Lines Matching +full:k +full:- +full:block

3 # This source code is licensed under the BSD-style license found in the
7 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x16c4__aarch64_neondot_cortex_a55
9 k-block: 16
11 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x16c4__aarch64_neondot_ld128
13 k-block: 16
17 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x1c4__armsimd32
19 k-block: 4
20 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x2c4__armsimd32
22 k-block: 4
23 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x1c4__armsimd32
25 k-block: 4
26 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x2c4__armsimd32
28 k-block: 4
31 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x8__neon_mlal_lane
33 k-block: 8
34 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x16__neon_mlal_lane
36 k-block: 8
37 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x16__neonv8_mlal_lane
39 k-block: 8
40 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x16c4__neondot
42 k-block: 8
43 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x16c4__neondot
45 k-block: 8
46 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x8__neon_mlal_lane
48 k-block: 8
49 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x16__neon_mlal_lane
51 k-block: 8
52 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x16__neonv8_mlal_lane
54 k-block: 8
55 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x16c4__neondot
57 k-block: 8
59 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x4c2__sse2_ld64
61 k-block: 8
62 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x4c2__sse41_ld64
64 k-block: 8
65 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x4c2__sse2_ld64
67 k-block: 8
68 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x4c2__sse41_ld64
70 k-block: 8
71 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x4c2__sse2_ld64
73 k-block: 8
74 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x4c2__sse41_ld64
76 k-block: 8
77 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x4c2__sse2_ld64
79 k-block: 8
80 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x4c2__sse41_ld64
82 k-block: 8
84 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x4c2__avx_ld64
86 k-block: 8
87 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x4c2__xop_ld64
89 k-block: 8
90 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x4c2__avx_ld64
92 k-block: 8
93 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x4c2__xop_ld64
95 k-block: 8
96 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x4c2__avx_ld64
98 k-block: 8
99 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x4c2__xop_ld64
101 k-block: 8
102 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x4c2__avx_ld64
104 k-block: 8
105 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x4c2__xop_ld64
107 k-block: 8
109 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x4c2__sse2_ld128
111 k-block: 8
112 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x4c2__sse41_ld128
114 k-block: 8
115 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x4c2__sse2_ld128
117 k-block: 8
118 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x4c2__sse41_ld128
120 k-block: 8
121 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x4c2__sse2_ld128
123 k-block: 8
124 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x4c2__sse41_ld128
126 k-block: 8
127 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x4c2__sse2_ld128
129 k-block: 8
130 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x4c2__sse41_ld128
132 k-block: 8
134 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x4c2__avx_ld128
136 k-block: 8
137 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x4c2__xop_ld128
139 k-block: 8
140 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x4c2__avx_ld128
142 k-block: 8
143 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x4c2__xop_ld128
145 k-block: 8
146 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x4c2__avx_ld128
148 k-block: 8
149 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x4c2__xop_ld128
151 k-block: 8
152 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x4c2__avx_ld128
154 k-block: 8
155 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x4c2__xop_ld128
157 k-block: 8
159 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x4c2s4__sse2_ld64
161 k-block: 8
162 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x4c2s4__sse41_ld64
164 k-block: 8
165 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x4c2s4__sse2_ld64
167 k-block: 8
168 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x4c2s4__sse41_ld64
170 k-block: 8
171 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x4c2s4__sse2_ld64
173 k-block: 8
174 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x4c2s4__sse41_ld64
176 k-block: 8
177 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x4c2s4__sse2_ld64
179 k-block: 8
180 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x4c2s4__sse41_ld64
182 k-block: 8
184 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x4c2s4__avx_ld64
186 k-block: 8
187 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x4c2s4__xop_ld64
189 k-block: 8
190 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x4c2s4__avx_ld64
192 k-block: 8
193 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x4c2s4__xop_ld64
195 k-block: 8
196 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x4c2s4__avx_ld64
198 k-block: 8
199 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x4c2s4__xop_ld64
201 k-block: 8
202 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x4c2s4__avx_ld64
204 k-block: 8
205 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x4c2s4__xop_ld64
207 k-block: 8
209 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x4c2s4__sse2_ld128
211 k-block: 8
212 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x4c2s4__sse41_ld128
214 k-block: 8
215 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x4c2s4__sse2_ld128
217 k-block: 8
218 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x4c2s4__sse41_ld128
220 k-block: 8
221 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x4c2s4__sse2_ld128
223 k-block: 8
224 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x4c2s4__sse41_ld128
226 k-block: 8
227 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x4c2s4__sse2_ld128
229 k-block: 8
230 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x4c2s4__sse41_ld128
232 k-block: 8
234 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x4c2s4__avx_ld128
236 k-block: 8
237 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x4c2s4__xop_ld128
239 k-block: 8
240 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x4c2s4__avx_ld128
242 k-block: 8
243 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x4c2s4__xop_ld128
245 k-block: 8
246 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x4c2s4__avx_ld128
248 k-block: 8
249 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x4c2s4__xop_ld128
251 k-block: 8
252 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x4c2s4__avx_ld128
254 k-block: 8
255 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x4c2s4__xop_ld128
257 k-block: 8
259 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x4c8__sse2_ld64
261 k-block: 8
262 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x4c8__sse41_ld64
264 k-block: 8
265 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x4c8__sse2_ld64
267 k-block: 8
268 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x4c8__sse41_ld64
270 k-block: 8
271 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x4c8__sse2_ld64
273 k-block: 8
274 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x4c8__sse41_ld64
276 k-block: 8
278 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x4c8__avx_ld64
280 k-block: 8
281 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x4c8__xop_ld64
283 k-block: 8
284 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x4c8__avx_ld64
286 k-block: 8
287 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x4c8__xop_ld64
289 k-block: 8
290 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x4c8__avx_ld64
292 k-block: 8
293 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x4c8__xop_ld64
295 k-block: 8
297 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x4c8__sse2_ld128
299 k-block: 8
300 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x4c8__sse41_ld128
302 k-block: 8
303 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x4c8__sse2_ld128
305 k-block: 8
306 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x4c8__sse41_ld128
308 k-block: 8
309 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x4c8__sse2_ld128
311 k-block: 8
312 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x4c8__sse41_ld128
314 k-block: 8
316 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x4c8__avx_ld128
318 k-block: 8
319 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x4c8__xop_ld128
321 k-block: 8
322 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x4c8__avx_ld128
324 k-block: 8
325 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x4c8__xop_ld128
327 k-block: 8
328 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x4c8__avx_ld128
330 k-block: 8
331 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x4c8__xop_ld128
333 k-block: 8
335 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x8c8__avx2
337 k-block: 8
338 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x8c8__avx2
340 k-block: 8
341 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x8c8__avx2
343 k-block: 8
345 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x16c8__avx512skx
347 k-block: 8
348 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x16c8__avx512skx
350 k-block: 8
351 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x16c8__avx512skx
353 k-block: 8
354 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x16c8__avx512skx
356 k-block: 8
358 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x4c2__wasmsimd_dot16x2_ld64
360 k-block: 8
361 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x4c2__wasmsimd_dot16x2_ld128
363 k-block: 8
364 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x4c2s4__wasmsimd_dot16x2_ld64
366 k-block: 8
367 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x4c2s4__wasmsimd_dot16x2_ld128
369 k-block: 8
370 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x4c8__wasmsimd_dot16x2_ld64
372 k-block: 8
373 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x4c8__wasmsimd_dot16x2_ld128
375 k-block: 8
376 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x4c2__wasmsimd_dot16x2_ld64
378 k-block: 8
379 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x4c2__wasmsimd_dot16x2_ld128
381 k-block: 8
382 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x4c2s4__wasmsimd_dot16x2_ld64
384 k-block: 8
385 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x4c2s4__wasmsimd_dot16x2_ld128
387 k-block: 8
388 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x4c8__wasmsimd_dot16x2_ld64
390 k-block: 8
391 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x4c8__wasmsimd_dot16x2_ld128
393 k-block: 8
394 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x4c2__wasmsimd_dot16x2_ld64
396 k-block: 8
397 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x4c2__wasmsimd_dot16x2_ld128
399 k-block: 8
400 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x4c2s4__wasmsimd_dot16x2_ld64
402 k-block: 8
403 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x4c2s4__wasmsimd_dot16x2_ld128
405 k-block: 8
406 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x4c8__wasmsimd_dot16x2_ld64
408 k-block: 8
409 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x4c8__wasmsimd_dot16x2_ld128
411 k-block: 8
412 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x4c2__wasmsimd_dot16x2_ld64
414 k-block: 8
415 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x4c2__wasmsimd_dot16x2_ld128
417 k-block: 8
418 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x4c2s4__wasmsimd_dot16x2_ld64
420 k-block: 8
421 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x4c2s4__wasmsimd_dot16x2_ld128
423 k-block: 8
424 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x4c8__wasmsimd_dot16x2_ld64
426 k-block: 8
427 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x4c8__wasmsimd_dot16x2_ld128
429 k-block: 8
431 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x2__wasm_fmagic
433 k-block: 1
434 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x4__wasm_fmagic
436 k-block: 1
437 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x2__wasm_fmagic
439 k-block: 1
440 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x4__wasm_fmagic
442 k-block: 1
443 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x2__wasm_fmagic
445 k-block: 1
446 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x4__wasm_fmagic
448 k-block: 1
449 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x2__wasm_fmagic
451 k-block: 1
452 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x4__wasm_fmagic
454 k-block: 1
456 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x2__scalar_fmagic
458 k-block: 1
459 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x2__scalar_imagic
461 k-block: 1
462 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x2__scalar_lrintf
464 k-block: 1
465 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x4__scalar_fmagic
467 k-block: 1
468 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x4__scalar_imagic
470 k-block: 1
471 - name: xnn_qu8_igemm_minmax_fp32_ukernel_1x4__scalar_lrintf
473 k-block: 1
474 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x2__scalar_fmagic
476 k-block: 1
477 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x2__scalar_imagic
479 k-block: 1
480 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x2__scalar_lrintf
482 k-block: 1
483 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x4__scalar_fmagic
485 k-block: 1
486 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x4__scalar_imagic
488 k-block: 1
489 - name: xnn_qu8_igemm_minmax_fp32_ukernel_2x4__scalar_lrintf
491 k-block: 1
492 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x2__scalar_fmagic
494 k-block: 1
495 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x2__scalar_imagic
497 k-block: 1
498 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x2__scalar_lrintf
500 k-block: 1
501 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x4__scalar_fmagic
503 k-block: 1
504 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x4__scalar_imagic
506 k-block: 1
507 - name: xnn_qu8_igemm_minmax_fp32_ukernel_3x4__scalar_lrintf
509 k-block: 1
510 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x2__scalar_fmagic
512 k-block: 1
513 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x2__scalar_imagic
515 k-block: 1
516 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x2__scalar_lrintf
518 k-block: 1
519 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x4__scalar_fmagic
521 k-block: 1
522 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x4__scalar_imagic
524 k-block: 1
525 - name: xnn_qu8_igemm_minmax_fp32_ukernel_4x4__scalar_lrintf
527 k-block: 1