Skip to content

Instantly share code, notes, and snippets.

@travisdowns

travisdowns/hsl_vs_skl_decode.txt

Created March 15, 2019 03:58
Show Gist options
  • Star 1 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save travisdowns/8b1d8d806cffa2a86cfabc744d4cf059 to your computer and use it in GitHub Desktop.
Save travisdowns/8b1d8d806cffa2a86cfabc744d4cf059 to your computer and use it in GitHub Desktop.
Download ZIP
Raw
hsl_vs_skl_decode.txt
Results of HSW vs SKL decode tests, for different patterns of multi-byte nops.
Tests run with:
./uarch-bench.sh --timer=libpfc --test-name=misc/decode* --extra-events=inst_retired,lsd.uops,idq.dsb_uops,idq.mite_uops --precision=3
Haswell (i7-4770) Results:
** Running group misc : Miscellaneous tests **
Benchmark Cycles INST_R LSD:UO IDQ:DS IDQ:MI
Decode 3-3-3-3-4 byte nops 0.252 1.000 0.000 0.000 1.000
Decode 3-3-3-3-3 byte nops 0.313 1.000 0.000 0.000 1.000
Decode 16x1 byte nops 0.250 1.000 0.000 0.000 1.000
Decode 8x2 byte nops 0.251 1.000 0.000 0.000 1.000
Decode 4x4 byte nops 0.254 1.000 0.000 0.000 1.000
Decode 6-6-4 byte nops 0.337 1.000 0.000 0.000 1.000
Decode 8-8 byte nops 0.505 1.000 0.000 0.000 1.000
Decode 8-8-3-3-3-3-4 byte nops 0.290 1.000 0.000 0.000 1.000
Decode 8-8-4-4-4-4 byte nops 0.337 1.000 0.000 0.000 1.000
Skylake (i7-6700HQ) results:
** Running group misc : Miscellaneous tests **
Benchmark Cycles INST_R LSD:UO IDQ:DS IDQ:MI
Decode 3-3-3-3-4 byte nops 0.251 1.000 0.000 0.000 1.000
Decode 3-3-3-3-3 byte nops 0.313 1.000 0.000 0.000 1.000
Decode 16x1 byte nops 0.250 1.000 0.000 0.000 1.000
Decode 8x2 byte nops 0.251 1.000 0.000 0.000 1.000
Decode 4x4 byte nops 0.252 1.000 0.000 0.000 1.000
Decode 6-6-4 byte nops 0.338 1.000 0.000 0.000 1.000
Decode 8-8 byte nops 0.514 1.000 0.000 0.000 1.000
Decode 8-8-3-3-3-3-4 byte nops 0.290 1.000 0.000 0.000 1.000
Decode 8-8-4-4-4-4 byte nops 0.334 1.000 0.000 0.000 1.000
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment