 422 254 256
kernel_flash_attn_ext_pf_amd_dk128_kf16_vq4_0 4719 17 422 254 256
kernel_flash_attn_ext_pf_amd_dk128_kf16_vq8_0 4707 20 422 254 256
kernel_flash_attn_ext_pf_amd_dk128_kf16_vturbo2 4703 18 422 254 256
kernel_flash_attn_ext_pf_amd_dk128_kf16_vturbo3 4723 19 422 255 256
kernel_flash_attn_ext_pf_amd_dk128_kf16_vturbo4 4720 19 422 254 256
kernel_flash_attn_ext_pf_amd_dk128_kq4_0_vf16 4721 17 422 254 256
kernel_flash_attn_ext_pf_amd_dk128_kq4_0_vq4_0 4795 18 422 254 256
kernel_flash_attn_ext_pf_amd_dk128_kq4_0_vq8_0 4787 21 422 254 256
kernel_flash_attn_ext_pf_amd_dk128_kq8_0_vf16 4706 20 422 254 256
kernel_flash_attn_ext_pf_amd_dk128_kq8_0_vq4_0 4786 21 422 254 256
kernel_flash_attn_ext_pf_amd_dk128_kq8_0_vq8_0 4772 24 422 254 256
kernel_flash_attn_ext_pf_amd_dk128_kq8_0_vturbo2 4769 22 422 254 256
kernel_flash_attn_ext_pf_amd_dk128_kq8_0_vturbo3 4787 23 422 254 256
kernel_flash_attn_ext_pf_amd_dk128_kq8_0_vturbo4 4789 23 422 254 256
kernel_flash_attn_ext_pf_amd_dk128_kturbo2_vf16 4701 18 422 254 256
kernel_flash_attn_ext_pf_amd_dk128_kturbo2_vq8_0 4772 22 422 254 256
kernel_flash_attn_ext_pf_amd_dk128_kturbo2_vturbo2 4724 20 422 254 256
kernel_flash_attn_ext_pf_amd_dk128_kturbo2_vturbo3 4777 21 422 254 256
kernel_flash_attn_ext_pf_amd_dk128_kturbo2_vturbo4 4774 21 422 254 256
kernel_flash_attn_ext_pf_amd_dk128_kturbo3_vf16 4725 19 422 254 256
kernel_flash_attn_ext_pf_amd_dk128_kturbo3_vq8_0 4793 23 422 254 256
kernel_flash_attn_ext_pf_amd_dk128_kturbo3_vturbo2 4780 21 422 254 256
kernel_flash_attn_ext_pf_amd_dk128_kturbo3_vturbo3 4767 22 422 254 256
kernel_flash_attn_ext_pf_amd_dk128_kturbo3_vturbo4 4796 22 422 254 256
kernel_flash_attn_ext_pf_amd_dk128_kturbo4_vf16 4725 19 422 254 256
kernel_flash_attn_ext_pf_amd_dk128_kturbo4_vq8_0 4795 23 422 254 256
kernel_flash_attn_ext_pf_amd_dk128_kturbo4_vturbo2 4780 21 422 254 256
kernel_flash_attn_ext_pf_amd_dk128_kturbo4_vturbo3 4800 22 422 254 256
kernel_flash_attn_ext_pf_amd_dk128_kturbo4_vturbo4 4751 22 422 254 256
kernel_flash_attn_ext_pf_amd_dk256_kf16_vf16 3815 8 462 255 256
kernel_flash_attn_ext_pf_amd_dk256_kf16_vq4_0 3878 9 462 255 256
kernel_flash_attn_ext_pf_amd_dk256_kf16_vq8_0 3867 12 462 255 256
kernel_flash_attn_ext_pf_amd_dk256_kf16_vturbo2 3878 9 462 255 256
kernel_flash_attn_ext_pf_amd_dk256_kf16_vturbo3 3899 10 462 255 256
kernel_flash_attn_ext_pf_amd_dk256_kf16_vturbo4 3894 10 462 255 256
kernel_flash_attn_ext_pf_amd_dk256_kq4_0_vf16 3880 9 462 255 256
kernel_flash_attn_ext_pf_amd_dk256_kq4_0_vq4_0 3961 10 462 255 256
kernel_flash_attn_ext_pf_amd_dk256_kq4_0_vq8_0 3951 13 462 255 256
kernel_flash_attn_ext_pf_amd_dk256_kq8_0_vf16 3874 12 462 255 256
kernel_flash_attn_ext_pf_amd_dk256_kq8_0_vq4_0 3954 13 462 255 256
kernel_flash_attn_ext_pf_amd_dk256_kq8_0_vq8_0 3949 16 462 255 256
kernel_flash_attn_ext_pf_amd_dk256_kq8_0_vturbo2 3954 13 462 255 256
kernel_flash_attn_ext_pf_amd_dk256_kq8_0_vturbo3 3977 14 462 255 256
kernel_flash_attn_ext_pf_amd_dk256_kq8_0_vturbo4 3972 14 462 255 256
kernel_flash_attn_ext_pf_amd_dk256_kturbo2_vf16 3882 9 462 255 256
kernel_flash_attn_ext_pf_amd_dk256_kturbo2_vq8_0 3946 13 462 255 256
kernel_flash_attn_ext_pf_amd_dk256_kturbo2_vturbo2 3919 10 462 255 256
kernel_flash_attn_ext_pf_amd_dk256_kturbo2_vturbo3 3979 11 462 255 256
kernel_flash_attn_ext_pf_amd_dk256_kturbo2_vturbo4 3972 11 462 255 256
kernel_flash_attn_ext_pf_amd_dk256_kturbo3_vf16 3905 10 462 255 256
kernel_flash_attn_ext_pf_amd_dk256_kturbo3_vq8_0 3977 14 462 255 256
kernel_flash_attn_ext_pf_amd_dk256_kturbo3_vturbo2 3982 11 462 255 256
kernel_flash_attn_ext_pf_amd_dk256_kturbo3_vturbo3 3962 12 462 255 256
kernel_flash_attn_ext_pf_amd_dk256_kturbo3_vturbo4 3996 12 462 255 256
kernel_flash_attn_ext_pf_amd_dk256_kturbo4_vf16 3896 10 462 255 256
kernel_flash_attn_ext_pf_amd_dk256_kturbo4_vq8_0 3967 14 462 255 256
kernel_flash_attn_ext_pf_amd_dk256_kturbo4_vturbo2 3970 11 462 255 256
kernel_flash_attn_ext_pf_amd_dk256_kturbo4_vturbo3 3988 12 462 255 256
kernel_flash_attn_ext_pf_amd_dk256_kturbo4_vturbo4 3940 12 462 255 256
kernel_flash_attn_ext_pf_amd_dk512_kf16_vf16 2517 2 356 245 448
kernel_flash_attn_ext_pf_amd_dk512_kq4_0_vf16 2583 3 356 255 448
kernel_flash_attn_ext_pf_amd_dk512_kq8_0_vf16 2567 6 356 248 448
kernel_flash_attn_ext_pf_amd_dk576_dv512_kf16_vf16 2556 2 356 183 448
kernel_flash_attn_ext_pf_amd_dk576_dv512_kq8_0_vf16 2560 6 356 181 448
kernel_flash_attn_ext_pf_amd_dk64_kf16_vf16 3127 16 246 223 256
kernel_flash_attn_ext_pf_amd_dk64_kf16_vq4_0 3193 17 246 232 256
kernel_flash_attn_ext_pf_amd_dk64_kf16_vq8_0 3184 20 247 251 256
kernel_flash_attn_ext_pf_amd_dk64_kq4_0_vf16 3194 17 246 232 256
kernel_flash_attn_ext_pf_amd_dk64_kq4_0_vq4_0 3257 18 246 223 256
kernel_flash_attn_ext_pf_amd_dk64_kq4_0_vq8_0 3248 21 246 224 256
kernel_flash_attn_ext_pf_amd_dk64_kq8_0_vf16 3182 20 247 251 256
kernel_flash_attn_ext_pf_amd_dk64_kq8_0_vq4_0 3246 21 246 224 256
kernel_flash_attn_ext_pf_amd_dk64_kq8_0_vq8_0 3234 24 246 223 256
kernel_flash_attn_ext_pf_amd_dk72_kf16_vf16 4259 16 340 255 256
kernel_flash_attn_ext_q4_0_dk112_dv112 FALLA
kernel_flash_attn_ext_q4_0_dk128_dv128 FALLA
kernel_flash_attn_ext_q4_0_dk192_dv128 FALLA
kernel_flash_attn_ext_q4_0_dk192_dv192 FALLA
kernel_flash_attn_ext_q4_0_dk256_dv256 FALLA
kernel_flash_attn_ext_q4_0_dk320_dv256 FALLA
kernel_flash_attn_ext_q4_0_dk32_dv32 FALLA
kernel_flash_attn_ext_q4_0_dk40_dv40 FALLA
kernel_flash_attn_ext_q4_0_dk48_dv48 FALLA
kernel_flash_attn_ext_q4_0_dk512_dv512 FALLA
kernel_flash_attn_ext_q4_0_dk576_dv512 FALLA
kernel_flash_attn_ext_q4_0_dk64_dv64 FALLA
kernel_flash_attn_ext_q4_0_dk72_dv72 FALLA
kernel_flash_attn_ext_q4_0_dk80_dv80 FALLA
kernel_flash_attn_ext_q4_0_dk96_dv96 FALLA
kernel_flash_attn_ext_q4_1_dk112_dv112 FALLA
kernel_flash_attn_ext_q4_1_dk128_dv128 FALLA
kernel_flash_attn_ext_q4_1_dk192_dv128 FALLA
kernel_flash_attn_ext_q4_1_dk192_dv192 FALLA
kernel_flash_attn_ext_q4_1_dk256_dv256 FALLA
kernel_flash_attn_ext_q4_1_dk320_dv256 FALLA
kernel_flash_attn_ext_q4_1_dk32_dv32 FALLA
kernel_flash_attn_ext_q4_1_dk40_dv40 FALLA
kernel_flash_attn_ext_q4_1_dk48_dv48 FALLA
kernel_flash_attn_ext_q4_1_dk512_dv512 FALLA
kernel_flash_attn_ext_q4_1_dk576_dv512 FALLA
kernel_flash_attn_ext_q4_1_dk64_dv64 FALLA
kernel_flash_attn_ext_q4_1_dk72_dv72 FALLA
kernel_flash_attn_ext_q4_1_dk80_dv80 FALLA
kernel_flash_attn_ext_q4_1_dk96_dv96 FALLA
kernel_flash_attn_ext_q5_0_dk112_dv112 FALLA
kernel_flash_attn_ext_q5_0_dk128_dv128 FALLA
kernel_flash_attn_ext_q5_0_dk192_dv128 FALLA
kernel_flash_attn_ext_q5_0_dk192_dv192 FALLA
kernel_flash_attn_ext_q5_0_dk256_dv256 FALLA
kernel_flash_attn_ext_q5_0_dk320_dv256 FALLA
kernel_flash_attn_ext_q5_0_dk32_dv32 FALLA
kernel_flash_attn_ext_q5_0_dk40_dv40 FALLA
kernel_flash_attn_ext_q5_0_dk48_dv48 FALLA
kernel_flash_attn_ext_q5_0_dk512_dv512 FALLA
kernel_flash_attn_ext_q5_0_dk576_dv512 FALLA
kernel_flash_attn_ext_q5_0_dk64_dv64 FALLA
kernel_flash_attn_ext_q5_0_dk72_dv72 FALLA
kernel_flash_attn_ext_q5_0_dk80_dv80 FALLA
kernel_flash_attn_ext_q5_0_dk96_dv96 FALLA
kernel_flash_attn_ext_q5_1_dk112_dv112 FALLA
kernel_flash_attn_ext_q5_1_dk128_dv128 FALLA
kernel_flash_attn_ext_q5_1_dk192_dv128 FALLA
kernel_flash_attn_ext_q5_1_dk192_dv192 FALLA
kernel_flash_attn_ext_q5_1_dk256_dv256 FALLA
kernel_flash_attn_ext_q5_1_dk320_dv256 FALLA
kernel_flash_attn_ext_q5_1_dk32_dv32 FALLA
kernel_flash_attn_ext_q5_1_dk40_dv40 FALLA
kernel_flash_attn_ext_q5_1_dk48_dv48 FALLA
kernel_flash_attn_ext_q5_1_dk512_dv512 FALLA
kernel_flash_attn_ext_q5_1_dk576_dv512 FALLA
kernel_flash_attn_ext_q5_1_dk64_dv64 FALLA
kernel_flash_attn_ext_q5_1_dk72_dv72 FALLA
kernel_flash_attn_ext_q5_1_dk80_dv80 FALLA
kernel_flash_attn_ext_q5_1_dk96_dv96 FALLA
kernel_flash_attn_ext_q8_0_dk112_dv112 FALLA
kernel_flash_attn_ext_q8_0_dk128_dv128 FALLA
kernel_flash_attn_ext_q8_0_dk192_dv128 FALLA
kernel_flash_attn_ext_q8_0_dk192_dv192 FALLA
kernel_flash_attn_ext_q8_0_dk256_dv256 FALLA
kernel_flash_attn_ext_q8_0_dk320_dv256 FALLA
kernel_flash_attn_ext_q8_0_dk32_dv32 FALLA
kernel_flash_attn_ext_q8_0_dk40_dv40 FALLA
kernel_flash_attn_ext_q8_0_dk48_dv48 FALLA
kernel_flash_attn_ext_q8_0_dk512_dv512 FALLA
kernel_flash_attn_ext_q8_0_dk576_dv512 FALLA
kernel_flash_attn_ext_q8_0_dk64_dv64 FALLA
kernel_flash_attn_ext_q8_0_dk72_dv72 FALLA
kernel_flash_attn_ext_q8_0_dk80_dv80 FALLA
kernel_flash_attn_ext_q8_0_dk96_dv96 FALLA
kernel_flash_attn_ext_tile_amd_dk128_kf16_vf16 760 2 7 186 1024
kernel_flash_attn_ext_tile_amd_dk128_kf16_vf16_w64 761 2 7 186 1024
kernel_flash_attn_ext_tile_amd_dk128_kf16_vq4_0 782 3 7 187 1024
kernel_flash_attn_ext_tile_amd_dk128_kf16_vq4_0_w64 782 3 7 187 1024
kernel_flash_attn_ext_tile_amd_dk128_kf16_vq8_0 767 6 7 187 1024
kernel_flash_attn_ext_tile_amd_dk128_kf16_vq8_0_w64 767 6 7 187 1024
kernel_flash_attn_ext_tile_amd_dk128_kf16_vturbo2 766 4 7 232 1024
kernel_flash_attn_ext_tile_amd_dk128_kf16_vturbo2_w64 766 4 7 232 1024
kernel_flash_attn_ext_tile_amd_dk128_kf16_vturbo3 790 5 7 248 1024
kernel_flash_attn_ext_tile_amd_dk128_kf16_vturbo3_w64 790 5 7 248 1024
kernel_flash_attn_ext_tile_amd_dk128_kf16_vturbo4 791 5 7 252 1024
kernel_flash_attn_ext_tile_amd_dk128_kf16_vturbo4_w64 791 5 7 252 1024
kernel_flash_attn_ext_tile_amd_dk128_kq4_0_vf16 783 3 7 186 1024
kernel_flash_attn_ext_tile_amd_dk128_kq4_0_vf16_w64 784 3 7 186 1024
kernel_flash_attn_ext_tile_amd_dk128_kq4_0_vq4_0 857 4 7 200 1024
kernel_flash_attn_ext_tile_amd_dk128_kq4_0_vq4_0_w64 858 4 7 200 1024
kernel_flash_attn_ext_tile_amd_dk128_kq4_0_vq8_0 842 7 7 200 1024
kernel_flash_attn_ext_tile_amd_dk128_kq4_0_vq8_0_w64 843 7 7 200 1024
kernel_flash_attn_ext_tile_amd_dk128_kq8_0_vf16 775 6 7 186 1024
kernel_flash_attn_ext_tile_amd_dk128_kq8_0_vf16_w64 776 6 7 186 1024
kernel_flash_attn_ext_tile_amd_dk128_kq8_0_vq4_0 849 7 7 200 1024
kernel_flash_attn_ext_tile_amd_dk128_kq8_0_vq4_0_w64 850 7 7 200 1024
kernel_flash_attn_ext_tile_amd_dk128_kq8_0_vq8_0 834 10 7 232 1024
kernel_flash_attn_ext_tile_amd_dk128_kq8_0_vq8_0_w64 835 10 7 232 1024
kernel_flash_attn_ext_tile_amd_dk128_kq8_0_vturbo2 833 8 7 232 1024
kernel_flash_attn_ext_tile_amd_dk128_kq8_0_vturbo2_w64 834 8 7 232 1024
kernel_flash_attn_ext_tile_amd_dk128_kq8_0_vturbo3 857 9 7 248 1024
kernel_flash_attn_ext_tile_amd_dk128_kq8_0_vturbo3_w64 858 9 7 248 1024
kernel_flash_attn_ext_tile_amd_dk128_kq8_0_vturbo4 847 9 7 252 1024
kernel_flash_attn_ext_tile_amd_dk128_kq8_0_vturbo4_w64 848 9 7 252 1024
kernel_flash_attn_ext_tile_amd_dk128_kturbo2_vf16 773 4 7 232 1024
kernel_flash_attn_ext_tile_amd_dk128_kturbo2_vf16_w64 774 4 7 232 1024
kernel_flash_attn_ext_tile_amd_dk128_kturbo2_vq8_0 833 8 7 232 1024
kernel_flash_attn_ext_tile_amd_dk128_kturbo2_vq8_0_w64 834 8 7 232 1024
kernel_flash_attn_ext_tile_amd_dk128_kturbo2_vturbo2 835 6 7 232 1024
kernel_flash_attn_ext_tile_amd_dk128_kturbo2_vturbo2_w64 836 6 7 232 1024
kernel_flash_attn_ext_tile_amd_dk128_kturbo2_vturbo3 850 7 7 248 1024
kernel_flash_attn_ext_tile_amd_dk128_kturbo2_vturbo3_w64 851 7 7 248 1024
kernel_flash_attn_ext_tile_amd_dk128_kturbo2_vturbo4 842 7 7 252 1024
kernel_flash_attn_ext_tile_amd_dk128_kturbo2_vturbo4_w64 843 7 7 252 1024
kernel_flash_attn_ext_tile_amd_dk128_kturbo3_vf16 789 5 7 248 1024
kernel_flash_attn_ext_tile_amd_dk128_kturbo3_vf16_w64 790 5 7 248 1024
kernel_flash_attn_ext_tile_amd_dk128_kturbo3_vq8_0 849 9 7 248 1024
kernel_flash_attn_ext_tile_amd_dk128_kturbo3_vq8_0_w64 850 9 7 248 1024
kernel_flash_attn_ext_tile_amd_dk128_kturbo3_vturbo2 842 7 7 248 1024
kernel_flash_attn_ext_tile_amd_dk128_kturbo3_vturbo2_w64 843 7 7 248 1024
kernel_flash_attn_ext_tile_amd_dk128_kturbo3_vturbo3 872 8 7 248 1024
kernel_flash_attn_ext_tile_amd_dk128_kturbo3_vturbo3_w64 872 8 7 248 1024
kernel_flash_attn_ext_tile_amd_dk128_kturbo3_vturbo4 866 8 7 252 1024
kernel_flash_attn_ext_tile_amd_dk128_kturbo3_vturbo4_w64 867 8 7 252 1024
kernel_flash_attn_ext_tile_amd_dk128_kturbo4_vf16 791 5 7 252 1024
kernel_flash_attn_ext_tile_amd_dk128_kturbo4_vf16_w64 792 5 7 252 1024
kernel_flash_attn_ext_tile_amd_dk128_kturbo4_vq8_0 848 9 7 252 1024
kernel_flash_attn_ext_tile_amd_dk128_kturbo4_vq8_0_w64 849 9 7 252 1024
kernel_flash_attn_ext_tile_amd_dk128_kturbo4_vturbo2 842 7 7 252 1024
kernel_flash_attn_ext_tile_amd_dk128_kturbo4_vturbo2_w64 843 7 7 252 1024
kernel_flash_attn_ext_tile_amd_dk128_kturbo4_vturbo3 866 8 7 252 1024
kernel_flash_attn_ext_tile_amd_dk128_kturbo4_vturbo3_w64 867 8 7 252 1024
kernel_flash_attn_ext_tile_amd_dk128_kturbo4_vturbo4 865 8 7 252 1024
kernel_flash_attn_ext_tile_amd_dk128_kturbo4_vturbo4_w64 865 8 7 252 1024
kernel_flash_attn_ext_tile_amd_dk256_kf16_vf16 828 2 9 186 1024
kernel_flash_attn_ext_tile_amd_dk256_kf16_vf16_w64 761 2 7 186 1024
kernel_flash_attn_ext_tile_amd_dk256_kf16_vq4_0 858 3 9 187 1024
kernel_flash_attn_ext_tile_amd_dk256_kf16_vq4_0_w64 782 3 7 187 1024
kernel_flash_attn_ext_tile_amd_dk256_kf16_vq8_0 843 6 9 187 1024
kernel_flash_attn_ext_tile_amd_dk256_kf16_vq8_0_w64 767 6 7 187 1024
kernel_flash_attn_ext_tile_amd_dk256_kf16_vturbo2 845 3 9 232 1024
kernel_flash_attn_ext_tile_amd_dk256_kf16_vturbo2_w64 777 3 7 232 1024
kernel_flash_attn_ext_tile_amd_dk256_kf16_vturbo3 869 4 9 248 1024
kernel_flash_attn_ext_tile_amd_dk256_kf16_vturbo3_w64 801 4 7 248 1024
kernel_flash_attn_ext_tile_amd_dk256_kf16_vturbo4 871 4 9 252 1024
kernel_flash_attn_ext_tile_amd_dk256_kf16_vturbo4_w64 795 4 7 252 1024
kernel_flash_attn_ext_tile_amd_dk256_kq4_0_vf16 859 3 9 186 1024
kernel_flash_attn_ext_tile_amd_dk256_kq4_0_vf16_w64 784 3 7 186 1024
kernel_flash_attn_ext_tile_amd_dk256_kq4_0_vq4_0 933 4 9 200 1024
kernel_flash_attn_ext_tile_amd_dk256_kq4_0_vq4_0_w64 858 4 7 200 1024
kernel_flash_attn_ext_tile_amd_dk256_kq4_0_vq8_0 918 7 9 251 1024
kernel_flash_attn_ext_tile_amd_dk256_kq4_0_vq8_0_w64 843 7 7 200 1024
kernel_flash_attn_ext_tile_amd_dk256_kq8_0_vf16 851 6 9 186 1024
kernel_flash_attn_ext_tile_amd_dk256_kq8_0_vf16_w64 776 6 7 186 1024
kernel_flash_attn_ext_tile_amd_dk256_kq8_0_vq4_0 925 7 9 246 1024
kernel_flash_attn_ext_tile_amd_dk256_kq8_0_vq4_0_w64 850 7 7 200 1024
kernel_flash_attn_ext_tile_amd_dk256_kq8_0_vq8_0 910 10 9 210 1024
kernel_flash_attn_ext_tile_amd_dk256_kq8_0_vq8_0_w64 835 10 7 232 1024
kernel_flash_attn_ext_tile_amd_dk256_kq8_0_vturbo2 912 7 9 232 1024
kernel_flash_attn_ext_tile_amd_dk256_kq8_0_vturbo2_w64 846 7 7 232 1024
kernel_flash_attn_ext_tile_amd_dk256_kq8_0_vturbo3 928 8 9 248 1024
kernel_flash_attn_ext_tile_amd_dk256_kq8_0_vturbo3_w64 869 8 7 248 1024
kernel_flash_attn_ext_tile_amd_dk256_kq8_0_vturbo4 927 8 9 252 1024
kernel_flash_attn_ext_tile_amd_dk256_kq8_0_vturbo4_w64 861 8 7 252 1024
kernel_flash_attn_ext_tile_amd_dk256_kturbo2_vf16 844 3 9 232 1024
kernel_flash_attn_ext_tile_amd_dk256_kturbo2_vf16_w64 777 3 7 232 1024
kernel_flash_attn_ext_tile_amd_dk256_kturbo2_vq8_0 904 7 9 232 1024
kernel_flash_attn_ext_tile_amd_dk256_kturbo2_vq8_0_w64 837 7 7 232 1024
kernel_flash_attn_ext_tile_amd_dk256_kturbo2_vturbo2 917 4 9 232 1024
kernel_flash_attn_ext_tile_amd_dk256_kturbo2_vturbo2_w64 848 4 7 232 1024
kernel_flash_attn_ext_tile_amd_dk256_kturbo2_vturbo3 932 5 9 248 1024
kernel_flash_attn_ext_tile_amd_dk256_kturbo2_vturbo3_w64 865 5 7 248 1024
kernel_flash_attn_ext_tile_amd_dk256_kturbo2_vturbo4 932 5 9 252 1024
kernel_flash_attn_ext_tile_amd_dk256_kturbo2_vturbo4_w64 857 5 7 252 1024
kernel_flash_attn_ext_tile_amd_dk256_kturbo3_vf16 868 4 9 248 1024
kernel_flash_attn_ext_tile_amd_dk256_kturbo3_vf16_w64 801 4 7 248 1024
kernel_flash_attn_ext_tile_amd_dk256_kturbo3_vq8_0 928 8 9 248 1024
kernel_flash_attn_ext_tile_amd_dk256_kturbo3_vq8_0_w64 861 8 7 248 1024
kernel_flash_attn_ext_tile_amd_dk256_kturbo3_vturbo2 940 5 9 248 1024
kernel_flash_attn_ext_tile_amd_dk256_kturbo3_vturbo2_w64 865 5 7 248 1024
kernel_flash_attn_ext_tile_amd_dk256_kturbo3_vturbo3 961 6 9 248 1024
kernel_flash_attn_ext_tile_amd_dk256_kturbo3_vturbo3_w64 892 6 7 248 1024
kernel_flash_attn_ext_tile_amd_dk256_kturbo3_vturbo4 956 6 9 252 1024
kernel_flash_attn_ext_tile_amd_dk256_kturbo3_vturbo4_w64 881 6 7 252 1024
kernel_flash_attn_ext_tile_amd_dk256_kturbo4_vf16 871 4 9 252 1024
kernel_flash_attn_ext_tile_amd_dk256_kturbo4_vf16_w64 796 4 7 252 1024
kernel_flash_attn_ext_tile_amd_dk256_kturbo4_vq8_0 928 8 9 252 1024
kernel_flash_attn_ext_tile_amd_dk256_kturbo4_vq8_0_w64 861 8 7 252 1024
kernel_flash_attn_ext_tile_amd_dk256_kturbo4_vturbo2 940 5 9 252 1024
kernel_flash_attn_ext_tile_amd_dk256_kturbo4_vturbo2_w64 865 5 7 252 1024
kernel_flash_attn_ext_tile_amd_dk256_kturbo4_vturbo3 956 6 9 252 1024
kernel_flash_attn_ext_tile_amd_dk256_kturbo4_vturbo3_w64 889 6 7 252 1024
kernel_flash_attn_ext_tile_amd_dk256_kturbo4_vturbo4 955 6 9 252 1024
kernel_flash_attn_ext_tile_amd_dk256_kturbo4_vturbo4_w64 879 6 7 252 1024
kernel_flash_attn_ext_tile_amd_dk384_kf16_vturbo2 931 3 11 232 1024
kernel_flash_attn_ext_tile_amd_dk384_kf16_vturbo2_w64 855 3 9 232 1024
kernel_flash_attn_ext_tile_amd_dk384_kf16_vturbo3 954 4 11 248 1024
kernel_flash_attn_ext_tile_amd_dk384_kf16_vturbo3_w64 879 4 9 248 1024
kernel_flash_attn_ext_tile_amd_dk384_kf16_vturbo4 954 4 11 252 1024
kernel_flash_attn_ext_tile_amd_dk384_kf16_vturbo4_w64 879 4 9 252 1024
kernel_flash_attn_ext_tile_amd_dk384_kq8_0_vturbo2 997 7 11 232 1024
kernel_flash_attn_ext_tile_amd_dk384_kq8_0_vturbo2_w64 915 7 9 232 1024
kernel_flash_attn_ext_tile_amd_dk384_kq8_0_vturbo3 1022 8 11 248 1024
kernel_flash_attn_ext_tile_amd_dk384_kq8_0_vturbo3_w64 947 8 9 248 1024
kernel_flash_attn_ext_tile_amd_dk384_kq8_0_vturbo4 1012 8 11 252 1024
kernel_flash_attn_ext_tile_amd_dk384_kq8_0_vturbo4_w64 939 8 9 252 1024
kernel_flash_attn_ext_tile_amd_dk384_kturbo2_vf16 937 3 11 232 1024
kernel_flash_attn_ext_tile_amd_dk384_kturbo2_vf16_w64 863 3 9 232 1024
kernel_flash_attn_ext_tile_amd_dk384_kturbo2_vq8_0 997 7 11 232 1024
kernel_flash_attn_ext_tile_amd_dk384_kturbo2_vq8_0_w64 915 7 9 232 1024
kernel_flash_attn_ext_tile_amd_dk384_kturbo2_vturbo2 1008 4 11 232 1024
kernel_flash_attn_ext_tile_amd_dk384_kturbo2_vturbo2_w64 945 4 9 232 1024
kernel_flash_attn_ext_tile_amd_dk384_kturbo2_vturbo3 1017 5 11 248 1024
kernel_flash_attn_ext_tile_amd_dk384_kturbo2_vturbo3_w64 952 5 9 248 1024
kernel_flash_attn_ext_tile_amd_dk384_kturbo2_vturbo4 1017 5 11 252 1024
kernel_flash_attn_ext_tile_amd_dk384_kturbo2_vturbo4_w64 951 5 9 252 1024
kernel_flash_attn_ext_tile_amd_dk384_kturbo3_vf16 961 4 11 248 1024
kernel_flash_attn_ext_tile_amd_dk384_kturbo3_vf16_w64 888 4 9 248 1024
kernel_flash_attn_ext_tile_amd_dk384_kturbo3_vq8_0 1022 8 11 248 1024
kernel_flash_attn_ext_tile_amd_dk384_kturbo3_vq8_0_w64 947 8 9 248 1024
kernel_flash_attn_ext_tile_amd_dk384_kturbo3_vturbo2 1017 5 11 248 1024
kernel_flash_attn_ext_tile_amd_dk384_kturbo3_vturbo2_w64 945 5 9 248 1024
kernel_flash_attn_ext_tile_amd_dk384_kturbo3_vturbo3 1059 6 11 248 1024
kernel_flash_attn_ext_tile_amd_dk384_kturbo3_vturbo3_w64 978 6 9 248 1024
kernel_flash_attn_ext_tile_amd_dk384_kturbo3_vturbo4 1041 6 11 252 1024
kernel_flash_attn_ext_tile_amd_dk384_kturbo3_vturbo4_w64 967 6 9 252 1024
kernel_flash_attn_ext_tile_amd_dk384_kturbo4_vf16 954 4 11 252 1024
kernel_flash_attn_ext_tile_amd_dk384_kturbo4_vf16_w64 880 4 9 252 1024
kernel_flash_attn_ext_tile_amd_dk384_kturbo4_vq8_0 1013 8 11 252 1024
kernel_flash_attn_ext_tile_amd_dk384_kturbo4_vq8_0_w64 931 8 9 252 1024
kernel_flash_attn_ext_tile_amd_dk384_kturbo4_vturbo2 1009 5 11 252 1024
kernel_flash_attn_ext_tile_amd_dk384_kturbo4_vturbo2_w64 935 5 9 252 1024
kernel_flash_attn_ext_tile_amd_dk384_kturbo4_vturbo3 1033 6 11 252 1024
kernel_flash_attn_ext_tile_amd_dk384_kturbo4_vturbo3_w64 967 6 9 252 1024
kernel_flash_attn_ext_tile_amd_dk384_kturbo4_vturbo4 1031 6 11 252 1024
kernel_flash_attn_ext_tile_amd_dk384_kturbo4_vturbo4_w64 956 6 9 252 1024
kernel_flash_attn_ext_tile_amd_dk512_kf16_vf16 979 2 13 186 768
kernel_flash_attn_ext_tile_amd_dk512_kf16_vf16_w64 834 2 9 186 1024
kernel_flash_attn_ext_tile_amd_dk512_kf16_vq4_0 1004 3 13 187 768
kernel_flash_attn_ext_tile_amd_dk512_kf16_vq4_0_w64 855 3 9 187 1024
kernel_flash_attn_ext_tile_amd_dk512_kf16_vq8_0 989 6 13 187 768
kernel_flash_attn_ext_tile_amd_dk512_kf16_vq8_0_w64 840 6 9 187 1024
kernel_flash_attn_ext_tile_amd_dk512_kf16_vturbo2 999 3 13 232 768
kernel_flash_attn_ext_tile_amd_dk512_kf16_vturbo2_w64 850 3 9 232 1024
kernel_flash_attn_ext_tile_amd_dk512_kf16_vturbo3 1023 4 13 248 768
kernel_flash_attn_ext_tile_amd_dk512_kf16_vturbo3_w64 874 4 9 248 1024
kernel_flash_attn_ext_tile_amd_dk512_kf16_vturbo4 1025 4 13 252 768
kernel_flash_attn_ext_tile_amd_dk512_kf16_vturbo4_w64 868 4 9 252 1024
kernel_flash_attn_ext_tile_amd_dk512_kq4_0_vf16 998 3 13 186 1024
kernel_flash_attn_ext_tile_amd_dk512_kq4_0_vf16_w64 857 3 9 186 1024
kernel_flash_attn_ext_tile_amd_dk512_kq4_0_vq4_0 1080 4 13 200 768
kernel_flash_attn_ext_tile_amd_dk512_kq4_0_vq4_0_w64 931 4 9 200 1024
kernel_flash_attn_ext_tile_amd_dk512_kq4_0_vq8_0 1064 7 13 236 768
kernel_flash_attn_ext_tile_amd_dk512_kq4_0_vq8_0_w64 916 7 9 251 1024
kernel_flash_attn_ext_tile_amd_dk512_kq8_0_vf16 1005 6 13 186 768
kernel_flash_attn_ext_tile_amd_dk512_kq8_0_vf16_w64 849 6 9 186 1024
kernel_flash_attn_ext_tile_amd_dk512_kq8_0_vq4_0 1071 7 13 200 768
kernel_flash_attn_ext_tile_amd_dk512_kq8_0_vq4_0_w64 923 7 9 246 1024
kernel_flash_attn_ext_tile_amd_dk512_kq8_0_vq8_0 1056 10 13 200 768
kernel_flash_attn_ext_tile_amd_dk512_kq8_0_vq8_0_w64 908 10 9 210 1024
kernel_flash_attn_ext_tile_amd_dk512_kq8_0_vturbo2 1066 7 13 236 768
kernel_flash_attn_ext_tile_amd_dk512_kq8_0_vturbo2_w64 918 7 9 232 1024
kernel_flash_attn_ext_tile_amd_dk512_kq8_0_vturbo3 1082 8 13 248 768
kernel_flash_attn_ext_tile_amd_dk512_kq8_0_vturbo3_w64 934 8 9 248 1024
kernel_flash_attn_ext_tile_amd_dk512_kq8_0_vturbo4 1081 8 13 252 768
kernel_flash_attn_ext_tile_amd_dk512_kq8_0_vturbo4_w64 925 8 9 252 1024
kernel_flash_attn_ext_tile_amd_dk512_kturbo2_vf16 998 3 13 232 768
kernel_flash_attn_ext_tile_amd_dk512_kturbo2_vf16_w64 850 3 9 232 1024
kernel_flash_attn_ext_tile_amd_dk512_kturbo2_vq8_0 1058 7 13 232 768
kernel_flash_attn_ext_tile_amd_dk512_kturbo2_vq8_0_w64 910 7 9 232 1024
kernel_flash_attn_ext_tile_amd_dk512_kturbo2_vturbo2 1073 4 13 232 768
kernel_flash_attn_ext_tile_amd_dk512_kturbo2_vturbo2_w64 924 4 9 232 1024
kernel_flash_attn_ext_tile_amd_dk512_kturbo2_vturbo3 1086 5 13 248 768
kernel_flash_attn_ext_tile_amd_dk512_kturbo2_vturbo3_w64 930 5 9 248 1024
kernel_flash_attn_ext_tile_amd_dk512_kturbo2_vturbo4 1078 5 13 252 768
kernel_flash_attn_ext_tile_amd_dk512_kturbo2_vturbo4_w64 930 5 9 252 1024
kernel_flash_attn_ext_tile_amd_dk512_kturbo3_vf16 1022 4 13 248 768
kernel_flash_attn_ext_tile_amd_dk512_kturbo3_vf16_w64 874 4 9 248 1024
kernel_flash_attn_ext_tile_amd_dk512_kturbo3_vq8_0 1082 8 13 248 768
kernel_flash_attn_ext_tile_amd_dk512_kturbo3_vq8_0_w64 934 8 9 248 1024
kernel_flash_attn_ext_tile_amd_dk512_kturbo3_vturbo2 1086 5 13 248 768
kernel_flash_attn_ext_tile_amd_dk512_kturbo3_vturbo2_w64 938 5 9 248 1024
kernel_flash_attn_ext_tile_amd_dk512_kturbo3_vturbo3 1115 6 13 248 768
kernel_flash_attn_ext_tile_amd_dk512_kturbo3_vturbo3_w64 967 6 9 248 1024
kernel_flash_attn_ext_tile_amd_dk512_kturbo3_vturbo4 1102 6 13 252 768
kernel_flash_attn_ext_tile_amd_dk512_kturbo3_vturbo4_w64 954 6 9 252 1024
kernel_flash_attn_ext_tile_amd_dk512_kturbo4_vf16 1017 4 13 252 768
kernel_flash_attn_ext_tile_amd_dk512_kturbo4_vf16_w64 869 4 9 252 1024
kernel_flash_attn_ext_tile_amd_dk512_kturbo4_vq8_0 1082 8 13 252 768
kernel_flash_attn_ext_tile_amd_dk512_kturbo4_vq8_0_w64 934 8 9 252 1024
kernel_flash_attn_ext_tile_amd_dk512_kturbo4_vturbo2 1086 5 13 252 768
kernel_flash_attn_ext_tile_amd_dk512_kturbo4_vturbo2_w64 938 5 9 252 1024
kernel_flash_attn_ext_tile_amd_dk512_kturbo4_vturbo3 1102 6 13 252 768
kernel_flash_attn_ext_tile_amd_dk512_kturbo4_vturbo3_w64 954 6 9 252 1024
kernel_flash_attn_ext_tile_amd_dk512_kturbo4_vturbo4 1111 6 13 252 768
kernel_flash_attn_ext_tile_amd_dk512_kturbo4_vturbo4_w64 953 6 9 252 1024
kernel_flash_attn_ext_tile_amd_dk576_dv512_kf16_vf16 1044 2 14 200 1024
kernel_flash_attn_ext_tile_amd_dk576_dv512_kf16_vf16_w64 853 2 10 255 1024
kernel_flash_attn_ext_tile_amd_dk576_dv512_kq8_0_vq8_0 1175 10 14 231 1024
kernel_flash_attn_ext_tile_amd_dk576_dv512_kq8_0_vq8_0_w64 980 10 10 187 1024
kernel_flash_attn_ext_tile_amd_dk640_kf16_vturbo2 1146 3 15 232 1024
kernel_flash_attn_ext_tile_amd_dk640_kf16_vturbo2_w64 926 3 11 232 1024
kernel_flash_attn_ext_tile_amd_dk640_kf16_vturbo3 1143 4 15 248 768
kernel_flash_attn_ext_tile_amd_dk640_kf16_vturbo3_w64 951 4 11 248 1024
kernel_flash_attn_ext_tile_amd_dk640_kf16_vturbo4 1106 4 15 252 768
kernel_flash_attn_ext_tile_amd_dk640_kf16_vturbo4_w64 949 4 11 252 1024
kernel_flash_attn_ext_tile_amd_dk640_kq8_0_vturbo2 1209 7 15 232 1024
kernel_flash_attn_ext_tile_amd_dk640_kq8_0_vturbo2_w64 989 7 11 232 1024
kernel_flash_attn_ext_tile_amd_dk640_kq8_0_vturbo3 1174 8 15 248 768
kernel_flash_attn_ext_tile_amd_dk640_kq8_0_vturbo3_w64 1014 8 11 248 1024
kernel_flash_attn_ext_tile_amd_dk640_kq8_0_vturbo4 1222 8 15 252 1024
kernel_flash_attn_ext_tile_amd_dk640_kq8_0_vturbo4_w64 1011 8 11 252 1024
kernel_flash_attn_ext_tile_amd_dk640_kturbo2_vf16 1161 3 15 232 1024
kernel_flash_attn_ext_tile_amd_dk640_kturbo2_vf16_w64 926 3 11 232 1024
kernel_flash_attn_ext_tile_amd_dk640_kturbo2_vq8_0 1209 7 15 232 1024
kernel_flash_attn_ext_tile_amd_dk640_kturbo2_vq8_0_w64 989 7 11 240 1024
kernel_flash_attn_ext_tile_amd_dk640_kturbo2_vturbo2 1224 4 15 232 1024
kernel_flash_attn_ext_tile_amd_dk640_kturbo2_vturbo2_w64 993 4 11 232 1024
kernel_flash_attn_ext_tile_amd_dk640_kturbo2_vturbo3 1171 5 15 248 768
kernel_flash_attn_ext_tile_amd_dk640_kturbo2_vturbo3_w64 1020 5 11 248 1024
kernel_flash_attn_ext_tile_amd_dk640_kturbo2_vturbo4 1229 5 15 252 1024
kernel_flash_attn_ext_tile_amd_dk640_kturbo2_vturbo4_w64 1018 5 11 252 1024
kernel_flash_attn_ext_tile_amd_dk640_kturbo3_vf16 1120 4 15 248 768
kernel_flash_attn_ext_tile_amd_dk640_kturbo3_vf16_w64 951 4 11 248 1024
kernel_flash_attn_ext_tile_amd_dk640_kturbo3_vq8_0 1174 8 15 248 768
kernel_flash_attn_ext_tile_amd_dk640_kturbo3_vq8_0_w64 1014 8 11 248 1024
kernel_flash_attn_ext_tile_amd_dk640_kturbo3_vturbo2 1171 5 15 248 768
kernel_flash_attn_ext_tile_amd_dk640_kturbo3_vturbo2_w64 1020 5 11 248 1024
kernel_flash_attn_ext_tile_amd_dk640_kturbo3_vturbo3 1207 6 15 248 768
kernel_flash_attn_ext_tile_amd_dk640_kturbo3_vturbo3_w64 1047 6 11 248 1024
kernel_flash_attn_ext_tile_amd_dk640_kturbo3_vturbo4 1194 6 15 252 768
kernel_flash_attn_ext_tile_amd_dk640_kturbo3_vturbo4_w64 1043 6 11 252 1024
kernel_flash_attn_ext_tile_amd_dk640_kturbo4_vf16 1106 4 15 252 768
kernel_flash_attn_ext_tile_amd_dk640_kturbo4_vf16_w64 950 4 11 252 1024
kernel_flash_attn_ext_tile_amd_dk640_kturbo4_vq8_0 1223 8 15 252 1024
kernel_flash_attn_ext_tile_amd_dk640_kturbo4_vq8_0_w64 1004 8 11 252 1024
kernel_flash_attn_ext_tile_amd_dk640_kturbo4_vturbo2 1221 5 15 252 1024
kernel_flash_attn_ext_tile_amd_dk640_kturbo4_vturbo2_w64 1010 5 11 252 1024
kernel_flash_attn_ext_tile_amd_dk640_kturbo4_vturbo3 1186 6 15 252 768
kernel_flash_attn_ext_tile_amd_dk640_kturbo4_vturbo3_w64 1035 6 11 252 1024
kernel_flash_attn_ext_tile_amd_dk640_kturbo4_vturbo4 1180 6 15 252 768
kernel_flash_attn_ext_tile_amd_dk640_kturbo4_vturbo4_w64 1024 6 11 252 1024
kernel_flash_attn_ext_tile_amd_dk64_kf16_vf16 734 2 7 208 1024
kernel_flash_attn_ext_tile_amd_dk64_kf16_vf16_w64 735 2 7 208 1024
kernel_flash_attn_ext_tile_amd_dk64_kf16_vq4_0 818 3 7 187 1024
kernel_flash_attn_ext_tile_amd_dk64_kf16_vq4_0_w64 819 3 7 187 1024
kernel_flash_attn_ext_tile_amd_dk64_kf16_vq8_0 801 6 7 187 1024
kernel_flash_attn_ext_tile_amd_dk64_kf16_vq8_0_w64 802 6 7 187 1024
kernel_flash_attn_ext_tile_amd_dk64_kq4_0_vf16 818 3 7 186 1024
kernel_flash_attn_ext_tile_amd_dk64_kq4_0_vf16_w64 819 3 7 186 1024
kernel_flash_attn_ext_tile_amd_dk64_kq4_0_vq4_0 848 4 7 187 1024
kernel_flash_attn_ext_tile_amd_dk64_kq4_0_vq4_0_w64 848 4 7 187 1024
kernel_flash_attn_ext_tile_amd_dk64_kq4_0_vq8_0 833 7 7 187 1024
kernel_flash_attn_ext_tile_amd_dk64_kq4_0_vq8_0_w64 833 7 7 187 1024
kernel_flash_attn_ext_tile_amd_dk64_kq8_0_vf16 808 6 7 186 1024
kernel_flash_attn_ext_tile_amd_dk64_kq8_0_vf16_w64 809 6 7 186 1024
kernel_flash_attn_ext_tile_amd_dk64_kq8_0_vq4_0 840 7 7 187 1024
kernel_flash_attn_ext_tile_amd_dk64_kq8_0_vq4_0_w64 840 7 7 187 1024
kernel_flash_attn_ext_tile_amd_dk64_kq8_0_vq8_0 825 10 7 187 1024
kernel_flash_attn_ext_tile_amd_dk64_kq8_0_vq8_0_w64 825 10 7 187 1024
kernel_flash_attn_ext_tile_amd_dk72_kf16_vf16 737 2 7 220 1024
kernel_flash_attn_ext_tile_amd_dk72_kf16_vf16_w64 738 2 7 220 1024
kernel_flash_attn_ext_vec_amd_dk128_kf16_vf16 885 1 12 187 576
kernel_flash_attn_ext_vec_amd_dk128_kf16_vf16_w64 838 1 8 212 576
kernel_flash_attn_ext_vec_amd_dk128_kf16_vq4_0 1005 5 12 187 640
kernel_flash_attn_ext_vec_amd_dk128_kf16_vq4_0_w64 960 5 8 187 640
kernel_flash_attn_ext_vec_amd_dk128_kf16_vq8_0 970 17 12 187 576
kernel_flash_attn_ext_vec_amd_dk128_kf16_vq8_0_w64 922 17 8 187 576
kernel_flash_attn_ext_vec_amd_dk128_kf16_vturbo2 1021 5 12 232 576
kernel_flash_attn_ext_vec_amd_dk128_kf16_vturbo2_w64 977 5 8 232 576
kernel_flash_attn_ext_vec_amd_dk128_kf16_vturbo2_wg32 1050 5 12 232 576
kernel_flash_attn_ext_vec_amd_dk128_kf16_vturbo3 1066 9 12 248 576
kernel_flash_attn_ext_vec_amd_dk128_kf16_vturbo3_w64 1036 9 8 248 512
kernel_flash_attn_ext_vec_amd_dk128_kf16_vturbo3_wg32 1089 9 12 248 512
kernel_flash_attn_ext_vec_amd_dk128_kf16_vturbo4 1052 9 12 252 576
kernel_flash_attn_ext_vec_amd_dk128_kf16_vturbo4_w64 1013 9 8 252 576
kernel_flash_attn_ext_vec_amd_dk128_kf16_vturbo4_wg32 1054 9 12 252 576
kernel_flash_attn_ext_vec_amd_dk128_kq4_0_vf16 890 3 10 187 768
kernel_flash_attn_ext_vec_amd_dk128_kq4_0_vf16_w64 953 5 8 187 576
kernel_flash_attn_ext_vec_amd_dk128_kq4_0_vq4_0 911 5 10 187 768
kernel_flash_attn_ext_vec_amd_dk128_kq4_0_vq4_0_w64 983 9 8 187 640
kernel_flash_attn_ext_vec_amd_dk128_kq4_0_vq8_0 918 11 10 187 768
kernel_flash_attn_ext_vec_amd_dk128_kq4_0_vq8_0_w64 961 21 8 187 576
kernel_flash_attn_ext_vec_amd_dk128_kq8_0_vf16 951 17 12 187 576
kernel_flash_attn_ext_vec_amd_dk128_kq8_0_vf16_w64 923 17 8 187 576
kernel_flash_attn_ext_vec_amd_dk128_kq8_0_vq4_0 1013 21 12 187 640
kernel_flash_attn_ext_vec_amd_dk128_kq8_0_vq4_0_w64 963 21 8 187 576
kernel_flash_attn_ext_vec_amd_dk128_kq8_0_vq8_0 992 33 12 187 576
kernel_flash_attn_ext_vec_amd_dk128_kq8_0_vq8_0_w64 944 33 8 187 576
kernel_flash_attn_ext_vec_amd_dk128_kq8_0_vturbo2 1056 21 12 232 576
kernel_flash_attn_ext_vec_amd_dk128_kq8_0_vturbo2_w64 1006 21 8 232 576
kernel_flash_attn_ext_vec_amd_dk128_kq8_0_vturbo2_wg32 1080 21 12 232 576
kernel_flash_attn_ext_vec_amd_dk128_kq8_0_vturbo3 1096 25 12 248 512
kernel_flash_attn_ext_vec_amd_dk128_kq8_0_vturbo3_w64 1053 25 8 248 512
kernel_flash_attn_ext_vec_amd_dk128_kq8_0_vturbo3_wg32 1144 25 12 248 512
kernel_flash_attn_ext_vec_amd_dk128_kq8_0_vturbo4 1068 25 12 252 576
kernel_flash_attn_ext_vec_amd_dk128_kq8_0_vturbo4_w64 1028 25 8 252 576
kernel_flash_attn_ext_vec_amd_dk128_kq8_0_vturbo4_wg32 1119 25 12 252 576
kernel_flash_attn_ext_vec_amd_dk128_kturbo2_vf16 1030 5 12 232 576
kernel_flash_attn_ext_vec_amd_dk128_kturbo2_vf16_w64 983 5 8 232 576
kernel_flash_attn_ext_vec_amd_dk128_kturbo2_vf16_wg32 1033 5 12 232 576
kernel_flash_attn_ext_vec_amd_dk128_kturbo2_vq8_0 1053 21 12 232 576
kernel_flash_attn_ext_vec_amd_dk128_kturbo2_vq8_0_w64 1015 21 8 232 576
kernel_flash_attn_ext_vec_amd_dk128_kturbo2_vq8_0_wg32 1097 21 12 232 576
kernel_flash_attn_ext_vec_amd_dk128_kturbo2_vturbo2 1108 9 12 240 576
kernel_flash_attn_ext_vec_amd_dk128_kturbo2_vturbo2_w64 1063 9 8 232 576
kernel_flash_attn_ext_vec_amd_dk128_kturbo2_vturbo2_wg32 1157 9 12 232 576
kernel_flash_attn_ext_vec_amd_dk128_kturbo2_vturbo3 1164 13 12 248 576
kernel_flash_attn_ext_vec_amd_dk128_kturbo2_vturbo3_w64 1123 13 8 248 512
kernel_flash_attn_ext_vec_amd_dk128_kturbo2_vturbo3_wg32 1210 13 12 248 576
kernel_flash_attn_ext_vec_amd_dk128_kturbo2_vturbo4 1135 13 12 252 576
kernel_flash_attn_ext_vec_amd_dk128_kturbo2_vturbo4_w64 1097 13 8 252 576
kernel_flash_attn_ext_vec_amd_dk128_kturbo2_vturbo4_wg32 1188 13 12 252 576
kernel_flash_attn_ext_vec_amd_dk128_kturbo3_vf16 1067 9 12 248 576
kernel_flash_attn_ext_vec_amd_dk128_kturbo3_vf16_w64 1039 9 8 248 512
kernel_flash_attn_ext_vec_amd_dk128_kturbo3_vf16_wg32 1098 9 12 248 512
kernel_flash_attn_ext_vec_amd_dk128_kturbo3_vq8_0 1102 25 12 248 512
kernel_flash_attn_ext_vec_amd_dk128_kturbo3_vq8_0_w64 1063 25 8 248 512
kernel_flash_attn_ext_vec_amd_dk128_kturbo3_vq8_0_wg32 1162 25 12 248 512
kernel_flash_attn_ext_vec_amd_dk128_kturbo3_vturbo2 1141 13 12 248 576
kernel_flash_attn_ext_vec_amd_dk128_kturbo3_vturbo2_w64 1126 13 8 248 576
kernel_flash_attn_ext_vec_amd_dk128_kturbo3_vturbo2_wg32 1202 13 12 248 576
kernel_flash_attn_ext_vec_amd_dk128_kturbo3_vturbo3 1203 17 12 248 576
kernel_flash_attn_ext_vec_amd_dk128_kturbo3_vturbo3_w64 1155 17 8 248 512
kernel_flash_attn_ext_vec_amd_dk128_kturbo3_vturbo3_wg32 1270 17 12 248 576
kernel_flash_attn_ext_vec_amd_dk128_kturbo3_vturbo4 1185 17 12 252 512
kernel_flash_attn_ext_vec_amd_dk128_kturbo3_vturbo4_w64 1150 17 8 252 512
kernel_flash_attn_ext_vec_amd_dk128_kturbo3_vturbo4_wg32 1239 17 12 252 512
kernel_flash_attn_ext_vec_amd_dk128_kturbo4_vf16 1052 9 12 252 640
kernel_flash_attn_ext_vec_amd_dk128_kturbo4_vf16_w64 1019 9 8 252 576
kernel_flash_attn_ext_vec_amd_dk128_kturbo4_vf16_wg32 1064 9 12 252 576
kernel_flash_attn_ext_vec_amd_dk128_kturbo4_vq8_0 1086 25 12 252 576
kernel_flash_attn_ext_vec_amd_dk128_kturbo4_vq8_0_w64 1039 25 8 252 576
kernel_flash_attn_ext_vec_amd_dk128_kturbo4_vq8_0_wg32 1113 25 12 252 576
kernel_flash_attn_ext_vec_amd_dk128_kturbo4_vturbo2 1136 13 12 252 576
kernel_flash_attn_ext_vec_amd_dk128_kturbo4_vturbo2_w64 1090 13 8 252 576
kernel_flash_attn_ext_vec_amd_dk128_kturbo4_vturbo2_wg32 1180 13 12 252 576
kernel_flash_attn_ext_vec_amd_dk128_kturbo4_vturbo3 1198 17 12 252 512
kernel_flash_attn_ext_vec_amd_dk128_kturbo4_vturbo3_w64 1161 17 8 252 512
kernel_flash_attn_ext_vec_amd_dk128_kturbo4_vturbo3_wg32 1234 17 12 252 512
kernel_flash_attn_ext_vec_amd_dk128_kturbo4_vturbo4 1136 17 12 252 576
kernel_flash_attn_ext_vec_amd_dk128_kturbo4_vturbo4_w64 1090 17 8 252 576
kernel_flash_attn_ext_vec_amd_dk128_kturbo4_vturbo4_wg32 1201 17 12 252 576
kernel_flash_attn_ext_vec_amd_dk256_kf16_vf16 847 1 8 244 576
kernel_flash_attn_ext_vec_amd_dk256_kf16_vf16_w64 848 1 8 244 576
kernel_flash_attn_ext_vec_amd_dk256_kf16_vf16_w64_wg32 912 1 8 187 576
kernel_flash_attn_ext_vec_amd_dk256_kf16_vf16_wg32 912 1 8 187 576
kernel_flash_attn_ext_vec_amd_dk256_kf16_vq4_0 997 5 8 187 576
kernel_flash_attn_ext_vec_amd_dk256_kf16_vq4_0_w64 998 5 8 187 576
kernel_flash_attn_ext_vec_amd_dk256_kf16_vq4_0_w64_wg32 1000 5 8 187 640
kernel_flash_attn_ext_vec_amd_dk256_kf16_vq4_0_wg32 999 5 8 187 640
kernel_flash_attn_ext_vec_amd_dk256_kf16_vq8_0 954 17 8 187 576
kernel_flash_attn_ext_vec_amd_dk256_kf16_vq8_0_w64 955 17 8 187 576
kernel_flash_attn_ext_vec_amd_dk256_kf16_vq8_0_w64_wg32 953 17 8 187 576
kernel_flash_attn_ext_vec_amd_dk256_kf16_vq8_0_wg32 952 17 8 187 576
kernel_flash_attn_ext_vec_amd_dk256_kf16_vturbo2 972 5 8 232 576
kernel_flash_attn_ext_vec_amd_dk256_kf16_vturbo2_w64 973 5 8 232 576
kernel_flash_attn_ext_vec_amd_dk256_kf16_vturbo2_w64_wg32 988 5 8 232 576
kernel_flash_attn_ext_vec_amd_dk256_kf16_vturbo2_wg32 987 5 8 232 576
kernel_flash_attn_ext_vec_amd_dk256_kf16_vturbo3 1032 9 8 248 576
kernel_flash_attn_ext_vec_amd_dk256_kf16_vturbo3_w64 1033 9 8 248 576
kernel_flash_attn_ext_vec_amd_dk256_kf16_vturbo3_w64_wg32 1034 9 8 248 576
kernel_flash_attn_ext_vec_amd_dk256_kf16_vturbo3_wg32 1033 9 8 248 576
kernel_flash_attn_ext_vec_amd_dk256_kf16_vturbo4 1015 9 8 252 576
kernel_flash_attn_ext_vec_amd_dk256_kf16_vturbo4_w64 1016 9 8 252 576
kernel_flash_attn_ext_vec_amd_dk256_kf16_vturbo4_w64_wg32 1008 9 8 252 576
kernel_flash_attn_ext_vec_amd_dk256_kf16_vturbo4_wg32 1007 9 8 252 576
kernel_flash_attn_ext_vec_amd_dk256_kq4_0_vf16 998 5 8 187 576
kernel_flash_attn_ext_vec_amd_dk256_kq4_0_vf16_w64 999 5 8 187 576
kernel_flash_attn_ext_vec_amd_dk256_kq4_0_vf16_w64_wg32 1005 5 8 187 576
kernel_flash_attn_ext_vec_amd_dk256_kq4_0_vf16_wg32 1004 5 8 187 576
kernel_flash_attn_ext_vec_amd_dk256_kq4_0_vq4_0 1052 9 8 187 576
kernel_flash_attn_ext_vec_amd_dk256_kq4_0_vq4_0_w64 1052 9 8 187 576
kernel_flash_attn_ext_vec_amd_dk256_kq4_0_vq4_0_w64_wg32 1096 9 8 200 576
kernel_flash_attn_ext_vec_amd_dk256_kq4_0_vq4_0_wg32 1095 9 8 200 576
kernel_flash_attn_ext_vec_amd_dk256_kq4_0_vq8_0 1059 21 8 187 512
kernel_flash_attn_ext_vec_amd_dk256_kq4_0_vq8_0_w64 1059 21 8 187 512
kernel_flash_attn_ext_vec_amd_dk256_kq4_0_vq8_0_w64_wg32 1094 21 8 200 512
kernel_flash_attn_ext_vec_amd_dk256_kq4_0_vq8_0_wg32 1093 21 8 200 512
kernel_flash_attn_ext_vec_amd_dk256_kq8_0_vf16 951 17 8 187 576
kernel_flash_attn_ext_vec_amd_dk256_kq8_0_vf16_w64 952 17 8 187 576
kernel_flash_attn_ext_vec_amd_dk256_kq8_0_vf16_w64_wg32 952 17 8 187 576
kernel_flash_attn_ext_vec_amd_dk256_kq8_0_vf16_wg32 951 17 8 187 576
kernel_flash_attn_ext_vec_amd_dk256_kq8_0_vq4_0 1053 21 8 187 576
kernel_flash_attn_ext_vec_amd_dk256_kq8_0_vq4_0_w64 1053 21 8 187 576
kernel_flash_attn_ext_vec_amd_dk256_kq8_0_vq4_0_w64_wg32 1089 21 8 200 576
kernel_flash_attn_ext_vec_amd_dk256_kq8_0_vq4_0_wg32 1088 21 8 200 576
kernel_flash_attn_ext_vec_amd_dk256_kq8_0_vq8_0 973 33 8 187 576
kernel_flash_attn_ext_vec_amd_dk256_kq8_0_vq8_0_w64 973 33 8 187 576
kernel_flash_attn_ext_vec_amd_dk256_kq8_0_vq8_0_w64_wg32 1019 33 8 200 576
kernel_flash_attn_ext_vec_amd_dk256_kq8_0_vq8_0_wg32 1018 33 8 200 576
kernel_flash_attn_ext_vec_amd_dk256_kq8_0_vturbo2 1031 21 8 232 576
kernel_flash_attn_ext_vec_amd_dk256_kq8_0_vturbo2_w64 1031 21 8 232 576
kernel_flash_attn_ext_vec_amd_dk256_kq8_0_vturbo2_w64_wg32 1070 21 8 232 576
kernel_flash_attn_ext_vec_amd_dk256_kq8_0_vturbo2_wg32 1069 21 8 232 576
kernel_flash_attn_ext_vec_amd_dk256_kq8_0_vturbo3 1068 25 8 248 512
kernel_flash_attn_ext_vec_amd_dk256_kq8_0_vturbo3_w64 1068 25 8 248 512
kernel_flash_attn_ext_vec_amd_dk256_kq8_0_vturbo3_w64_wg32 1107 25 8 248 512
kernel_flash_attn_ext_vec_amd_dk256_kq8_0_vturbo3_wg32 1106 25 8 248 512
kernel_flash_attn_ext_vec_amd_dk256_kq8_0_vturbo4 1036 25 8 252 576
kernel_flash_attn_ext_vec_amd_dk256_kq8_0_vturbo4_w64 1036 25 8 252 576
kernel_flash_attn_ext_vec_amd_dk256_kq8_0_vturbo4_w64_wg32 1090 25 8 252 576
kernel_flash_attn_ext_vec_amd_dk256_kq8_0_vturbo4_wg32 1089 25 8 252 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo2_vf16 988 5 8 232 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo2_vf16_w64 989 5 8 232 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo2_vf16_w64_wg32 996 5 8 232 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo2_vf16_wg32 995 5 8 232 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo2_vq8_0 1031 21 8 232 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo2_vq8_0_w64 1032 21 8 232 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo2_vq8_0_w64_wg32 1079 21 8 232 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo2_vq8_0_wg32 1078 21 8 232 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo2_vturbo2 1055 9 8 232 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo2_vturbo2_w64 1056 9 8 232 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo2_vturbo2_w64_wg32 1119 9 8 232 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo2_vturbo2_wg32 1118 9 8 232 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo2_vturbo3 1103 13 8 248 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo2_vturbo3_w64 1104 13 8 248 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo2_vturbo3_w64_wg32 1168 13 8 248 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo2_vturbo3_wg32 1167 13 8 248 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo2_vturbo4 1081 13 8 252 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo2_vturbo4_w64 1082 13 8 252 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo2_vturbo4_w64_wg32 1133 13 8 252 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo2_vturbo4_wg32 1132 13 8 252 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo3_vf16 1037 9 8 248 512
kernel_flash_attn_ext_vec_amd_dk256_kturbo3_vf16_w64 1038 9 8 248 512
kernel_flash_attn_ext_vec_amd_dk256_kturbo3_vf16_w64_wg32 1035 9 8 248 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo3_vf16_wg32 1034 9 8 248 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo3_vq8_0 1077 25 8 248 512
kernel_flash_attn_ext_vec_amd_dk256_kturbo3_vq8_0_w64 1078 25 8 248 512
kernel_flash_attn_ext_vec_amd_dk256_kturbo3_vq8_0_w64_wg32 1125 25 8 248 512
kernel_flash_attn_ext_vec_amd_dk256_kturbo3_vq8_0_wg32 1124 25 8 248 512
kernel_flash_attn_ext_vec_amd_dk256_kturbo3_vturbo2 1096 13 8 248 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo3_vturbo2_w64 1097 13 8 248 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo3_vturbo2_w64_wg32 1161 13 8 248 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo3_vturbo2_wg32 1160 13 8 248 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo3_vturbo3 1146 17 8 248 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo3_vturbo3_w64 1147 17 8 248 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo3_vturbo3_w64_wg32 1198 17 8 248 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo3_vturbo3_wg32 1197 17 8 248 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo3_vturbo4 1130 17 8 252 512
kernel_flash_attn_ext_vec_amd_dk256_kturbo3_vturbo4_w64 1131 17 8 252 512
kernel_flash_attn_ext_vec_amd_dk256_kturbo3_vturbo4_w64_wg32 1176 17 8 252 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo3_vturbo4_wg32 1175 17 8 252 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo4_vf16 1010 9 8 252 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo4_vf16_w64 1011 9 8 252 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo4_vf16_w64_wg32 1018 9 8 252 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo4_vf16_wg32 1017 9 8 252 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo4_vq8_0 1047 25 8 252 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo4_vq8_0_w64 1048 25 8 252 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo4_vq8_0_w64_wg32 1086 25 8 252 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo4_vq8_0_wg32 1085 25 8 252 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo4_vturbo2 1082 13 8 252 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo4_vturbo2_w64 1083 13 8 252 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo4_vturbo2_w64_wg32 1130 13 8 252 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo4_vturbo2_wg32 1129 13 8 252 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo4_vturbo3 1122 17 8 252 512
kernel_flash_attn_ext_vec_amd_dk256_kturbo4_vturbo3_w64 1123 17 8 252 512
kernel_flash_attn_ext_vec_amd_dk256_kturbo4_vturbo3_w64_wg32 1170 17 8 252 512
kernel_flash_attn_ext_vec_amd_dk256_kturbo4_vturbo3_wg32 1169 17 8 252 512
kernel_flash_attn_ext_vec_amd_dk256_kturbo4_vturbo4 1089 17 8 252 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo4_vturbo4_w64 1090 17 8 252 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo4_vturbo4_w64_wg32 1137 17 8 252 576
kernel_flash_attn_ext_vec_amd_dk256_kturbo4_vturbo4_wg32 1137 17 8 252 576
kernel_flash_attn_ext_vec_amd_dk384_kf16_vturbo2 914 4 7 232 768
kernel_flash_attn_ext_vec_amd_dk384_kf16_vturbo2_w64 950 4 7 232 640
kernel_flash_attn_ext_vec_amd_dk384_kf16_vturbo3 967 7 7 248 640
kernel_flash_attn_ext_vec_amd_dk384_kf16_vturbo3_w64 978 7 7 248 640
kernel_flash_attn_ext_vec_amd_dk384_kf16_vturbo4 925 7 7 252 640
kernel_flash_attn_ext_vec_amd_dk384_kf16_vturbo4_w64 946 7 7 252 640
kernel_flash_attn_ext_vec_amd_dk384_kq8_0_vturbo2 949 16 7 232 640
kernel_flash_attn_ext_vec_amd_dk384_kq8_0_vturbo2_w64 955 16 7 232 640
kernel_flash_attn_ext_vec_amd_dk384_kq8_0_vturbo3 971 19 7 248 640
kernel_flash_attn_ext_vec_amd_dk384_kq8_0_vturbo3_w64 988 19 7 248 640
kernel_flash_attn_ext_vec_amd_dk384_kq8_0_vturbo4 942 19 7 252 640
kernel_flash_attn_ext_vec_amd_dk384_kq8_0_vturbo4_w64 980 19 7 252 640
kernel_flash_attn_ext_vec_amd_dk384_kturbo2_vf16 916 4 7 232 640
kernel_flash_attn_ext_vec_amd_dk384_kturbo2_vf16_w64 943 4 7 232 640
kernel_flash_attn_ext_vec_amd_dk384_kturbo2_vq8_0 948 16 7 232 640
kernel_flash_attn_ext_vec_amd_dk384_kturbo2_vq8_0_w64 973 16 7 232 640
kernel_flash_attn_ext_vec_amd_dk384_kturbo2_vturbo2 966 7 7 232 768
kernel_flash_attn_ext_vec_amd_dk384_kturbo2_vturbo2_w64 1001 7 7 232 640
kernel_flash_attn_ext_vec_amd_dk384_kturbo2_vturbo3 1023 10 7 248 640
kernel_flash_attn_ext_vec_amd_dk384_kturbo2_vturbo3_w64 1031 10 7 248 640
kernel_flash_attn_ext_vec_amd_dk384_kturbo2_vturbo4 977 10 7 252 640
kernel_flash_attn_ext_vec_amd_dk384_kturbo2_vturbo4_w64 1017 10 7 252 640
kernel_flash_attn_ext_vec_amd_dk384_kturbo3_vf16 967 7 7 248 640
kernel_flash_attn_ext_vec_amd_dk384_kturbo3_vf16_w64 988 7 7 248 640
kernel_flash_attn_ext_vec_amd_dk384_kturbo3_vq8_0 981 19 7 248 640
kernel_flash_attn_ext_vec_amd_dk384_kturbo3_vq8_0_w64 998 19 7 248 640
kernel_flash_attn_ext_vec_amd_dk384_kturbo3_vturbo2 1030 10 7 248 640
kernel_flash_attn_ext_vec_amd_dk384_kturbo3_vturbo2_w64 1048 10 7 248 640
kernel_flash_attn_ext_vec_amd_dk384_kturbo3_vturbo3 1049 13 7 248 640
kernel_flash_attn_ext_vec_amd_dk384_kturbo3_vturbo3_w64 1062 13 7 248 640
kernel_flash_attn_ext_vec_amd_dk384_kturbo3_vturbo4 1025 13 7 252 640
kernel_flash_attn_ext_vec_amd_dk384_kturbo3_vturbo4_w64 1056 13 7 252 640
kernel_flash_attn_ext_vec_amd_dk384_kturbo4_vf16 929 7 7 252 640
kernel_flash_attn_ext_vec_amd_dk384_kturbo4_vf16_w64 956 7 7 252 640
kernel_flash_attn_ext_vec_amd_dk384_kturbo4_vq8_0 949 19 7 252 640
kernel_flash_attn_ext_vec_amd_dk384_kturbo4_vq8_0_w64 976 19 7 252 640
kernel_flash_attn_ext_vec_amd_dk384_kturbo4_vturbo2 983 10 7 252 640
kernel_flash_attn_ext_vec_amd_dk384_kturbo4_vturbo2_w64 1017 10 7 252 640
kernel_flash_attn_ext_vec_amd_dk384_kturbo4_vturbo3 1034 13 7 252 640
kernel_flash_attn_ext_vec_amd_dk384_kturbo4_vturbo3_w64 1049 13 7 252 576
kernel_flash_attn_ext_vec_amd_dk384_kturbo4_vturbo4 974 13 7 252 768
kernel_flash_attn_ext_vec_amd_dk384_kturbo4_vturbo4_w64 998 13 7 252 640
kernel_flash_attn_ext_vec_amd_dk512_kf16_vf16 843 1 8 230 576
kernel_flash_attn_ext_vec_amd_dk512_kf16_vf16_w64 845 1 8 246 640
kernel_flash_attn_ext_vec_amd_dk512_kf16_vf16_w64_wg32 914 1 8 187 640
kernel_flash_attn_ext_vec_amd_dk512_kf16_vf16_wg32 895 1 8 187 576
kernel_flash_attn_ext_vec_amd_dk512_kf16_vq4_0 988 5 8 187 576
kernel_flash_attn_ext_vec_amd_dk512_kf16_vq4_0_w64 999 5 8 187 576
kernel_flash_attn_ext_vec_amd_dk512_kf16_vq4_0_w64_wg32 1009 5 8 187 640
kernel_flash_attn_ext_vec_amd_dk512_kf16_vq4_0_wg32 997 5 8 187 576
kernel_flash_attn_ext_vec_amd_dk512_kf16_vq8_0 928 17 8 187 576
kernel_flash_attn_ext_vec_amd_dk512_kf16_vq8_0_w64 951 17 8 187 576
kernel_flash_attn_ext_vec_amd_dk512_kf16_vq8_0_w64_wg32 962 17 8 187 576
kernel_flash_attn_ext_vec_amd_dk512_kf16_vq8_0_wg32 951 17 8 187 576
kernel_flash_attn_ext_vec_amd_dk512_kf16_vturbo2 981 5 8 232 576
kernel_flash_attn_ext_vec_amd_dk512_kf16_vturbo2_w64 1012 5 8 232 576
kernel_flash_attn_ext_vec_amd_dk512_kf16_vturbo2_w64_wg32 1009 5 8 232 576
kernel_flash_attn_ext_vec_amd_dk512_kf16_vturbo2_wg32 992 5 8 232 576
kernel_flash_attn_ext_vec_amd_dk512_kf16_vturbo3 1038 9 8 248 576
kernel_flash_attn_ext_vec_amd_dk512_kf16_vturbo3_w64 1052 9 8 248 576
kernel_flash_attn_ext_vec_amd_dk512_kf16_vturbo3_w64_wg32 1063 9 8 248 576
kernel_flash_attn_ext_vec_amd_dk512_kf16_vturbo3_wg32 1036 9 8 248 576
kernel_flash_attn_ext_vec_amd_dk512_kf16_vturbo4 972 9 8 252 576
kernel_flash_attn_ext_vec_amd_dk512_kf16_vturbo4_w64 1007 9 8 252 576
kernel_flash_attn_ext_vec_amd_dk512_kf16_vturbo4_w64_wg32 1012 9 8 252 576
kernel_flash_attn_ext_vec_amd_dk512_kf16_vturbo4_wg32 993 9 8 252 576
kernel_flash_attn_ext_vec_amd_dk512_kq4_0_vf16 996 5 8 187 576
kernel_flash_attn_ext_vec_amd_dk512_kq4_0_vf16_w64 1009 5 8 187 576
kernel_flash_attn_ext_vec_amd_dk512_kq4_0_vf16_w64_wg32 1014 5 8 187 576
kernel_flash_attn_ext_vec_amd_dk512_kq4_0_vf16_wg32 995 5 8 187 576
kernel_flash_attn_ext_vec_amd_dk512_kq4_0_vq4_0 1050 9 8 187 576
kernel_flash_attn_ext_vec_amd_dk512_kq4_0_vq4_0_w64 1062 9 8 187 576
kernel_flash_attn_ext_vec_amd_dk512_kq4_0_vq4_0_w64_wg32 1104 9 8 200 576
kernel_flash_attn_ext_vec_amd_dk512_kq4_0_vq4_0_wg32 1084 9 8 200 576
kernel_flash_attn_ext_vec_amd_dk512_kq4_0_vq8_0 1043 21 8 187 512
kernel_flash_attn_ext_vec_amd_dk512_kq4_0_vq8_0_w64 1061 21 8 187 512
kernel_flash_attn_ext_vec_amd_dk512_kq4_0_vq8_0_w64_wg32 1103 21 8 200 512
kernel_flash_attn_ext_vec_amd_dk512_kq4_0_vq8_0_wg32 1094 21 8 200 512
kernel_flash_attn_ext_vec_amd_dk512_kq8_0_vf16 926 17 8 187 576
kernel_flash_attn_ext_vec_amd_dk512_kq8_0_vf16_w64 950 17 8 187 576
kernel_flash_attn_ext_vec_amd_dk512_kq8_0_vf16_w64_wg32 961 17 8 187 576
kernel_flash_attn_ext_vec_amd_dk512_kq8_0_vf16_wg32 943 17 8 187 576
kernel_flash_attn_ext_vec_amd_dk512_kq8_0_vq4_0 1045 21 8 187 576
kernel_flash_attn_ext_vec_amd_dk512_kq8_0_vq4_0_w64 1055 21 8 187 576
kernel_flash_attn_ext_vec_amd_dk512_kq8_0_vq4_0_w64_wg32 1097 21 8 200 576
kernel_flash_attn_ext_vec_amd_dk512_kq8_0_vq4_0_wg32 1088 21 8 200 576
kernel_flash_attn_ext_vec_amd_dk512_kq8_0_vq8_0 966 33 8 187 576
kernel_flash_attn_ext_vec_amd_dk512_kq8_0_vq8_0_w64 983 33 8 187 576
kernel_flash_attn_ext_vec_amd_dk512_kq8_0_vq8_0_w64_wg32 1036 33 8 200 576
kernel_flash_attn_ext_vec_amd_dk512_kq8_0_vq8_0_wg32 1001 33 8 200 576
kernel_flash_attn_ext_vec_amd_dk512_kq8_0_vturbo2 1021 21 8 232 576
kernel_flash_attn_ext_vec_amd_dk512_kq8_0_vturbo2_w64 1037 21 8 232 576
kernel_flash_attn_ext_vec_amd_dk512_kq8_0_vturbo2_w64_wg32 1090 21 8 232 576
kernel_flash_attn_ext_vec_amd_dk512_kq8_0_vturbo2_wg32 1057 21 8 232 576
kernel_flash_attn_ext_vec_amd_dk512_kq8_0_vturbo3 1065 25 8 248 512
kernel_flash_attn_ext_vec_amd_dk512_kq8_0_vturbo3_w64 1075 25 8 248 512
kernel_flash_attn_ext_vec_amd_dk512_kq8_0_vturbo3_w64_wg32 1136 25 8 248 576
kernel_flash_attn_ext_vec_amd_dk512_kq8_0_vturbo3_wg32 1109 25 8 248 576
kernel_flash_attn_ext_vec_amd_dk512_kq8_0_vturbo4 1031 25 8 252 576
kernel_flash_attn_ext_vec_amd_dk512_kq8_0_vturbo4_w64 1045 25 8 252 576
kernel_flash_attn_ext_vec_amd_dk512_kq8_0_vturbo4_w64_wg32 1084 25 8 252 576
kernel_flash_attn_ext_vec_amd_dk512_kq8_0_vturbo4_wg32 1065 25 8 252 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo2_vf16 984 5 8 232 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo2_vf16_w64 1013 5 8 232 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo2_vf16_w64_wg32 1009 5 8 232 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo2_vf16_wg32 993 5 8 232 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo2_vq8_0 1028 21 8 232 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo2_vq8_0_w64 1046 21 8 232 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo2_vq8_0_w64_wg32 1100 21 8 232 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo2_vq8_0_wg32 1059 21 8 232 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo2_vturbo2 1075 9 8 232 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo2_vturbo2_w64 1082 9 8 232 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo2_vturbo2_w64_wg32 1152 9 8 232 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo2_vturbo2_wg32 1130 9 8 232 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo2_vturbo3 1116 13 8 248 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo2_vturbo3_w64 1142 13 8 248 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo2_vturbo3_w64_wg32 1189 13 8 248 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo2_vturbo3_wg32 1155 13 8 248 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo2_vturbo4 1065 13 8 252 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo2_vturbo4_w64 1090 13 8 252 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo2_vturbo4_w64_wg32 1137 13 8 252 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo2_vturbo4_wg32 1126 13 8 252 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo3_vf16 1035 9 8 248 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo3_vf16_w64 1049 9 8 248 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo3_vf16_w64_wg32 1064 9 8 248 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo3_vf16_wg32 1045 9 8 248 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo3_vq8_0 1072 25 8 248 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo3_vq8_0_w64 1093 25 8 248 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo3_vq8_0_w64_wg32 1147 25 8 248 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo3_vq8_0_wg32 1118 25 8 248 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo3_vturbo2 1125 13 8 248 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo3_vturbo2_w64 1143 13 8 248 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo3_vturbo2_w64_wg32 1189 13 8 248 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo3_vturbo2_wg32 1171 13 8 248 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo3_vturbo3 1154 17 8 248 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo3_vturbo3_w64 1174 17 8 248 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo3_vturbo3_w64_wg32 1242 17 8 248 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo3_vturbo3_wg32 1220 17 8 248 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo3_vturbo4 1120 17 8 252 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo3_vturbo4_w64 1152 17 8 252 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo3_vturbo4_w64_wg32 1191 17 8 252 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo3_vturbo4_wg32 1164 17 8 252 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo4_vf16 970 9 8 252 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo4_vf16_w64 1014 9 8 252 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo4_vf16_w64_wg32 1015 9 8 252 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo4_vf16_wg32 999 9 8 252 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo4_vq8_0 1028 25 8 252 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo4_vq8_0_w64 1053 25 8 252 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo4_vq8_0_w64_wg32 1098 25 8 252 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo4_vq8_0_wg32 1074 25 8 252 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo4_vturbo2 1070 13 8 252 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo4_vturbo2_w64 1088 13 8 252 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo4_vturbo2_w64_wg32 1141 13 8 252 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo4_vturbo2_wg32 1132 13 8 252 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo4_vturbo3 1116 17 8 252 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo4_vturbo3_w64 1144 17 8 252 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo4_vturbo3_w64_wg32 1190 17 8 252 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo4_vturbo3_wg32 1169 17 8 252 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo4_vturbo4 1054 17 8 252 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo4_vturbo4_w64 1067 17 8 252 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo4_vturbo4_w64_wg32 1133 17 8 252 576
kernel_flash_attn_ext_vec_amd_dk512_kturbo4_vturbo4_wg32 1105 17 8 252 576
kernel_flash_attn_ext_vec_amd_dk576_dv512_kf16_vf16 880 1 8 187 640
kernel_flash_attn_ext_vec_amd_dk576_dv512_kf16_vf16_w64 776 1 6 187 1024
kernel_flash_attn_ext_vec_amd_dk576_dv512_kf16_vf16_w64_wg32 812 1 6 200 1024
kernel_flash_attn_ext_vec_amd_dk576_dv512_kf16_vf16_wg32 902 1 8 200 640
kernel_flash_attn_ext_vec_amd_dk576_dv512_kq8_0_vq8_0 1030 37 8 230 640
kernel_flash_attn_ext_vec_amd_dk576_dv512_kq8_0_vq8_0_w64 874 21 6 187 1024
kernel_flash_attn_ext_vec_amd_dk576_dv512_kq8_0_vq8_0_w64_wg32 935 21 6 187 1024
kernel_flash_attn_ext_vec_amd_dk576_dv512_kq8_0_vq8_0_wg32 1096 37 8 187 640
kernel_flash_attn_ext_vec_amd_dk640_kf16_vturbo2 1051 6 9 232 512
kernel_flash_attn_ext_vec_amd_dk640_kf16_vturbo2_w64 1075 6 9 232 512
kernel_flash_attn_ext_vec_amd_dk640_kf16_vturbo3 1115 11 9 248 512
kernel_flash_attn_ext_vec_amd_dk640_kf16_vturbo3_w64 1149 11 9 248 576
kernel_flash_attn_ext_vec_amd_dk640_kf16_vturbo4 1054 11 9 252 512
kernel_flash_attn_ext_vec_amd_dk640_kf16_vturbo4_w64 1066 11 9 252 512
kernel_flash_attn_ext_vec_amd_dk640_kq8_0_vturbo2 1108 26 9 232 512
kernel_flash_attn_ext_vec_amd_dk640_kq8_0_vturbo2_w64 1122 26 9 232 512
kernel_flash_attn_ext_vec_amd_dk640_kq8_0_vturbo3 1189 31 9 248 640
kernel_flash_attn_ext_vec_amd_dk640_kq8_0_vturbo3_w64 1192 31 9 248 576
kernel_flash_attn_ext_vec_amd_dk640_kq8_0_vturbo4 1108 31 9 252 512
kernel_flash_attn_ext_vec_amd_dk640_kq8_0_vturbo4_w64 1120 31 9 252 512
kernel_flash_attn_ext_vec_amd_dk640_kturbo2_vf16 1057 6 9 232 512
kernel_flash_attn_ext_vec_amd_dk640_kturbo2_vf16_w64 1066 6 9 232 512
kernel_flash_attn_ext_vec_amd_dk640_kturbo2_vq8_0 1101 26 9 232 512
kernel_flash_attn_ext_vec_amd_dk640_kturbo2_vq8_0_w64 1133 26 9 232 512
kernel_flash_attn_ext_vec_amd_dk640_kturbo2_vturbo2 1161 11 9 232 512
kernel_flash_attn_ext_vec_amd_dk640_kturbo2_vturbo2_w64 1180 11 9 232 512
kernel_flash_attn_ext_vec_amd_dk640_kturbo2_vturbo3 1225 16 9 248 512
kernel_flash_attn_ext_vec_amd_dk640_kturbo2_vturbo3_w64 1260 16 9 248 640
kernel_flash_attn_ext_vec_amd_dk640_kturbo2_vturbo4 1166 16 9 252 512
kernel_flash_attn_ext_vec_amd_dk640_kturbo2_vturbo4_w64 1183 16 9 252 512
kernel_flash_attn_ext_vec_amd_dk640_kturbo3_vf16 1102 11 9 248 512
kernel_flash_attn_ext_vec_amd_dk640_kturbo3_vf16_w64 1149 11 9 248 448
kernel_flash_attn_ext_vec_amd_dk640_kturbo3_vq8_0 1165 31 9 248 512
kernel_flash_attn_ext_vec_amd_dk640_kturbo3_vq8_0_w64 1191 31 9 248 576
kernel_flash_attn_ext_vec_amd_dk640_kturbo3_vturbo2 1225 16 9 248 512
kernel_flash_attn_ext_vec_amd_dk640_kturbo3_vturbo2_w64 1241 16 9 248 512
kernel_flash_attn_ext_vec_amd_dk640_kturbo3_vturbo3 1265 21 9 248 512
kernel_flash_attn_ext_vec_amd_dk640_kturbo3_vturbo3_w64 1324 21 9 248 640
kernel_flash_attn_ext_vec_amd_dk640_kturbo3_vturbo4 1222 21 9 252 512
kernel_flash_attn_ext_vec_amd_dk640_kturbo3_vturbo4_w64 1247 21 9 252 512
kernel_flash_attn_ext_vec_amd_dk640_kturbo4_vf16 1068 11 9 252 512
kernel_flash_attn_ext_vec_amd_dk640_kturbo4_vf16_w64 1071 11 9 252 448
kernel_flash_attn_ext_vec_amd_dk640_kturbo4_vq8_0 1112 31 9 252 512
kernel_flash_attn_ext_vec_amd_dk640_kturbo4_vq8_0_w64 1130 31 9 252 576
kernel_flash_attn_ext_vec_amd_dk640_kturbo4_vturbo2 1170 16 9 252 512
kernel_flash_attn_ext_vec_amd_dk640_kturbo4_vturbo2_w64 1182 16 9 252 512
kernel_flash_attn_ext_vec_amd_dk640_kturbo4_vturbo3 1229 21 9 252 512
kernel_flash_attn_ext_vec_amd_dk640_kturbo4_vturbo3_w64 1256 21 9 252 576
kernel_flash_attn_ext_vec_amd_dk640_kturbo4_vturbo4 1135 21 9 252 512
kernel_flash_attn_ext_vec_amd_dk640_kturbo4_vturbo4_w64 1153 21 9 252 512
kernel_flash_attn_ext_vec_amd_dk64_kf16_vf16 865 1 16 200 1024
kernel_flash_attn_ext_vec_amd_dk64_kf16_vf16_w64 796 1 10 200 1024
kernel_flash_attn_ext_vec_amd_dk64_kf16_vq4_0 884 3 13 187 1024
kernel_flash_attn_ext_vec_amd_dk64_kf16_vq4_0_w64 852 3 10 200 1024
kernel_flash_attn_ext_vec_amd_dk64_kf16_vq8_0 898 9 16 200 1024
kernel_flash_attn_ext_vec_amd_dk64_kf16_vq8_0_w64 838 9 10 216 1024
kernel_flash_attn_ext_vec_amd_dk64_kq4_0_vf16 890 3 13 187 768
kernel_flash_attn_ext_vec_amd_dk64_kq4_0_vf16_w64 850 3 10 200 768
kernel_flash_attn_ext_vec_amd_dk64_kq4_0_vq4_0 942 5 13 187 1024
kernel_flash_attn_ext_vec_amd_dk64_kq4_0_vq4_0_w64 914 5 10 187 1024
kernel_flash_attn_ext_vec_amd_dk64_kq4_0_vq8_0 946 11 13 187 768
kernel_flash_attn_ext_vec_amd_dk64_kq4_0_vq8_0_w64 907 11 10 187 768
kernel_flash_attn_ext_vec_amd_dk64_kq8_0_vf16 896 9 16 200 1024
kernel_flash_attn_ext_vec_amd_dk64_kq8_0_vf16_w64 836 9 10 216 768
kernel_flash_attn_ext_vec_amd_dk64_kq8_0_vq4_0 935 11 13 187 1024
kernel_flash_attn_ext_vec_amd_dk64_kq8_0_vq4_0_w64 906 11 10 187 1024
kernel_flash_attn_ext_vec_amd_dk64_kq8_0_vq8_0 948 17 16 192 1024
kernel_flash_attn_ext_vec_amd_dk64_kq8_0_vq8_0_w64 879 17 10 187 1024
kernel_flash_attn_ext_vec_amd_dk72_kf16_vf16 821 1 10 201 1024
kernel_flash_attn_ext_vec_amd_dk72_kf16_vf16_w64 775 1 6 201 1024
kernel_flash_attn_ext_vec_bf16_dk128_dv128 FALLA
kernel_flash_attn_ext_vec_bf16_dk192_dv128 FALLA
kernel_flash_attn_ext_vec_bf16_dk192_dv192 FALLA
kernel_flash_attn_ext_vec_bf16_dk256_dv256 FALLA
kernel_flash_attn_ext_vec_bf16_dk320_dv256 FALLA
kernel_flash_attn_ext_vec_bf16_dk32_dv32 FALLA
kernel_flash_attn_ext_vec_bf16_dk512_dv512 FALLA
kernel_flash_attn_ext_vec_bf16_dk576_dv512 FALLA
kernel_flash_attn_ext_vec_bf16_dk64_dv64 FALLA
kernel_flash_attn_ext_vec_bf16_dk96_dv96 FALLA
kernel_flash_attn_ext_vec_f16_dk128_dv128 FALLA
kernel_flash_attn_ext_vec_f16_dk192_dv128 FALLA
kernel_flash_attn_ext_vec_f16_dk192_dv192 FALLA
kernel_flash_attn_ext_vec_f16_dk256_dv256 FALLA
kernel_flash_attn_ext_vec_f16_dk320_dv256 FALLA
kernel_flash_attn_ext_vec_f16_dk32_dv32 FALLA
kernel_flash_attn_ext_vec_f16_dk512_dv512 FALLA
kernel_flash_attn_ext_vec_f16_dk576_dv512 FALLA
kernel_flash_attn_ext_vec_f16_dk64_dv64 FALLA
kernel_flash_attn_ext_vec_f16_dk96_dv96 FALLA
kernel_flash_attn_ext_vec_f32_dk128_dv128 FALLA
kernel_flash_attn_ext_vec_f32_dk192_dv128 FALLA
kernel_flash_attn_ext_vec_f32_dk192_dv192 FALLA
kernel_flash_attn_ext_vec_f32_dk256_dv256 FALLA
kernel_flash_attn_ext_vec_f32_dk320_dv256 FALLA
kernel_flash_attn_ext_vec_f32_dk32_dv32 FALLA
kernel_flash_attn_ext_vec_f32_dk512_dv512 FALLA
kernel_flash_attn_ext_vec_f32_dk576_dv512 FALLA
kernel_flash_attn_ext_vec_f32_dk64_dv64 FALLA
kernel_flash_attn_ext_vec_f32_dk96_dv96 FALLA
kernel_flash_attn_ext_vec_q4_0_dk128_dv128 FALLA
kernel_flash_attn_ext_vec_q4_0_dk192_dv128 FALLA
kernel_flash_attn_ext_vec_q4_0_dk192_dv192 FALLA
kernel_flash_attn_ext_vec_q4_0_dk256_dv256 FALLA
kernel_flash_attn_ext_vec_q4_0_dk320_dv256 FALLA
kernel_flash_attn_ext_vec_q4_0_dk32_dv32 FALLA
kernel_flash_attn_ext_vec_q4_0_dk512_dv512 FALLA
kernel_flash_attn_ext_vec_q4_0_dk576_dv512 FALLA
kernel_flash_attn_ext_vec_q4_0_dk64_dv64 FALLA
kernel_flash_attn_ext_vec_q4_0_dk96_dv96 FALLA
kernel_flash_attn_ext_vec_q4_1_dk128_dv128 FALLA
kernel_flash_attn_ext_vec_q4_1_dk192_dv128 FALLA
kernel_flash_attn_ext_vec_q4_1_dk192_dv192 FALLA
kernel_flash_attn_ext_vec_q4_1_dk256_dv256 FALLA
kernel_flash_attn_ext_vec_q4_1_dk320_dv256 FALLA
kernel_flash_attn_ext_vec_q4_1_dk32_dv32 FALLA
kernel_flash_attn_ext_vec_q4_1_dk512_dv512 FALLA
kernel_flash_attn_ext_vec_q4_1_dk576_dv512 FALLA
kernel_flash_attn_ext_vec_q4_1_dk64_dv64 FALLA
kernel_flash_attn_ext_vec_q4_1_dk96_dv96 FALLA
kernel_flash_attn_ext_vec_q5_0_dk128_dv128 FALLA
kernel_flash_attn_ext_vec_q5_0_dk192_dv128 FALLA
kernel_flash_attn_ext_vec_q5_0_dk192_dv192 FALLA
kernel_flash_attn_ext_vec_q5_0_dk256_dv256 FALLA
kernel_flash_attn_ext_vec_q5_0_dk320_dv256 FALLA
kernel_flash_attn_ext_vec_q5_0_dk32_dv32 FALLA
kernel_flash_attn_ext_vec_q5_0_dk512_dv512 FALLA
kernel_flash_attn_ext_vec_q5_0_dk576_dv512 FALLA
kernel_flash_attn_ext_vec_q5_0_dk64_dv64 FALLA
kernel_flash_attn_ext_vec_q5_0_dk96_dv96 FALLA
kernel_flash_attn_ext_vec_q5_1_dk128_dv128 FALLA
kernel_flash_attn_ext_vec_q5_1_dk192_dv128 FALLA
kernel_flash_attn_ext_vec_q5_1_dk192_dv192 FALLA
kernel_flash_attn_ext_vec_q5_1_dk256_dv256 FALLA
kernel_flash_attn_ext_vec_q5_1_dk320_dv256 FALLA
kernel_flash_attn_ext_vec_q5_1_dk32_dv32 FALLA
kernel_flash_attn_ext_vec_q5_1_dk512_dv512 FALLA
kernel_flash_attn_ext_vec_q5_1_dk576_dv512 FALLA
kernel_flash_attn_ext_vec_q5_1_dk64_dv64 FALLA
kernel_flash_attn_ext_vec_q5_1_dk96_dv96 FALLA
kernel_flash_attn_ext_vec_q8_0_dk128_dv128 FALLA
kernel_flash_attn_ext_vec_q8_0_dk192_dv128 FALLA
kernel_flash_attn_ext_vec_q8_0_dk192_dv192 FALLA
kernel_flash_attn_ext_vec_q8_0_dk256_dv256 FALLA
kernel_flash_attn_ext_vec_q8_0_dk320_dv256 FALLA
kernel_flash_attn_ext_vec_q8_0_dk32_dv32 FALLA
kernel_flash_attn_ext_vec_q8_0_dk512_dv512 FALLA
kernel_flash_attn_ext_vec_q8_0_dk576_dv512 FALLA
kernel_flash_attn_ext_vec_q8_0_dk64_dv64 FALLA
kernel_flash_attn_ext_vec_q8_0_dk96_dv96 FALLA
kernel_flash_attn_ext_vec_reduce FALLA
kernel_fwht_f32_128 277 0 0 186 1024
kernel_fwht_f32_128_w64 238 0 0 186 1024
kernel_fwht_f32_256 391 0 0 186 1024
kernel_fwht_f32_256_w64 307 0 0 186 1024
kernel_fwht_f32_512 627 0 0 186 1024
kernel_fwht_f32_512_w64 442 0 0 186 1024
kernel_fwht_f32_64 260 0 0 186 1024
kernel_fwht_f32_64_w64 243 0 0 186 1024
kernel_gated_delta_net_f32_1 FALLA
kernel_gated_delta_net_f32_2 FALLA
kernel_gated_delta_net_f32_4 FALLA
kernel_geglu_erf_f16 182 0 0 182 1024
kernel_geglu_erf_f32 180 0 0 182 1024
kernel_geglu_f16 270 0 0 182 1024
kernel_geglu_f32 268 0 0 182 1024
kernel_geglu_quick_f16 208 0 0 186 1024
kernel_geglu_quick_f32 208 0 0 186 1024
kernel_get_rows_bf16 213 0 0 187 1024
kernel_get_rows_f16 202 0 0 187 1024
kernel_get_rows_f32 201 0 0 187 1024
kernel_get_rows_i32 201 0 0 187 1024
kernel_get_rows_iq1_m 2025 12 0 187 1024
kernel_get_rows_iq1_s 1989 9 0 187 1024
kernel_get_rows_iq2_s 2039 6 0 187 1024
kernel_get_rows_iq2_xs 1224 5 0 251 1024
kernel_get_rows_iq2_xxs 859 7 0 251 1024
kernel_get_rows_iq3_s 825 8 0 187 1024
kernel_get_rows_iq3_xxs 651 10 0 251 1024
kernel_get_rows_iq4_nl 329 1 0 187 1024
kernel_get_rows_iq4_xs 357 1 0 187 1024
kernel_get_rows_mxfp4 393 17 0 187 1024
kernel_get_rows_q1_0 294 2 0 187 1024
kernel_get_rows_q2_0 305 4 0 187 1024
kernel_get_rows_q2_K 328 4 0 224 1024
kernel_get_rows_q3_K 415 4 0 192 1024
kernel_get_rows_q4_0 290 9 0 187 1024
kernel_get_rows_q4_1 292 10 0 187 1024
kernel_get_rows_q4_K 325 0 0 208 1024
kernel_get_rows_q5_0 383 4 0 187 1024
kernel_get_rows_q5_1 378 4 0 187 1024
kernel_get_rows_q5_K 416 7 0 187 1024
kernel_get_rows_q6_K 388 1 0 187 1024
kernel_get_rows_q8_0 289 16 0 187 1024
kernel_group_norm_f32 533 0 6 206 1024
kernel_im2col_ext_f16 572 0 0 204 1024
kernel_im2col_ext_f32 571 0 0 204 1024
kernel_im2col_f16 412 0 0 204 1024
kernel_im2col_f32 410 0 0 204 1024
kernel_l2_norm_f32_f32 271 0 3 183 1024
kernel_l2_norm_f32_f32_4 247 0 3 183 1024
kernel_lightning_indexer_bf16 FALLA
kernel_lightning_indexer_f16 FALLA
kernel_lightning_indexer_f32 FALLA
kernel_lightning_indexer_q4_0 FALLA
kernel_lightning_indexer_q4_1 FALLA
kernel_lightning_indexer_q5_0 FALLA
kernel_lightning_indexer_q5_1 FALLA
kernel_lightning_indexer_q8_0 FALLA
kernel_memset_i64 103 0 0 186 1024
kernel_mul_mm_bf16_f32 FALLA
kernel_mul_mm_f16_f16 FALLA
kernel_mul_mm_f16_f32 FALLA
kernel_mul_mm_f16_f32_w FALLA
kernel_mul_mm_f32_f16 FALLA
kernel_mul_mm_f32_f32 FALLA
kernel_mul_mm_id_bf16_f32 FALLA
kernel_mul_mm_id_f16_f16 FALLA
kernel_mul_mm_id_f16_f32 FALLA
kernel_mul_mm_id_f16_f32_w FALLA
kernel_mul_mm_id_f32_f16 FALLA
kernel_mul_mm_id_f32_f32 FALLA
kernel_mul_mm_id_iq1_m_f16 FALLA
kernel_mul_mm_id_iq1_m_f32 FALLA
kernel_mul_mm_id_iq1_m_f32_w FALLA
kernel_mul_mm_id_iq1_s_f16 FALLA
kernel_mul_mm_id_iq1_s_f32 FALLA
kernel_mul_mm_id_iq1_s_f32_w FALLA
kernel_mul_mm_id_iq2_s_f16 FALLA
kernel_mul_mm_id_iq2_s_f32 FALLA
kernel_mul_mm_id_iq2_s_f32_w FALLA
kernel_mul_mm_id_iq2_xs_f16 FALLA
kernel_mul_mm_id_iq2_xs_f32 FALLA
kernel_mul_mm_id_iq2_xs_f32_w FALLA
kernel_mul_mm_id_iq2_xxs_f16 FALLA
kernel_mul_mm_id_iq2_xxs_f32 FALLA
kernel_mul_mm_id_iq2_xxs_f32_w FALLA
kernel_mul_mm_id_iq3_s_f16 FALLA
kernel_mul_mm_id_iq3_s_f32 FALLA
kernel_mul_mm_id_iq3_s_f32_w FALLA
kernel_mul_mm_id_iq3_xxs_f16 FALLA
kernel_mul_mm_id_iq3_xxs_f32 FALLA
kernel_mul_mm_id_iq3_xxs_f32_w FALLA
kernel_mul_mm_id_iq4_nl_f16 FALLA
kernel_mul_mm_id_iq4_nl_f32 FALLA
kernel_mul_mm_id_iq4_nl_f32_w FALLA
kernel_mul_mm_id_iq4_xs_f16 FALLA
kernel_mul_mm_id_iq4_xs_f32 FALLA
kernel_mul_mm_id_iq4_xs_f32_w FALLA
kernel_mul_mm_id_map0_ne20_1 177 1 2 182 1024
kernel_mul_mm_id_map0_ne20_10 250 10 20 183 1024
kernel_mul_mm_id_map0_ne20_16 298 16 32 192 1024
kernel_mul_mm_id_map0_ne20_2 188 2 4 182 1024
kernel_mul_mm_id_map0_ne20_22 342 22 44 183 1024
kernel_mul_mm_id_map0_ne20_4 203 4 8 182 1024
kernel_mul_mm_id_map0_ne20_5 214 5 10 182 1024
kernel_mul_mm_id_map0_ne20_6 224 6 12 182 1024
kernel_mul_mm_id_map0_ne20_8 237 8 16 182 1024
kernel_mul_mm_id_mxfp4_f16 FALLA
kernel_mul_mm_id_mxfp4_f32 FALLA
kernel_mul_mm_id_mxfp4_f32_w FALLA
kernel_mul_mm_id_q1_0_f16 FALLA
kernel_mul_mm_id_q1_0_f32 FALLA
kernel_mul_mm_id_q2_0_f16 FALLA
kernel_mul_mm_id_q2_0_f32 FALLA
kernel_mul_mm_id_q2_K_f16 FALLA
kernel_mul_mm_id_q2_K_f32 FALLA
kernel_mul_mm_id_q2_K_f32_w FALLA
kernel_mul_mm_id_q3_K_f16 FALLA
kernel_mul_mm_id_q3_K_f32 FALLA
kernel_mul_mm_id_q3_K_f32_w FALLA
kernel_mul_mm_id_q4_0_f16 FALLA
kernel_mul_mm_id_q4_0_f32 FALLA
kernel_mul_mm_id_q4_0_f32_w FALLA
kernel_mul_mm_id_q4_1_f16 FALLA
kernel_mul_mm_id_q4_1_f32 FALLA
kernel_mul_mm_id_q4_1_f32_w FALLA
kernel_mul_mm_id_q4_K_f16 FALLA
kernel_mul_mm_id_q4_K_f32 FALLA
kernel_mul_mm_id_q4_K_f32_w FALLA
kernel_mul_mm_id_q5_0_f16 FALLA
kernel_mul_mm_id_q5_0_f32 FALLA
kernel_mul_mm_id_q5_0_f32_w FALLA
kernel_mul_mm_id_q5_1_f16 FALLA
kernel_mul_mm_id_q5_1_f32 FALLA
kernel_mul_mm_id_q5_1_f32_w FALLA
kernel_mul_mm_id_q5_K_f16 FALLA
kernel_mul_mm_id_q5_K_f32 FALLA
kernel_mul_mm_id_q5_K_f32_w FALLA
kernel_mul_mm_id_q6_K_f16 FALLA
kernel_mul_mm_id_q6_K_f32 FALLA
kernel_mul_mm_id_q6_K_f32_w FALLA
kernel_mul_mm_id_q8_0_f16 FALLA
kernel_mul_mm_id_q8_0_f32 FALLA
kernel_mul_mm_id_q8_0_f32_w FALLA
kernel_mul_mm_iq1_m_f16 FALLA
kernel_mul_mm_iq1_m_f32 FALLA
kernel_mul_mm_iq1_m_f32_w FALLA
kernel_mul_mm_iq1_s_f16 FALLA
kernel_mul_mm_iq1_s_f32 FALLA
kernel_mul_mm_iq1_s_f32_w FALLA
kernel_mul_mm_iq2_s_f16 FALLA
kernel_mul_mm_iq2_s_f32 FALLA
kernel_mul_mm_iq2_s_f32_w FALLA
kernel_mul_mm_iq2_xs_f16 FALLA
kernel_mul_mm_iq2_xs_f32 FALLA
kernel_mul_mm_iq2_xs_f32_w FALLA
kernel_mul_mm_iq2_xxs_f16 FALLA
kernel_mul_mm_iq2_xxs_f32 FALLA
kernel_mul_mm_iq2_xxs_f32_w FALLA
kernel_mul_mm_iq3_s_f16 FALLA
kernel_mul_mm_iq3_s_f32 FALLA
kernel_mul_mm_iq3_s_f32_w FALLA
kernel_mul_mm_iq3_xxs_f16 FALLA
kernel_mul_mm_iq3_xxs_f32 FALLA
kernel_mul_mm_iq3_xxs_f32_w FALLA
kernel_mul_mm_iq4_nl_f16 FALLA
kernel_mul_mm_iq4_nl_f32 FALLA
kernel_mul_mm_iq4_nl_f32_w FALLA
kernel_mul_mm_iq4_xs_f16 FALLA
kernel_mul_mm_iq4_xs_f32 FALLA
kernel_mul_mm_iq4_xs_f32_w FALLA
kernel_mul_mm_mxfp4_f16 FALLA
kernel_mul_mm_mxfp4_f32 FALLA
kernel_mul_mm_mxfp4_f32_w FALLA
kernel_mul_mm_q1_0_f16 FALLA
kernel_mul_mm_q1_0_f32 FALLA
kernel_mul_mm_q2_0_f16 FALLA
kernel_mul_mm_q2_0_f32 FALLA
kernel_mul_mm_q2_K_f16 FALLA
kernel_mul_mm_q2_K_f32 FALLA
kernel_mul_mm_q2_K_f32_w FALLA
kernel_mul_mm_q3_K_f16 FALLA
kernel_mul_mm_q3_K_f32 FALLA
kernel_mul_mm_q3_K_f32_w FALLA
kernel_mul_mm_q4_0_f16 FALLA
kernel_mul_mm_q4_0_f32 FALLA
kernel_mul_mm_q4_0_f32_w FALLA
kernel_mul_mm_q4_1_f16 FALLA
kernel_mul_mm_q4_1_f32 FALLA
kernel_mul_mm_q4_1_f32_w FALLA
kernel_mul_mm_q4_K_f16 FALLA
kernel_mul_mm_q4_K_f32 FALLA
kernel_mul_mm_q4_K_f32_w FALLA
kernel_mul_mm_q5_0_f16 FALLA
kernel_mul_mm_q5_0_f32 FALLA
kernel_mul_mm_q5_0_f32_w FALLA
kernel_mul_mm_q5_1_f16 FALLA
kernel_mul_mm_q5_1_f32 FALLA
kernel_mul_mm_q5_1_f32_w FALLA
kernel_mul_mm_q5_K_f16 FALLA
kernel_mul_mm_q5_K_f32 FALLA
kernel_mul_mm_q5_K_f32_w FALLA
kernel_mul_mm_q6_K_f16 FALLA
kernel_mul_mm_q6_K_f32 FALLA
kernel_mul_mm_q6_K_f32_w FALLA
kernel_mul_mm_q8_0_f16 FALLA
kernel_mul_mm_q8_0_f32 FALLA
kernel_mul_mm_q8_0_f32_w FALLA
kernel_mul_mv_bf16_bf16 FALLA
kernel_mul_mv_bf16_bf16_4 FALLA
kernel_mul_mv_bf16_bf16_short FALLA
kernel_mul_mv_bf16_f32 FALLA
kernel_mul_mv_bf16_f32_4 FALLA
kernel_mul_mv_bf16_f32_short FALLA
kernel_mul_mv_ext_bf16_f32_r1_2 FALLA
kernel_mul_mv_ext_bf16_f32_r1_3 FALLA
kernel_mul_mv_ext_bf16_f32_r1_4 FALLA
kernel_mul_mv_ext_bf16_f32_r1_5 FALLA
kernel_mul_mv_ext_f16_f32_r1_2 FALLA
kernel_mul_mv_ext_f16_f32_r1_3 FALLA
kernel_mul_mv_ext_f16_f32_r1_4 FALLA
kernel_mul_mv_ext_f16_f32_r1_5 FALLA
kernel_mul_mv_ext_f32_f32_r1_2 FALLA
kernel_mul_mv_ext_f32_f32_r1_3 FALLA
kernel_mul_mv_ext_f32_f32_r1_4 FALLA
kernel_mul_mv_ext_f32_f32_r1_5 FALLA
kernel_mul_mv_ext_iq4_nl_f32_r1_2 FALLA
kernel_mul_mv_ext_iq4_nl_f32_r1_3 FALLA
kernel_mul_mv_ext_iq4_nl_f32_r1_4 FALLA
kernel_mul_mv_ext_iq4_nl_f32_r1_5 FALLA
kernel_mul_mv_ext_mxfp4_f32_r1_2 FALLA
kernel_mul_mv_ext_mxfp4_f32_r1_3 FALLA
kernel_mul_mv_ext_mxfp4_f32_r1_4 FALLA
kernel_mul_mv_ext_mxfp4_f32_r1_5 FALLA
kernel_mul_mv_ext_q1_0_f32_r1_2 FALLA
kernel_mul_mv_ext_q1_0_f32_r1_3 FALLA
kernel_mul_mv_ext_q1_0_f32_r1_4 FALLA
kernel_mul_mv_ext_q1_0_f32_r1_5 FALLA
kernel_mul_mv_ext_q2_0_f32_r1_2 FALLA
kernel_mul_mv_ext_q2_0_f32_r1_3 FALLA
kernel_mul_mv_ext_q2_0_f32_r1_4 FALLA
kernel_mul_mv_ext_q2_0_f32_r1_5 FALLA
kernel_mul_mv_ext_q2_K_f32_r1_2 FALLA
kernel_mul_mv_ext_q2_K_f32_r1_2_nr0_2 FALLA
kernel_mul_mv_ext_q2_K_f32_r1_3 FALLA
kernel_mul_mv_ext_q2_K_f32_r1_3_nr0_2 FALLA
kernel_mul_mv_ext_q2_K_f32_r1_4 FALLA
kernel_mul_mv_ext_q2_K_f32_r1_4_nr0_2 FALLA
kernel_mul_mv_ext_q2_K_f32_r1_5 FALLA
kernel_mul_mv_ext_q2_K_f32_r1_5_nr0_2 FALLA
kernel_mul_mv_ext_q2_K_hdr_f32_r1_2 FALLA
kernel_mul_mv_ext_q2_K_hdr_f32_r1_2_nr0_2 FALLA
kernel_mul_mv_ext_q2_K_hdr_f32_r1_3 FALLA
kernel_mul_mv_ext_q2_K_hdr_f32_r1_3_nr0_2 FALLA
kernel_mul_mv_ext_q2_K_hdr_f32_r1_4 FALLA
kernel_mul_mv_ext_q2_K_hdr_f32_r1_4_nr0_2 FALLA
kernel_mul_mv_ext_q2_K_hdr_f32_r1_5 FALLA
kernel_mul_mv_ext_q2_K_hdr_f32_r1_5_nr0_2 FALLA
kernel_mul_mv_ext_q3_K_f32_r1_2 FALLA
kernel_mul_mv_ext_q3_K_f32_r1_2_nr0_2 FALLA
kernel_mul_mv_ext_q3_K_f32_r1_3 FALLA
kernel_mul_mv_ext_q3_K_f32_r1_3_nr0_2 FALLA
kernel_mul_mv_ext_q3_K_f32_r1_4 FALLA
kernel_mul_mv_ext_q3_K_f32_r1_4_nr0_2 FALLA
kernel_mul_mv_ext_q3_K_f32_r1_5 FALLA
kernel_mul_mv_ext_q3_K_f32_r1_5_nr0_2 FALLA
kernel_mul_mv_ext_q3_K_hdr_f32_r1_2 FALLA
kernel_mul_mv_ext_q3_K_hdr_f32_r1_2_nr0_2 FALLA
kernel_mul_mv_ext_q3_K_hdr_f32_r1_3 FALLA
kernel_mul_mv_ext_q3_K_hdr_f32_r1_3_nr0_2 FALLA
kernel_mul_mv_ext_q3_K_hdr_f32_r1_4 FALLA
kernel_mul_mv_ext_q3_K_hdr_f32_r1_4_nr0_2 FALLA
kernel_mul_mv_ext_q3_K_hdr_f32_r1_5 FALLA
kernel_mul_mv_ext_q3_K_hdr_f32_r1_5_nr0_2 FALLA
kernel_mul_mv_ext_q4_0_f32_r1_2 FALLA
kernel_mul_mv_ext_q4_0_f32_r1_3 FALLA
kernel_mul_mv_ext_q4_0_f32_r1_4 FALLA
kernel_mul_mv_ext_q4_0_f32_r1_5 FALLA
kernel_mul_mv_ext_q4_1_f32_r1_2 FALLA
kernel_mul_mv_ext_q4_1_f32_r1_3 FALLA
kernel_mul_mv_ext_q4_1_f32_r1_4 FALLA
kernel_mul_mv_ext_q4_1_f32_r1_5 FALLA
kernel_mul_mv_ext_q4_K_f32_r1_2 FALLA
kernel_mul_mv_ext_q4_K_f32_r1_2_nr0_2 FALLA
kernel_mul_mv_ext_q4_K_f32_r1_3 FALLA
kernel_mul_mv_ext_q4_K_f32_r1_3_nr0_2 FALLA
kernel_mul_mv_ext_q4_K_f32_r1_4 FALLA
kernel_mul_mv_ext_q4_K_f32_r1_4_nr0_2 FALLA
kernel_mul_mv_ext_q4_K_f32_r1_5 FALLA
kernel_mul_mv_ext_q4_K_f32_r1_5_nr0_2 FALLA
kernel_mul_mv_ext_q4_K_raw_f32_r1_2 FALLA
kernel_mul_mv_ext_q4_K_raw_f32_r1_2_nr0_2 FALLA
kernel_mul_mv_ext_q4_K_raw_f32_r1_3 FALLA
kernel_mul_mv_ext_q4_K_raw_f32_r1_3_nr0_2 FALLA
kernel_mul_mv_ext_q4_K_raw_f32_r1_4 FALLA
kernel_mul_mv_ext_q4_K_raw_f32_r1_4_nr0_2 FALLA
kernel_mul_mv_ext_q4_K_raw_f32_r1_5 FALLA
kernel_mul_mv_ext_q4_K_raw_f32_r1_5_nr0_2 FALLA
kernel_mul_mv_ext_q5_0_f32_r1_2 FALLA
kernel_mul_mv_ext_q5_0_f32_r1_3 FALLA
kernel_mul_mv_ext_q5_0_f32_r1_4 FALLA
kernel_mul_mv_ext_q5_0_f32_r1_5 FALLA
kernel_mul_mv_ext_q5_1_f32_r1_2 FALLA
kernel_mul_mv_ext_q5_1_f32_r1_3 FALLA
kernel_mul_mv_ext_q5_1_f32_r1_4 FALLA
kernel_mul_mv_ext_q5_1_f32_r1_5 FALLA
kernel_mul_mv_ext_q5_K_f32_r1_2 FALLA
kernel_mul_mv_ext_q5_K_f32_r1_2_nr0_2 FALLA
kernel_mul_mv_ext_q5_K_f32_r1_3 FALLA
kernel_mul_mv_ext_q5_K_f32_r1_3_nr0_2 FALLA
kernel_mul_mv_ext_q5_K_f32_r1_4 FALLA
kernel_mul_mv_ext_q5_K_f32_r1_4_nr0_2 FALLA
kernel_mul_mv_ext_q5_K_f32_r1_5 FALLA
kernel_mul_mv_ext_q5_K_f32_r1_5_nr0_2 FALLA
kernel_mul_mv_ext_q5_K_hdr_f32_r1_2 FALLA
kernel_mul_mv_ext_q5_K_hdr_f32_r1_2_nr0_2 FALLA
kernel_mul_mv_ext_q5_K_hdr_f32_r1_3 FALLA
kernel_mul_mv_ext_q5_K_hdr_f32_r1_3_nr0_2 FALLA
kernel_mul_mv_ext_q5_K_hdr_f32_r1_4 FALLA
kernel_mul_mv_ext_q5_K_hdr_f32_r1_4_nr0_2 FALLA
kernel_mul_mv_ext_q5_K_hdr_f32_r1_5 FALLA
kernel_mul_mv_ext_q5_K_hdr_f32_r1_5_nr0_2 FALLA
kernel_mul_mv_ext_q6_K_f32_r1_2 FALLA
kernel_mul_mv_ext_q6_K_f32_r1_2_nr0_2 FALLA
kernel_mul_mv_ext_q6_K_f32_r1_3 FALLA
kernel_mul_mv_ext_q6_K_f32_r1_3_nr0_2 FALLA
kernel_mul_mv_ext_q6_K_f32_r1_4 FALLA
kernel_mul_mv_ext_q6_K_f32_r1_4_nr0_2 FALLA
kernel_mul_mv_ext_q6_K_f32_r1_5 FALLA
kernel_mul_mv_ext_q6_K_f32_r1_5_nr0_2 FALLA
kernel_mul_mv_ext_q8_0_f32_r1_2 FALLA
kernel_mul_mv_ext_q8_0_f32_r1_3 FALLA
kernel_mul_mv_ext_q8_0_f32_r1_4 FALLA
kernel_mul_mv_ext_q8_0_f32_r1_5 FALLA
kernel_mul_mv_f16_f16 FALLA
kernel_mul_mv_f16_f16_4 FALLA
kernel_mul_mv_f16_f16_short FALLA
kernel_mul_mv_f16_f32 FALLA
kernel_mul_mv_f16_f32_4 FALLA
kernel_mul_mv_f16_f32_short FALLA
kernel_mul_mv_f32_f32 FALLA
kernel_mul_mv_f32_f32_4 FALLA
kernel_mul_mv_f32_f32_short FALLA
kernel_mul_mv_id_bf16_f32 FALLA
kernel_mul_mv_id_bf16_f32_4 FALLA
kernel_mul_mv_id_f16_f32 FALLA
kernel_mul_mv_id_f16_f32_4 FALLA
kernel_mul_mv_id_f32_f32 FALLA
kernel_mul_mv_id_f32_f32_4 FALLA
kernel_mul_mv_id_iq1_m_f32 FALLA
kernel_mul_mv_id_iq1_s_f32 FALLA
kernel_mul_mv_id_iq2_s_f32 FALLA
kernel_mul_mv_id_iq2_xs_f32 FALLA
kernel_mul_mv_id_iq2_xxs_f32 FALLA
kernel_mul_mv_id_iq3_s_f32 FALLA
kernel_mul_mv_id_iq3_xxs_f32 FALLA
kernel_mul_mv_id_iq4_nl_f32 FALLA
kernel_mul_mv_id_iq4_xs_f32 FALLA
kernel_mul_mv_id_mxfp4_f32 FALLA
kernel_mul_mv_id_q1_0_f32 FALLA
kernel_mul_mv_id_q2_0_f32 FALLA
kernel_mul_mv_id_q2_K_f32 FALLA
kernel_mul_mv_id_q3_K_f32 FALLA
kernel_mul_mv_id_q4_0_f32 FALLA
kernel_mul_mv_id_q4_1_f32 FALLA
kernel_mul_mv_id_q4_K_f32 FALLA
kernel_mul_mv_id_q5_0_f32 FALLA
kernel_mul_mv_id_q5_1_f32 FALLA
kernel_mul_mv_id_q5_K_f32 FALLA
kernel_mul_mv_id_q6_K_f32 FALLA
kernel_mul_mv_id_q8_0_f32 FALLA
kernel_mul_mv_iq1_m_f32 FALLA
kernel_mul_mv_iq1_s_f32 FALLA
kernel_mul_mv_iq2_s_f32 FALLA
kernel_mul_mv_iq2_xs_f32 FALLA
kernel_mul_mv_iq2_xxs_f32 FALLA
kernel_mul_mv_iq3_s_f32 FALLA
kernel_mul_mv_iq3_xxs_f32 FALLA
kernel_mul_mv_iq4_nl_f32 FALLA
kernel_mul_mv_iq4_xs_f32 FALLA
kernel_mul_mv_mxfp4_f32 FALLA
kernel_mul_mv_q1_0_f32 FALLA
kernel_mul_mv_q2_0_f32 FALLA
kernel_mul_mv_q2_K_f32 FALLA
kernel_mul_mv_q3_K_f32 FALLA
kernel_mul_mv_q4_0_f32 FALLA
kernel_mul_mv_q4_1_f32 FALLA
kernel_mul_mv_q4_K_f32 FALLA
kernel_mul_mv_q4_K_f32_r4 FALLA
kernel_mul_mv_q5_0_f32 FALLA
kernel_mul_mv_q5_1_f32 FALLA
kernel_mul_mv_q5_K_f32 FALLA
kernel_mul_mv_q6_K_f32 FALLA
kernel_mul_mv_q8_0_f32 FALLA
kernel_mul_mv_qvk_q2_K_q4_K_q2_K_f32 FALLA
kernel_mul_mv_qvk_q4_0_q4_0_q4_0_f32 FALLA
kernel_mul_mv_qvk_q4_K_q6_K_q4_K_f32 FALLA
kernel_norm_f32 307 0 5 185 1024
kernel_norm_f32_4 328 0 5 202 1024
kernel_norm_mul_add_f32 562 0 5 185 1024
kernel_norm_mul_add_f32_4 589 0 5 202 1024
kernel_norm_mul_f32 438 0 5 185 1024
kernel_norm_mul_f32_4 472 0 5 202 1024
kernel_op_sum_f32 231 0 2 206 1024
kernel_opt_step_adamw_f32 162 0 0 187 1024
kernel_opt_step_sgd_f32 122 0 0 186 1024
kernel_pad_f32 333 0 0 183 1024
kernel_pad_f32_4 318 0 0 182 1024
kernel_pad_reflect_1d_f32 226 0 0 204 1024
kernel_pool_1d_avg_f32 460 0 0 184 1024
kernel_pool_1d_max_f32 451 0 0 188 1024
kernel_pool_2d_avg_f32 535 0 0 184 1024
kernel_pool_2d_max_f32 530 0 0 188 1024
kernel_reglu_f16 166 1 0 182 1024
kernel_reglu_f32 164 0 0 200 1024
kernel_repeat_bf16 332 1 0 185 1024
kernel_repeat_f16 303 1 0 228 1024
kernel_repeat_f32 302 0 0 228 1024
kernel_repeat_i16 303 1 0 228 1024
kernel_repeat_i32 302 0 0 228 1024
kernel_rms_norm_f32 233 0 3 185 1024
kernel_rms_norm_f32_4 251 0 3 202 1024
kernel_rms_norm_mul_add_f32 493 0 3 185 1024
kernel_rms_norm_mul_add_f32_4 511 0 3 202 1024
kernel_rms_norm_mul_f32 373 0 3 185 1024
kernel_rms_norm_mul_f32_4 390 0 3 220 1024
kernel_roll_f32 174 0 0 182 1024
kernel_rope_multi_f16 FALLA
kernel_rope_multi_f32 FALLA
kernel_rope_neox_f16 FALLA
kernel_rope_neox_f32 FALLA
kernel_rope_norm_f16 FALLA
kernel_rope_norm_f32 FALLA
kernel_rope_vision_f16 FALLA
kernel_rope_vision_f32 FALLA
kernel_rwkv_wkv6_f32 1161 0 67 187 512
kernel_rwkv_wkv7_f32 953 0 83 187 576
kernel_set_rows_bf16_i32_bf16 303 1 0 218 1024
kernel_set_rows_bf16_i64_bf16 304 1 0 218 1024
kernel_set_rows_f16_i32_f16 295 1 0 187 1024
kernel_set_rows_f16_i64_f16 295 1 0 187 1024
kernel_set_rows_f32_i32_bf16 311 0 0 187 1024
kernel_set_rows_f32_i32_f16 297 0 0 187 1024
kernel_set_rows_f32_i32_f32 295 0 0 187 1024
kernel_set_rows_f32_i32_iq4_nl 2627 0 0 187 1024
kernel_set_rows_f32_i32_q4_0 798 0 0 187 1024
kernel_set_rows_f32_i32_q4_1 833 0 0 187 1024
kernel_set_rows_f32_i32_q5_0 980 0 0 187 1024
kernel_set_rows_f32_i32_q5_1 914 0 0 187 1024
kernel_set_rows_f32_i32_q8_0 778 0 0 187 1024
kernel_set_rows_f32_i64_bf16 312 0 0 187 1024
kernel_set_rows_f32_i64_f16 297 0 0 187 1024
kernel_set_rows_f32_i64_f32 295 0 0 187 1024
kernel_set_rows_f32_i64_iq4_nl 2627 0 0 187 1024
kernel_set_rows_f32_i64_q4_0 799 0 0 187 1024
kernel_set_rows_f32_i64_q4_1 834 0 0 187 1024
kernel_set_rows_f32_i64_q5_0 981 0 0 187 1024
kernel_set_rows_f32_i64_q5_1 915 0 0 187 1024
kernel_set_rows_f32_i64_q8_0 779 0 0 187 1024
kernel_set_rows_turbo2_i32 3530 8 0 232 448
kernel_set_rows_turbo2_i64 3531 8 0 232 448
kernel_set_rows_turbo3_i32 846 0 0 248 1024
kernel_set_rows_turbo3_i32_w64 748 0 0 248 1024
kernel_set_rows_turbo3_i64 843 0 0 248 1024
kernel_set_rows_turbo3_i64_w64 747 0 0 248 1024
kernel_set_rows_turbo4_i32 1017 0 0 252 1024
kernel_set_rows_turbo4_i32_w64 813 0 0 252 1024
kernel_set_rows_turbo4_i64 1013 0 0 252 1024
kernel_set_rows_turbo4_i64_w64 812 0 0 252 1024
kernel_silu_back_f32 131 0 0 186 1024
kernel_snake_bf16 181 0 0 183 1024
kernel_snake_f16 182 0 0 224 1024
kernel_snake_f32 179 0 0 183 1024
kernel_soft_max_f16 537 0 6 188 1024
kernel_soft_max_f16_4 592 0 6 222 1024
kernel_soft_max_f32 528 0 6 188 1024
kernel_soft_max_f32_4 583 0 6 202 1024
kernel_solve_tri_f32 FALLA
kernel_ssm_conv_f32_f32 147 0 0 187 1024
kernel_ssm_conv_f32_f32_4 164 0 0 206 1024
kernel_ssm_conv_f32_f32_batched FALLA
kernel_ssm_conv_f32_f32_batched_4 FALLA
kernel_ssm_conv_f32_f32_decode 198 0 0 187 1024
kernel_ssm_conv_f32_f32_decode_4 179 0 0 192 1024
kernel_ssm_conv_f32_f32_decode_silu 171 0 0 187 1024
kernel_ssm_conv_f32_f32_decode_silu_4 187 0 0 206 1024
kernel_ssm_scan_f32 702 0 7 183 1024
kernel_sum_rows_f32_f32 FALLA
kernel_sum_rows_f32_f32_4 FALLA
kernel_swiglu_f16 171 0 0 187 1024
kernel_swiglu_f32 170 0 0 187 1024
kernel_swiglu_oai_f16 172 0 0 187 1024
kernel_swiglu_oai_f32 171 0 0 187 1024
kernel_timestep_embedding_f32 178 0 0 186 1024
kernel_topk_moe_f32 757 0 51 184 1024
kernel_tri_bf16_0 173 0 0 204 1024
kernel_tri_bf16_1 173 0 0 204 1024
kernel_tri_bf16_2 173 0 0 204 1024
kernel_tri_bf16_3 173 0 0 204 1024
kernel_tri_f16_0 173 1 0 185 1024
kernel_tri_f16_1 173 1 0 185 1024
kernel_tri_f16_2 172 1 0 185 1024
kernel_tri_f16_3 172 1 0 185 1024
kernel_tri_f32_0 170 0 0 185 1024
kernel_tri_f32_1 170 0 0 185 1024
kernel_tri_f32_2 169 0 0 185 1024
kernel_tri_f32_3 169 0 0 185 1024
kernel_turbo_wht 2627 192 0 186 768
kernel_unary_f16_f16 FALLA
kernel_unary_f16_f16_4 FALLA
kernel_unary_f32_f32 FALLA
kernel_unary_f32_f32_4 FALLA
kernel_upscale_bicubic_f32 761 0 0 183 1024
kernel_upscale_bilinear_f32 FALLA
kernel_upscale_nearest_f32 369 0 0 186 1024
TOTAL volcados=1036 fallos=566
