mirrors/tinygrad
2
0
Fork 0
mirror of https://github.com/tinygrad/tinygrad.git synced 2026-06-24 02:14:17 +00:00
Code Issues Projects Releases Packages Wiki Activity
tinygrad/extra/sqtt
History
Exact
qazal 8c174bdad4
 viz/sqtt: correct exec pipes (#15885) 
* wmma

* p2

* test

* left

* work

* pickle

* handwritten failing tests

* start work

* test the pipes

* empirical evidence

* update rdna4 enum types

* VALU pipe 1

* TRANSCENDENTAL pipe

* transcendental function units

* reorder

* wmma pipe

* cleanup and notes

* smaller

* work

* diff cleanup

* pickle

* use se:1

* int
2026-04-28 05:05:49 +09:00
..
examples viz/sqtt: correct exec pipes (#15885) 2026-04-28 05:05:49 +09:00
install_rocprof_decoder.py sqtt: match amd's TS_DELTA_SHORT offset (#15901) 2026-04-24 06:41:22 +03:00
README.md sqtt: remove old files (#15108) 2026-03-03 22:43:24 +09:00
rgptool.py diff devices for sdma (#14589) 2026-02-06 16:39:12 +03:00
roc.py viz: no global state (#15705) 2026-04-13 21:35:20 +09:00
sqtt.h amd: sqtt on gfx12 (#12564) 2025-10-10 17:54:14 +08:00

README.md

SQTT Profiling

Getting SQ Thread Trace

VIZ=2 to enable SQTT profiling.

SQTT_ITRACE_SE_MASK=X to select shader engines for instruction tracing, -1 = all, 0 = disabled, >0 = SE bitmask, default 0b11.

SQTT_BUFFER_SIZE=X to change size of SQTT buffer (per shader engine, 6 SEs on 7900xtx) in megabytes, default 256.

Viewing the traces

  • Web UI: tinygrad/viz/serve.py
  • Command line: python -m tinygrad.renderer.amd.sqtt