Lines Matching +full:k +full:- +full:block
3 # This source code is licensed under the BSD-style license found in the
7 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x8__aarch32_neon_mlal_lane_cortex_a7
9 k-block: 8
11 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x8__aarch32_neon_mlal_lane_prfm_cortex_a7
13 k-block: 8
15 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x8__aarch32_neonv8_mlal_lane_cortex_a35
17 k-block: 8
19 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x8__aarch32_neonv8_mlal_lane_prfm_cortex_a35
21 k-block: 8
23 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x8__aarch32_neon_mlal_lane_cortex_a7
25 k-block: 8
27 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x8__aarch32_neon_mlal_lane_cortex_a53
29 k-block: 8
31 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x8__aarch32_neon_mlal_lane_ld64
33 k-block: 8
35 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x8__aarch32_neon_mlal_lane_prfm_cortex_a7
37 k-block: 8
39 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x8__aarch32_neon_mlal_lane_prfm_cortex_a53
41 k-block: 8
43 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x8__aarch32_neon_mlal_lane_prfm_ld64
45 k-block: 8
47 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x8__aarch32_neonv8_mlal_lane_cortex_a35
49 k-block: 8
51 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x8__aarch32_neonv8_mlal_lane_cortex_a53
53 k-block: 8
55 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x8__aarch32_neonv8_mlal_lane_ld64
57 k-block: 8
59 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x8__aarch32_neonv8_mlal_lane_prfm_cortex_a35
61 k-block: 8
63 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x8__aarch32_neonv8_mlal_lane_prfm_cortex_a53
65 k-block: 8
67 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x8__aarch32_neonv8_mlal_lane_prfm_ld64
69 k-block: 8
71 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x8c4__aarch32_neondot_cortex_a55
73 k-block: 8
75 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x8c4__aarch32_neondot_ld64
77 k-block: 8
80 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x8c8__aarch64_neon_mlal
82 k-block: 16
84 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x8c8__aarch64_neon_mlal_cortex_a53
86 k-block: 16
88 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x8c8__aarch64_neon_mlal_prfm
90 k-block: 16
92 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x8c8__aarch64_neon_mlal_prfm_cortex_a53
94 k-block: 16
96 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x16c4__aarch64_neondot_ld32
98 k-block: 4
100 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x16c4__aarch64_neondot_ld64
102 k-block: 8
104 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x8c8__aarch64_neon_mlal
106 k-block: 16
108 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x8c8__aarch64_neon_mlal_cortex_a53
110 k-block: 16
112 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x8c8__aarch64_neon_mlal_prfm
114 k-block: 16
116 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x8c8__aarch64_neon_mlal_prfm_cortex_a53
118 k-block: 16
120 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x8c8__aarch64_neon_mull
122 k-block: 8
124 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x8c16__aarch64_neon_mlal
126 k-block: 16
128 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x16__aarch64_neon_mlal_lane_cortex_a53
130 k-block: 8
132 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x16__aarch64_neon_mlal_lane_ld64
134 k-block: 8
136 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x16__aarch64_neon_mlal_lane_prfm_cortex_a53
138 k-block: 8
140 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x16__aarch64_neon_mlal_lane_prfm_ld64
142 k-block: 8
144 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x16c4__aarch64_neondot_cortex_a55
146 k-block: 16
148 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x16c4__aarch64_neondot_ld32
150 k-block: 4
152 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x16c4__aarch64_neondot_ld64
154 k-block: 8
156 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x16c4__aarch64_neondot_ld128
158 k-block: 16
162 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x1c4__armsimd32
164 k-block: 4
165 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x2c4__armsimd32
167 k-block: 4
168 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x1c4__armsimd32
170 k-block: 4
171 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x2c4__armsimd32
173 k-block: 4
176 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x8__neon_mlal_lane
178 k-block: 8
179 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x8__neon_mlal_lane_prfm
181 k-block: 8
182 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x8__neonv8_mlal_lane
184 k-block: 8
185 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x8__neonv8_mlal_lane_prfm
187 k-block: 8
188 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x8c2__neon_mlal_dup
190 k-block: 16
191 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x8c2__neon_mlal_ld1r
193 k-block: 16
194 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x8c2__neon_mlal_ld2r
196 k-block: 16
197 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x8c2__neon_mlal_ld4r
199 k-block: 16
200 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x8c2__neonv8_mlal_dup
202 k-block: 16
203 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x8c2__neonv8_mlal_ld1r
205 k-block: 16
206 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x8c2__neonv8_mlal_ld2r
208 k-block: 16
209 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x8c2__neonv8_mlal_ld4r
211 k-block: 16
212 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x8c2s4__neon_mlal
214 k-block: 16
215 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x8c2s4__neonv8_mlal
217 k-block: 16
218 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x8c4__neon_mlal_dup
220 k-block: 16
221 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x8c4__neon_mlal_ld1r
223 k-block: 16
224 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x8c4__neon_mlal_ld2r
226 k-block: 16
227 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x8c4__neondot
229 k-block: 8
230 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x8c4__neonv8_mlal_dup
232 k-block: 16
233 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x8c4__neonv8_mlal_ld1r
235 k-block: 16
236 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x8c4__neonv8_mlal_ld2r
238 k-block: 16
239 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x8c4s2__neon_mlal
241 k-block: 16
242 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x8c4s2__neonv8_mlal
244 k-block: 16
245 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x8c8__neon_mlal
247 k-block: 16
248 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x8c8__neonv8_mlal
250 k-block: 16
251 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x16__neon_mlal_lane
253 k-block: 8
254 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x16__neon_mlal_lane_prfm
256 k-block: 8
257 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x16__neonv8_mlal_lane
259 k-block: 8
260 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x16__neonv8_mlal_lane_prfm
262 k-block: 8
263 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x16c4__neondot
265 k-block: 8
266 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x8__neon_mlal_lane
268 k-block: 8
269 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x8__neon_mlal_lane_prfm
271 k-block: 8
272 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x8__neonv8_mlal_lane
274 k-block: 8
275 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x8__neonv8_mlal_lane_prfm
277 k-block: 8
278 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x8c2__neon_mlal_dup
280 k-block: 16
281 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x8c2__neon_mlal_ld1r
283 k-block: 16
284 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x8c2__neon_mlal_ld2r
286 k-block: 16
287 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x8c2__neon_mlal_ld4r
289 k-block: 16
290 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x8c2__neonv8_mlal_dup
292 k-block: 16
293 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x8c2__neonv8_mlal_ld1r
295 k-block: 16
296 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x8c2__neonv8_mlal_ld2r
298 k-block: 16
299 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x8c2__neonv8_mlal_ld4r
301 k-block: 16
302 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x8c2s4__neon_mlal
304 k-block: 16
305 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x8c2s4__neonv8_mlal
307 k-block: 16
308 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x8c4__neon_mlal_dup
310 k-block: 16
311 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x8c4__neon_mlal_ld1r
313 k-block: 16
314 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x8c4__neon_mlal_ld2r
316 k-block: 16
317 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x8c4__neonv8_mlal_dup
319 k-block: 16
320 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x8c4__neonv8_mlal_ld1r
322 k-block: 16
323 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x8c4__neonv8_mlal_ld2r
325 k-block: 16
326 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x8c4s2__neon_mlal
328 k-block: 16
329 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x8c4s2__neonv8_mlal
331 k-block: 16
332 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x8c8__neon_mlal
334 k-block: 16
335 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x8c8__neonv8_mlal
337 k-block: 16
338 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x16__neon_mlal_lane
340 k-block: 8
341 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x16__neon_mlal_lane_prfm
343 k-block: 8
344 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x16__neonv8_mlal_lane
346 k-block: 8
347 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x16__neonv8_mlal_lane_prfm
349 k-block: 8
350 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x8__neon_mlal_lane
352 k-block: 8
353 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x8__neon_mlal_lane_prfm
355 k-block: 8
356 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x8__neonv8_mlal_lane
358 k-block: 8
359 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x8__neonv8_mlal_lane_prfm
361 k-block: 8
362 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x16__neon_mlal_lane
364 k-block: 8
365 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x16__neon_mlal_lane_prfm
367 k-block: 8
368 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x16__neonv8_mlal_lane
370 k-block: 8
371 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x16__neonv8_mlal_lane_prfm
373 k-block: 8
374 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x8__neon_mlal_lane
376 k-block: 8
377 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x8__neon_mlal_lane_prfm
379 k-block: 8
380 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x8__neonv8_mlal_lane
382 k-block: 8
383 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x8__neonv8_mlal_lane_prfm
385 k-block: 8
386 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x8c4__neondot
388 k-block: 8
389 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x16__neon_mlal_lane
391 k-block: 8
392 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x16__neon_mlal_lane_prfm
394 k-block: 8
395 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x16__neonv8_mlal_lane
397 k-block: 8
398 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x16__neonv8_mlal_lane_prfm
400 k-block: 8
401 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x16c4__neondot
403 k-block: 8
404 - name: xnn_qc8_gemm_minmax_fp32_ukernel_6x8__neon_mlal_lane
406 k-block: 8
407 - name: xnn_qc8_gemm_minmax_fp32_ukernel_6x8__neon_mlal_lane_prfm
409 k-block: 8
410 - name: xnn_qc8_gemm_minmax_fp32_ukernel_6x8__neonv8_mlal_lane
412 k-block: 8
413 - name: xnn_qc8_gemm_minmax_fp32_ukernel_6x8__neonv8_mlal_lane_prfm
415 k-block: 8
416 - name: xnn_qc8_gemm_minmax_fp32_ukernel_6x8c4__neondot
418 k-block: 8
419 - name: xnn_qc8_gemm_minmax_fp32_ukernel_6x16__neon_mlal_lane
421 k-block: 8
422 - name: xnn_qc8_gemm_minmax_fp32_ukernel_6x16__neon_mlal_lane_prfm
424 k-block: 8
425 - name: xnn_qc8_gemm_minmax_fp32_ukernel_6x16__neonv8_mlal_lane
427 k-block: 8
428 - name: xnn_qc8_gemm_minmax_fp32_ukernel_6x16__neonv8_mlal_lane_prfm
430 k-block: 8
431 - name: xnn_qc8_gemm_minmax_fp32_ukernel_6x16c4__neondot
433 k-block: 8
434 - name: xnn_qc8_gemm_minmax_fp32_ukernel_8x8c4__neondot
436 k-block: 8
437 - name: xnn_qc8_gemm_minmax_fp32_ukernel_8x16c4__neondot
439 k-block: 8
441 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x4c2__sse2_ld64
443 k-block: 8
444 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x4c2__sse41_ld64
446 k-block: 8
447 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x4c2__sse2_ld64
449 k-block: 8
450 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x4c2__sse41_ld64
452 k-block: 8
453 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x4c2__sse2_ld64
455 k-block: 8
456 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x4c2__sse41_ld64
458 k-block: 8
459 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x4c2__sse2_ld64
461 k-block: 8
462 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x4c2__sse41_ld64
464 k-block: 8
466 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x4c2__avx_ld64
468 k-block: 8
469 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x4c2__xop_ld64
471 k-block: 8
472 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x4c2__avx_ld64
474 k-block: 8
475 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x4c2__xop_ld64
477 k-block: 8
478 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x4c2__avx_ld64
480 k-block: 8
481 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x4c2__xop_ld64
483 k-block: 8
484 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x4c2__avx_ld64
486 k-block: 8
487 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x4c2__xop_ld64
489 k-block: 8
491 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x4c2__sse2_ld128
493 k-block: 8
494 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x4c2__sse41_ld128
496 k-block: 8
497 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x4c2__sse2_ld128
499 k-block: 8
500 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x4c2__sse41_ld128
502 k-block: 8
503 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x4c2__sse2_ld128
505 k-block: 8
506 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x4c2__sse41_ld128
508 k-block: 8
509 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x4c2__sse2_ld128
511 k-block: 8
512 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x4c2__sse41_ld128
514 k-block: 8
516 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x4c2__avx_ld128
518 k-block: 8
519 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x4c2__xop_ld128
521 k-block: 8
522 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x4c2__avx_ld128
524 k-block: 8
525 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x4c2__xop_ld128
527 k-block: 8
528 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x4c2__avx_ld128
530 k-block: 8
531 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x4c2__xop_ld128
533 k-block: 8
534 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x4c2__avx_ld128
536 k-block: 8
537 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x4c2__xop_ld128
539 k-block: 8
541 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x4c2s4__sse2_ld64
543 k-block: 8
544 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x4c2s4__sse41_ld64
546 k-block: 8
547 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x4c2s4__sse2_ld64
549 k-block: 8
550 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x4c2s4__sse41_ld64
552 k-block: 8
553 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x4c2s4__sse2_ld64
555 k-block: 8
556 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x4c2s4__sse41_ld64
558 k-block: 8
559 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x4c2s4__sse2_ld64
561 k-block: 8
562 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x4c2s4__sse41_ld64
564 k-block: 8
566 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x4c2s4__avx_ld64
568 k-block: 8
569 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x4c2s4__xop_ld64
571 k-block: 8
572 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x4c2s4__avx_ld64
574 k-block: 8
575 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x4c2s4__xop_ld64
577 k-block: 8
578 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x4c2s4__avx_ld64
580 k-block: 8
581 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x4c2s4__xop_ld64
583 k-block: 8
584 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x4c2s4__avx_ld64
586 k-block: 8
587 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x4c2s4__xop_ld64
589 k-block: 8
591 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x4c2s4__sse2_ld128
593 k-block: 8
594 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x4c2s4__sse41_ld128
596 k-block: 8
597 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x4c2s4__sse2_ld128
599 k-block: 8
600 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x4c2s4__sse41_ld128
602 k-block: 8
603 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x4c2s4__sse2_ld128
605 k-block: 8
606 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x4c2s4__sse41_ld128
608 k-block: 8
609 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x4c2s4__sse2_ld128
611 k-block: 8
612 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x4c2s4__sse41_ld128
614 k-block: 8
616 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x4c2s4__avx_ld128
618 k-block: 8
619 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x4c2s4__xop_ld128
621 k-block: 8
622 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x4c2s4__avx_ld128
624 k-block: 8
625 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x4c2s4__xop_ld128
627 k-block: 8
628 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x4c2s4__avx_ld128
630 k-block: 8
631 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x4c2s4__xop_ld128
633 k-block: 8
634 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x4c2s4__avx_ld128
636 k-block: 8
637 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x4c2s4__xop_ld128
639 k-block: 8
641 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x4c8__sse2_ld64
643 k-block: 8
644 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x4c8__sse41_ld64
646 k-block: 8
647 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x4c8__sse2_ld64
649 k-block: 8
650 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x4c8__sse41_ld64
652 k-block: 8
653 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x4c8__sse2_ld64
655 k-block: 8
656 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x4c8__sse41_ld64
658 k-block: 8
660 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x4c8__avx_ld64
662 k-block: 8
663 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x4c8__xop_ld64
665 k-block: 8
666 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x4c8__avx_ld64
668 k-block: 8
669 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x4c8__xop_ld64
671 k-block: 8
672 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x4c8__avx_ld64
674 k-block: 8
675 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x4c8__xop_ld64
677 k-block: 8
679 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x4c8__sse2_ld128
681 k-block: 8
682 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x4c8__sse41_ld128
684 k-block: 8
685 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x4c8__sse2_ld128
687 k-block: 8
688 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x4c8__sse41_ld128
690 k-block: 8
691 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x4c8__sse2_ld128
693 k-block: 8
694 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x4c8__sse41_ld128
696 k-block: 8
698 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x4c8__avx_ld128
700 k-block: 8
701 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x4c8__xop_ld128
703 k-block: 8
704 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x4c8__avx_ld128
706 k-block: 8
707 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x4c8__xop_ld128
709 k-block: 8
710 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x4c8__avx_ld128
712 k-block: 8
713 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x4c8__xop_ld128
715 k-block: 8
717 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x8c8__avx2
719 k-block: 8
720 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x8c8__avx2
722 k-block: 8
723 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x8c8__avx2
725 k-block: 8
726 - name: xnn_qc8_gemm_xw_minmax_fp32_ukernel_1x8c8__avx2
728 k-block: 8
729 - name: xnn_qc8_gemm_xw_minmax_fp32_ukernel_2x8c8__avx2
731 k-block: 8
732 - name: xnn_qc8_gemm_xw_minmax_fp32_ukernel_3x8c8__avx2
734 k-block: 8
736 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x16c8__avx512skx
738 k-block: 8
739 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x16c8__avx512skx
741 k-block: 8
742 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x16c8__avx512skx
744 k-block: 8
745 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x16c8__avx512skx
747 k-block: 8
749 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x4c2__wasmsimd_dot16x2_ld64
751 k-block: 8
752 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x4c2__wasmsimd_dot16x2_ld128
754 k-block: 8
755 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x4c2s4__wasmsimd_dot16x2_ld64
757 k-block: 8
758 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x4c2s4__wasmsimd_dot16x2_ld128
760 k-block: 8
761 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x4c8__wasmsimd_dot16x2_ld64
763 k-block: 8
764 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x4c8__wasmsimd_dot16x2_ld128
766 k-block: 8
767 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x4c2__wasmsimd_dot16x2_ld64
769 k-block: 8
770 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x4c2__wasmsimd_dot16x2_ld128
772 k-block: 8
773 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x4c2s4__wasmsimd_dot16x2_ld64
775 k-block: 8
776 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x4c2s4__wasmsimd_dot16x2_ld128
778 k-block: 8
779 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x4c8__wasmsimd_dot16x2_ld64
781 k-block: 8
782 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x4c8__wasmsimd_dot16x2_ld128
784 k-block: 8
785 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x4c2__wasmsimd_dot16x2_ld64
787 k-block: 8
788 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x4c2__wasmsimd_dot16x2_ld128
790 k-block: 8
791 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x4c2s4__wasmsimd_dot16x2_ld64
793 k-block: 8
794 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x4c2s4__wasmsimd_dot16x2_ld128
796 k-block: 8
797 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x4c8__wasmsimd_dot16x2_ld64
799 k-block: 8
800 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x4c8__wasmsimd_dot16x2_ld128
802 k-block: 8
803 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x4c2__wasmsimd_dot16x2_ld64
805 k-block: 8
806 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x4c2__wasmsimd_dot16x2_ld128
808 k-block: 8
809 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x4c2s4__wasmsimd_dot16x2_ld64
811 k-block: 8
812 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x4c2s4__wasmsimd_dot16x2_ld128
814 k-block: 8
815 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x4c8__wasmsimd_dot16x2_ld64
817 k-block: 8
818 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x4c8__wasmsimd_dot16x2_ld128
820 k-block: 8
822 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x2__wasm_fmagic
824 k-block: 1
825 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x4__wasm_fmagic
827 k-block: 1
828 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x2__wasm_fmagic
830 k-block: 1
831 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x4__wasm_fmagic
833 k-block: 1
834 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x2__wasm_fmagic
836 k-block: 1
837 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x4__wasm_fmagic
839 k-block: 1
840 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x2__wasm_fmagic
842 k-block: 1
843 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x4__wasm_fmagic
845 k-block: 1
847 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x2__scalar_fmagic
849 k-block: 1
850 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x2__scalar_imagic
852 k-block: 1
853 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x2__scalar_lrintf
855 k-block: 1
856 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x4__scalar_fmagic
858 k-block: 1
859 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x4__scalar_imagic
861 k-block: 1
862 - name: xnn_qc8_gemm_minmax_fp32_ukernel_1x4__scalar_lrintf
864 k-block: 1
865 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x2__scalar_fmagic
867 k-block: 1
868 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x2__scalar_imagic
870 k-block: 1
871 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x2__scalar_lrintf
873 k-block: 1
874 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x4__scalar_fmagic
876 k-block: 1
877 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x4__scalar_imagic
879 k-block: 1
880 - name: xnn_qc8_gemm_minmax_fp32_ukernel_2x4__scalar_lrintf
882 k-block: 1
883 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x2__scalar_fmagic
885 k-block: 1
886 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x2__scalar_imagic
888 k-block: 1
889 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x2__scalar_lrintf
891 k-block: 1
892 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x4__scalar_fmagic
894 k-block: 1
895 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x4__scalar_imagic
897 k-block: 1
898 - name: xnn_qc8_gemm_minmax_fp32_ukernel_3x4__scalar_lrintf
900 k-block: 1
901 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x2__scalar_fmagic
903 k-block: 1
904 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x2__scalar_imagic
906 k-block: 1
907 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x2__scalar_lrintf
909 k-block: 1
910 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x4__scalar_fmagic
912 k-block: 1
913 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x4__scalar_imagic
915 k-block: 1
916 - name: xnn_qc8_gemm_minmax_fp32_ukernel_4x4__scalar_lrintf
918 k-block: 1
920 - name: xnn_generate_qc8_gemm_fp32_ukernel_4x8__aarch32_neonv8_mlal_lane_ld64
922 k-block: 8
924 - name: xnn_generate_qc8_gemm_fp32_ukernel_4x8__aarch32_neonv8_mlal_lane_prfm_ld64
926 k-block: 8
928 - name: xnn_generate_qc8_gemm_fp32_ukernel_4x8c4__aarch32_neondot_ld64
930 k-block: 8