362 commits

Author	SHA1	Message	Date
George Hotz	f3cb4c3eef	oops	2025-04-01 23:44:44 +08:00
George Hotz	6ecaf11224	ugh many hacks	2025-04-01 23:33:09 +08:00
George Hotz	8b24f9cb0d	oops, didn't mean to change that	2025-04-01 17:55:04 +08:00
George Hotz	797e512c00	all correct	2025-04-01 17:51:24 +08:00
George Hotz	f600482982	correctness	2025-04-01 17:27:16 +08:00
George Hotz	da35edbb55	reenable that upcast	2025-04-01 17:09:02 +08:00
George Hotz	8340d9c1c2	disable padding	2025-04-01 16:27:54 +08:00
George Hotz	e6e0c0ec86	should work	2025-04-01 15:25:15 +08:00
George Hotz	eb606d7230	MULTICORE=1 PYTHONPATH=. QUANTIZE=1 DEBUG=2 DEVECTORIZE=0 python3 extra/replay_pkl.py /tmp/im.pkl	2025-03-31 15:37:07 +08:00
George Hotz	996d0ac1d2	multicore all the way	2025-03-31 14:17:19 +08:00
George Hotz	273dde69bd	remove range split support	2025-03-31 12:43:21 +08:00
George Hotz	9b19129e87	mc	2025-03-31 11:34:22 +08:00
George Hotz	48221d9024	2 global dim	2025-03-31 11:25:12 +08:00
George Hotz	abc90024ac	hand coded opts	2025-03-31 10:44:09 +08:00
George Hotz	e0fd84dd64	add locals	2025-03-28 18:52:48 +08:00
George Hotz	0aa7031b5f	simpler	2025-03-28 17:42:14 +08:00
George Hotz	d32ad080c3	fast 66	2025-03-27 16:47:58 +08:00
George Hotz	6d860389f4	issue	2025-03-27 16:21:02 +08:00
George Hotz	38488ec3b0	extend to 128	2025-03-27 10:35:06 +08:00
George Hotz	ff96f0adae	7	2025-03-27 00:29:00 +08:00
George Hotz	5dd59a6096	touchup	2025-03-27 00:23:58 +08:00
George Hotz	a436d7542f	up 7	2025-03-27 00:00:40 +08:00
George Hotz	928994c6ea	bugfix	2025-03-26 20:10:15 +08:00
George Hotz	290ba9ee37	more cleanups	2025-03-26 17:59:26 +08:00
George Hotz	8660fecb02	unroll both sides	2025-03-26 15:12:02 +08:00
George Hotz	f1ff18acec	prepad kernel weights	2025-03-26 12:13:46 +08:00
George Hotz	f6e64a5e8e	optional conv	2025-03-25 19:00:48 +08:00
George Hotz	554a490751	Merge branch 'master' into dsp_search	2025-03-24 12:29:22 +08:00
George Hotz	de7d6cec3a	hotfix: DEBUG 5 prints the ast	2025-03-24 11:43:11 +08:00
George Hotz	fd73ec2b1b	knum	2025-03-22 18:59:54 +08:00
George Hotz	e1d2bec4a4	opt	2025-03-22 18:52:56 +08:00
George Hotz	26b02a037c	fix 33	2025-03-22 17:17:47 +08:00
Ignacio Sica	eddafb84e5	Bugfix for TC=3 (#9464) ... * wrong but uses less shared * for size 8 tc1 with devectorize in 0 loads into local before wmma and works * improvements over tc1 devectorize * fix tc=3 * works for handcoded tc opts * clean bugfix tc=3 * fix * revert changes	2025-03-21 16:43:42 -07:00
chenyu	6da78164f9	assert Kernel ast.op to be Ops.SINK [pr] (#9539) ... rest of the code assumes self.ast is defined anyway	2025-03-21 18:09:44 -04:00
George Hotz	0416b0998d	revert those	2025-03-21 17:15:38 +08:00
George Hotz	5ce951fb34	l2	2025-03-21 11:14:12 +08:00
George Hotz	e5ccd9e846	work	2025-03-20 15:20:03 +08:00
George Hotz	52ae9af4dd	Fast DSP for MobileNetV2 (try 2) (#9467) ... * Fast DSP for MobileNetV2 (try 2) * enable fast path on uchar * fix tests	2025-03-17 15:10:36 +08:00
chenyu	ca5064a5b6	remove Kernel.float4_axis [pr] (#9448)	2025-03-14 17:54:32 -04:00
George Hotz	a73d8717f3	fast amd gemm (#9318) ... * 50 TFLOP AMD gemm * add lds tiling * register tiling * flip locals * work * comment * remove those	2025-03-03 12:01:14 +08:00
chenyu	2e7c2780a9	CLANG -> CPU (#9189)	2025-02-20 18:03:09 -05:00
qazal	cf315d544b	rename can_pad arg to cache [pr] (#9170)	2025-02-19 12:24:59 +01:00
George Hotz	a330f3338c	save applied opts in ProgramSpec [pr] (#9150)	2025-02-19 00:40:03 +08:00
George Hotz	ff9b985d9f	hotfix: View Base AST	2025-02-18 18:48:34 +08:00
George Hotz	ddddcc165b	colors back in DEBUG=2 [pr] (#9155)	2025-02-18 16:17:57 +08:00
George Hotz	caee42e8a6	Revert "name from uops [pr] (#9151)" (#9154) ... This reverts commit 28897be9a2.	2025-02-18 16:06:44 +08:00
George Hotz	28897be9a2	name from uops [pr] (#9151)	2025-02-18 15:52:03 +08:00
George Hotz	a4dab3ec3f	add name uop (#9149) ... * add name uop, TODO: refactor renderer to use * renderer uses name uop * fix tests * render * ptx	2025-02-18 15:26:58 +08:00
George Hotz	df3b320f46	rewriter -> devectorizer [pr] (#9147)	2025-02-18 12:42:08 +08:00
George Hotz	9289425170	add ast to ProgramSpec + pre matcher [pr] (#9128) ... * add ast to ProgramSpec + pre matcher [pr] * cleaner cast + test fix	2025-02-17 16:39:14 +08:00