AVX2/Neon: Add native `pointwise` and `pointwise_acc native` #468

jakemas · 2025-09-19T06:26:46Z

Resolves AVX2: Add poly_pointwise_montgomery assembly #337
Resolves Add Neon polyvecl_pointwise_acc_montgomery #257
Resolves AArch64: Add poly_pointwise_montgomery assembly #336
Resolves Add Neon poly_pointwise_montgomery #207
Resolves Achieve competitive performance on x86_64 #303

github-actions

Mac Mini (M1, 2020) benchmarks (opt)

Benchmark suite	Current: `216176f`	Previous: `abf8281`	Ratio
`ML-DSA-44 keypair`	`46222` cycles	`47882` cycles	`0.97`
`ML-DSA-44 sign`	`132581` cycles	`151579` cycles	`0.87`
`ML-DSA-44 verify`	`47863` cycles	`50930` cycles	`0.94`
`ML-DSA-65 keypair`	`81125` cycles	`83679` cycles	`0.97`
`ML-DSA-65 sign`	`219258` cycles	`249921` cycles	`0.88`
`ML-DSA-65 verify`	`80130` cycles	`84582` cycles	`0.95`
`ML-DSA-87 keypair`	`132284` cycles	`135655` cycles	`0.98`
`ML-DSA-87 sign`	`280799` cycles	`314509` cycles	`0.89`
`ML-DSA-87 verify`	`130347` cycles	`136425` cycles	`0.96`

This comment was automatically generated by workflow using github-action-benchmark.

github-actions

Mac Mini (M1, 2020) benchmarks (no-opt)

Benchmark suite	Current: `216176f`	Previous: `abf8281`	Ratio
`ML-DSA-44 keypair`	`115046` cycles	`115049` cycles	`1.00`
`ML-DSA-44 sign`	`431544` cycles	`431559` cycles	`1.00`
`ML-DSA-44 verify`	`122135` cycles	`122151` cycles	`1.00`
`ML-DSA-65 keypair`	`197045` cycles	`197046` cycles	`1.00`
`ML-DSA-65 sign`	`701180` cycles	`701106` cycles	`1.00`
`ML-DSA-65 verify`	`197639` cycles	`197624` cycles	`1.00`
`ML-DSA-87 keypair`	`325309` cycles	`325321` cycles	`1.00`
`ML-DSA-87 sign`	`884623` cycles	`884597` cycles	`1.00`
`ML-DSA-87 verify`	`328854` cycles	`328981` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

github-actions

Arm Cortex-A76 (Raspberry Pi 5) benchmarks (opt)

Benchmark suite	Current: `216176f`	Previous: `abf8281`	Ratio
`ML-DSA-44 keypair`	`115170` cycles	`115355` cycles	`1.00`
`ML-DSA-44 sign`	`377460` cycles	`380513` cycles	`0.99`
`ML-DSA-44 verify`	`120403` cycles	`120663` cycles	`1.00`
`ML-DSA-65 keypair`	`199241` cycles	`199958` cycles	`1.00`
`ML-DSA-65 sign`	`623626` cycles	`630974` cycles	`0.99`
`ML-DSA-65 verify`	`198418` cycles	`199181` cycles	`1.00`
`ML-DSA-87 keypair`	`325995` cycles	`327118` cycles	`1.00`
`ML-DSA-87 sign`	`792017` cycles	`797677` cycles	`0.99`
`ML-DSA-87 verify`	`324893` cycles	`326175` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

github-actions

Arm Cortex-A76 (Raspberry Pi 5) benchmarks (no-opt)

Benchmark suite	Current: `216176f`	Previous: `abf8281`	Ratio
`ML-DSA-44 keypair`	`213153` cycles	`213177` cycles	`1.00`
`ML-DSA-44 sign`	`780479` cycles	`781343` cycles	`1.00`
`ML-DSA-44 verify`	`229843` cycles	`230299` cycles	`1.00`
`ML-DSA-65 keypair`	`380942` cycles	`380958` cycles	`1.00`
`ML-DSA-65 sign`	`1303904` cycles	`1291486` cycles	`1.01`
`ML-DSA-65 verify`	`372619` cycles	`372866` cycles	`1.00`
`ML-DSA-87 keypair`	`609714` cycles	`608974` cycles	`1.00`
`ML-DSA-87 sign`	`1641842` cycles	`1641956` cycles	`1.00`
`ML-DSA-87 verify`	`621880` cycles	`621291` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

Intel Xeon 4th gen (c7i)

Benchmark suite	Current: `216176f`	Previous: `abf8281`	Ratio
`ML-DSA-44 keypair`	`35330` cycles	`35216` cycles	`1.00`
`ML-DSA-44 sign`	`120523` cycles	`125138` cycles	`0.96`
`ML-DSA-44 verify`	`38328` cycles	`39121` cycles	`0.98`
`ML-DSA-65 keypair`	`62840` cycles	`63454` cycles	`0.99`
`ML-DSA-65 sign`	`202193` cycles	`208653` cycles	`0.97`
`ML-DSA-65 verify`	`62950` cycles	`64036` cycles	`0.98`
`ML-DSA-87 keypair`	`94928` cycles	`96943` cycles	`0.98`
`ML-DSA-87 sign`	`234690` cycles	`251543` cycles	`0.93`
`ML-DSA-87 verify`	`95237` cycles	`96969` cycles	`0.98`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

Intel Xeon 4th gen (c7i) (no-opt)

Benchmark suite	Current: `216176f`	Previous: `abf8281`	Ratio
`ML-DSA-44 keypair`	`95724` cycles	`95709` cycles	`1.00`
`ML-DSA-44 sign`	`345661` cycles	`345734` cycles	`1.00`
`ML-DSA-44 verify`	`101364` cycles	`101338` cycles	`1.00`
`ML-DSA-65 keypair`	`165054` cycles	`164680` cycles	`1.00`
`ML-DSA-65 sign`	`568432` cycles	`568266` cycles	`1.00`
`ML-DSA-65 verify`	`165543` cycles	`165518` cycles	`1.00`
`ML-DSA-87 keypair`	`270522` cycles	`270264` cycles	`1.00`
`ML-DSA-87 sign`	`725298` cycles	`725295` cycles	`1.00`
`ML-DSA-87 verify`	`274135` cycles	`273484` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

Intel Xeon 3rd gen (c6i)

Benchmark suite	Current: `216176f`	Previous: `abf8281`	Ratio
`ML-DSA-44 keypair`	`56763` cycles	`57825` cycles	`0.98`
`ML-DSA-44 sign`	`180516` cycles	`190230` cycles	`0.95`
`ML-DSA-44 verify`	`61258` cycles	`63174` cycles	`0.97`
`ML-DSA-65 keypair`	`100306` cycles	`102038` cycles	`0.98`
`ML-DSA-65 sign`	`297435` cycles	`317731` cycles	`0.94`
`ML-DSA-65 verify`	`100556` cycles	`104176` cycles	`0.97`
`ML-DSA-87 keypair`	`153797` cycles	`157788` cycles	`0.97`
`ML-DSA-87 sign`	`353873` cycles	`378266` cycles	`0.94`
`ML-DSA-87 verify`	`152631` cycles	`158668` cycles	`0.96`

This comment was automatically generated by workflow using github-action-benchmark.

github-actions

Arm Cortex-A55 (Snapdragon 888) benchmarks (opt)

Benchmark suite	Current: `216176f`	Previous: `abf8281`	Ratio
`ML-DSA-44 keypair`	`287596` cycles	`288573` cycles	`1.00`
`ML-DSA-44 sign`	`928120` cycles	`930921` cycles	`1.00`
`ML-DSA-44 verify`	`290687` cycles	`295012` cycles	`0.99`
`ML-DSA-65 keypair`	`488563` cycles	`490626` cycles	`1.00`
`ML-DSA-65 sign`	`1513930` cycles	`1551572` cycles	`0.98`
`ML-DSA-65 verify`	`476293` cycles	`481319` cycles	`0.99`
`ML-DSA-87 keypair`	`841378` cycles	`833327` cycles	`1.01`
`ML-DSA-87 sign`	`2065613` cycles	`2076843` cycles	`0.99`
`ML-DSA-87 verify`	`821369` cycles	`817076` cycles	`1.01`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

AMD EPYC 3rd gen (c6a)

Benchmark suite	Current: `216176f`	Previous: `abf8281`	Ratio
`ML-DSA-44 keypair`	`69182` cycles	`72965` cycles	`0.95`
`ML-DSA-44 sign`	`185561` cycles	`209026` cycles	`0.89`
`ML-DSA-44 verify`	`69174` cycles	`74357` cycles	`0.93`
`ML-DSA-65 keypair`	`119498` cycles	`122701` cycles	`0.97`
`ML-DSA-65 sign`	`296122` cycles	`330324` cycles	`0.90`
`ML-DSA-65 verify`	`115326` cycles	`121320` cycles	`0.95`
`ML-DSA-87 keypair`	`205751` cycles	`208588` cycles	`0.99`
`ML-DSA-87 sign`	`392231` cycles	`432454` cycles	`0.91`
`ML-DSA-87 verify`	`196117` cycles	`203545` cycles	`0.96`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

Graviton4

Benchmark suite	Current: `216176f`	Previous: `abf8281`	Ratio
`ML-DSA-44 keypair`	`69390` cycles	`69452` cycles	`1.00`
`ML-DSA-44 sign`	`214041` cycles	`214770` cycles	`1.00`
`ML-DSA-44 verify`	`72664` cycles	`72508` cycles	`1.00`
`ML-DSA-65 keypair`	`123061` cycles	`122780` cycles	`1.00`
`ML-DSA-65 sign`	`351927` cycles	`353195` cycles	`1.00`
`ML-DSA-65 verify`	`120662` cycles	`120430` cycles	`1.00`
`ML-DSA-87 keypair`	`201101` cycles	`200488` cycles	`1.00`
`ML-DSA-87 sign`	`450403` cycles	`451124` cycles	`1.00`
`ML-DSA-87 verify`	`197903` cycles	`198358` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

Intel Xeon 3rd gen (c6i) (no-opt)

Benchmark suite	Current: `216176f`	Previous: `abf8281`	Ratio
`ML-DSA-44 keypair`	`157843` cycles	`157874` cycles	`1.00`
`ML-DSA-44 sign`	`564586` cycles	`563418` cycles	`1.00`
`ML-DSA-44 verify`	`169419` cycles	`169267` cycles	`1.00`
`ML-DSA-65 keypair`	`269879` cycles	`269343` cycles	`1.00`
`ML-DSA-65 sign`	`926876` cycles	`928710` cycles	`1.00`
`ML-DSA-65 verify`	`274388` cycles	`274926` cycles	`1.00`
`ML-DSA-87 keypair`	`449964` cycles	`450143` cycles	`1.00`
`ML-DSA-87 sign`	`1177029` cycles	`1177838` cycles	`1.00`
`ML-DSA-87 verify`	`459130` cycles	`458629` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

AMD EPYC 3rd gen (c6a) (no-opt)

Benchmark suite	Current: `216176f`	Previous: `abf8281`	Ratio
`ML-DSA-44 keypair`	`136077` cycles	`135835` cycles	`1.00`
`ML-DSA-44 sign`	`543045` cycles	`543662` cycles	`1.00`
`ML-DSA-44 verify`	`148907` cycles	`148551` cycles	`1.00`
`ML-DSA-65 keypair`	`227365` cycles	`227293` cycles	`1.00`
`ML-DSA-65 sign`	`882082` cycles	`880495` cycles	`1.00`
`ML-DSA-65 verify`	`235758` cycles	`235973` cycles	`1.00`
`ML-DSA-87 keypair`	`376752` cycles	`376279` cycles	`1.00`
`ML-DSA-87 sign`	`1103368` cycles	`1099997` cycles	`1.00`
`ML-DSA-87 verify`	`389430` cycles	`388895` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

AMD EPYC 4th gen (c7a)

Benchmark suite	Current: `216176f`	Previous: `abf8281`	Ratio
`ML-DSA-44 keypair`	`42196` cycles	`42495` cycles	`0.99`
`ML-DSA-44 sign`	`132313` cycles	`136953` cycles	`0.97`
`ML-DSA-44 verify`	`44148` cycles	`45651` cycles	`0.97`
`ML-DSA-65 keypair`	`73634` cycles	`73514` cycles	`1.00`
`ML-DSA-65 sign`	`212291` cycles	`222961` cycles	`0.95`
`ML-DSA-65 verify`	`73489` cycles	`75361` cycles	`0.98`
`ML-DSA-87 keypair`	`109492` cycles	`111831` cycles	`0.98`
`ML-DSA-87 sign`	`249171` cycles	`264029` cycles	`0.94`
`ML-DSA-87 verify`	`109508` cycles	`113897` cycles	`0.96`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

⚠️ Performance Alert ⚠️

Possible performance regression was detected for benchmark 'AMD EPYC 4th gen (c7a)'.
Benchmark result of this commit is worse than the previous benchmark result exceeding threshold 1.03.

Benchmark suite	Current: `0241a1e`	Previous: `a50c7c5`	Ratio
`ML-DSA-65 keypair`	`74894` cycles	`72607` cycles	`1.03`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

Graviton4 (no-opt)

Benchmark suite	Current: `216176f`	Previous: `abf8281`	Ratio
`ML-DSA-44 keypair`	`132632` cycles	`132765` cycles	`1.00`
`ML-DSA-44 sign`	`498393` cycles	`498360` cycles	`1.00`
`ML-DSA-44 verify`	`144871` cycles	`144978` cycles	`1.00`
`ML-DSA-65 keypair`	`227282` cycles	`227374` cycles	`1.00`
`ML-DSA-65 sign`	`814627` cycles	`813162` cycles	`1.00`
`ML-DSA-65 verify`	`231687` cycles	`231727` cycles	`1.00`
`ML-DSA-87 keypair`	`374382` cycles	`374649` cycles	`1.00`
`ML-DSA-87 sign`	`1021450` cycles	`1021467` cycles	`1.00`
`ML-DSA-87 verify`	`383912` cycles	`383727` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

Graviton2

Benchmark suite	Current: `216176f`	Previous: `abf8281`	Ratio
`ML-DSA-44 keypair`	`115227` cycles	`116321` cycles	`0.99`
`ML-DSA-44 sign`	`377680` cycles	`382918` cycles	`0.99`
`ML-DSA-44 verify`	`120445` cycles	`121647` cycles	`0.99`
`ML-DSA-65 keypair`	`199386` cycles	`200280` cycles	`1.00`
`ML-DSA-65 sign`	`623881` cycles	`631576` cycles	`0.99`
`ML-DSA-65 verify`	`198543` cycles	`199420` cycles	`1.00`
`ML-DSA-87 keypair`	`326380` cycles	`327953` cycles	`1.00`
`ML-DSA-87 sign`	`792806` cycles	`799114` cycles	`0.99`
`ML-DSA-87 verify`	`325168` cycles	`326610` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

Graviton3

Benchmark suite	Current: `216176f`	Previous: `abf8281`	Ratio
`ML-DSA-44 keypair`	`73683` cycles	`73940` cycles	`1.00`
`ML-DSA-44 sign`	`227417` cycles	`228396` cycles	`1.00`
`ML-DSA-44 verify`	`77957` cycles	`78071` cycles	`1.00`
`ML-DSA-65 keypair`	`129692` cycles	`129923` cycles	`1.00`
`ML-DSA-65 sign`	`376961` cycles	`377186` cycles	`1.00`
`ML-DSA-65 verify`	`128989` cycles	`129040` cycles	`1.00`
`ML-DSA-87 keypair`	`208475` cycles	`210651` cycles	`0.99`
`ML-DSA-87 sign`	`478022` cycles	`478561` cycles	`1.00`
`ML-DSA-87 verify`	`208461` cycles	`210198` cycles	`0.99`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

AMD EPYC 4th gen (c7a) (no-opt)

Benchmark suite	Current: `216176f`	Previous: `abf8281`	Ratio
`ML-DSA-44 keypair`	`120750` cycles	`120120` cycles	`1.01`
`ML-DSA-44 sign`	`452766` cycles	`453371` cycles	`1.00`
`ML-DSA-44 verify`	`131683` cycles	`132326` cycles	`1.00`
`ML-DSA-65 keypair`	`204985` cycles	`204716` cycles	`1.00`
`ML-DSA-65 sign`	`737151` cycles	`737570` cycles	`1.00`
`ML-DSA-65 verify`	`210237` cycles	`210009` cycles	`1.00`
`ML-DSA-87 keypair`	`338901` cycles	`338444` cycles	`1.00`
`ML-DSA-87 sign`	`941134` cycles	`941628` cycles	`1.00`
`ML-DSA-87 verify`	`348225` cycles	`349377` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

Graviton3 (no-opt)

Benchmark suite	Current: `216176f`	Previous: `abf8281`	Ratio
`ML-DSA-44 keypair`	`138577` cycles	`138584` cycles	`1.00`
`ML-DSA-44 sign`	`494733` cycles	`495481` cycles	`1.00`
`ML-DSA-44 verify`	`148719` cycles	`148760` cycles	`1.00`
`ML-DSA-65 keypair`	`241622` cycles	`241312` cycles	`1.00`
`ML-DSA-65 sign`	`810227` cycles	`809760` cycles	`1.00`
`ML-DSA-65 verify`	`241033` cycles	`240909` cycles	`1.00`
`ML-DSA-87 keypair`	`396517` cycles	`396477` cycles	`1.00`
`ML-DSA-87 sign`	`1031718` cycles	`1031613` cycles	`1.00`
`ML-DSA-87 verify`	`402457` cycles	`402260` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

Graviton2 (no-opt)

Benchmark suite	Current: `216176f`	Previous: `abf8281`	Ratio
`ML-DSA-44 keypair`	`213591` cycles	`213647` cycles	`1.00`
`ML-DSA-44 sign`	`781116` cycles	`794200` cycles	`0.98`
`ML-DSA-44 verify`	`230160` cycles	`230157` cycles	`1.00`
`ML-DSA-65 keypair`	`381239` cycles	`381964` cycles	`1.00`
`ML-DSA-65 sign`	`1287102` cycles	`1286398` cycles	`1.00`
`ML-DSA-65 verify`	`373072` cycles	`373972` cycles	`1.00`
`ML-DSA-87 keypair`	`609771` cycles	`609842` cycles	`1.00`
`ML-DSA-87 sign`	`1644018` cycles	`1645519` cycles	`1.00`
`ML-DSA-87 verify`	`623636` cycles	`621691` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

github-actions

SpacemiT K1 8 (Banana Pi F3) benchmarks (no-opt)

Benchmark suite	Current: `216176f`	Previous: `abf8281`	Ratio
`ML-DSA-44 keypair`	`823292` cycles	`822607` cycles	`1.00`
`ML-DSA-44 sign`	`3331458` cycles	`3327961` cycles	`1.00`
`ML-DSA-44 verify`	`920849` cycles	`919517` cycles	`1.00`
`ML-DSA-65 keypair`	`1396888` cycles	`1395382` cycles	`1.00`
`ML-DSA-65 sign`	`5441990` cycles	`5429294` cycles	`1.00`
`ML-DSA-65 verify`	`1464465` cycles	`1462157` cycles	`1.00`
`ML-DSA-87 keypair`	`2301520` cycles	`2304525` cycles	`1.00`
`ML-DSA-87 sign`	`6814378` cycles	`6826659` cycles	`1.00`
`ML-DSA-87 verify`	`2397724` cycles	`2401170` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

github-actions

Arm Cortex-A55 (Snapdragon 888) benchmarks (no-opt)

Benchmark suite	Current: `216176f`	Previous: `abf8281`	Ratio
`ML-DSA-44 keypair`	`466206` cycles	`465321` cycles	`1.00`
`ML-DSA-44 sign`	`2215738` cycles	`2229058` cycles	`0.99`
`ML-DSA-44 verify`	`550432` cycles	`547134` cycles	`1.01`
`ML-DSA-65 keypair`	`778286` cycles	`779224` cycles	`1.00`
`ML-DSA-65 sign`	`3634433` cycles	`3638239` cycles	`1.00`
`ML-DSA-65 verify`	`849942` cycles	`850223` cycles	`1.00`
`ML-DSA-87 keypair`	`1266642` cycles	`1267615` cycles	`1.00`
`ML-DSA-87 sign`	`4543465` cycles	`4512261` cycles	`1.01`
`ML-DSA-87 verify`	`1376276` cycles	`1373672` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

github-actions

Arm Cortex-A72 (Raspberry Pi 4) benchmarks (opt)

Benchmark suite	Current: `216176f`	Previous: `abf8281`	Ratio
`ML-DSA-44 keypair`	`229216` cycles	`241131` cycles	`0.95`
`ML-DSA-44 sign`	`680980` cycles	`692486` cycles	`0.98`
`ML-DSA-44 verify`	`232699` cycles	`231043` cycles	`1.01`
`ML-DSA-65 keypair`	`386953` cycles	`393697` cycles	`0.98`
`ML-DSA-65 sign`	`1066946` cycles	`1084494` cycles	`0.98`
`ML-DSA-65 verify`	`372780` cycles	`375628` cycles	`0.99`
`ML-DSA-87 keypair`	`660337` cycles	`671942` cycles	`0.98`
`ML-DSA-87 sign`	`1450115` cycles	`1488161` cycles	`0.97`
`ML-DSA-87 verify`	`649151` cycles	`646858` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

github-actions

⚠️ Performance Alert ⚠️

Possible performance regression was detected for benchmark 'Arm Cortex-A72 (Raspberry Pi 4) benchmarks (opt)'.
Benchmark result of this commit is worse than the previous benchmark result exceeding threshold 1.03.

Benchmark suite	Current: `a41adf6`	Previous: `abf8281`	Ratio
`ML-DSA-65 keypair`	`411845` cycles	`393697` cycles	`1.05`
`ML-DSA-65 sign`	`1135899` cycles	`1084494` cycles	`1.05`
`ML-DSA-65 verify`	`389216` cycles	`375628` cycles	`1.04`

This comment was automatically generated by workflow using github-action-benchmark.

github-actions

Arm Cortex-A72 (Raspberry Pi 4) benchmarks (no-opt)

Benchmark suite	Current: `216176f`	Previous: `abf8281`	Ratio
`ML-DSA-44 keypair`	`317658` cycles	`301890` cycles	`1.05`
`ML-DSA-44 sign`	`1209526` cycles	`1200079` cycles	`1.01`
`ML-DSA-44 verify`	`341046` cycles	`332889` cycles	`1.02`
`ML-DSA-65 keypair`	`558435` cycles	`568930` cycles	`0.98`
`ML-DSA-65 sign`	`1943000` cycles	`1985832` cycles	`0.98`
`ML-DSA-65 verify`	`532271` cycles	`539286` cycles	`0.99`
`ML-DSA-87 keypair`	`866747` cycles	`863211` cycles	`1.00`
`ML-DSA-87 sign`	`2482734` cycles	`2481765` cycles	`1.00`
`ML-DSA-87 verify`	`881465` cycles	`888876` cycles	`0.99`

This comment was automatically generated by workflow using github-action-benchmark.

github-actions

⚠️ Performance Alert ⚠️

Possible performance regression was detected for benchmark 'Arm Cortex-A72 (Raspberry Pi 4) benchmarks (no-opt)'.
Benchmark result of this commit is worse than the previous benchmark result exceeding threshold 1.03.

Benchmark suite	Current: `216176f`	Previous: `abf8281`	Ratio
`ML-DSA-44 keypair`	`317658` cycles	`301890` cycles	`1.05`

This comment was automatically generated by workflow using github-action-benchmark.

mkannwischer

Thanks @jakemas and @jammychiou1! This is almost looking good to me.
Please fix the comment, and then clean up the commit history.

This will conclude #303 once merged.

mldsa/native/api.h

Co-authored-by: jammychiou1 <[email protected]> Signed-off-by: Jake Massimo <[email protected]> Signed-off-by: jammychiou1 <[email protected]>

Signed-off-by: Jake Massimo <[email protected]>

This commit adds a native implementation of poly_pointwise_montgomery written from scratch. Co-authored-by: Matthias J. Kannwischer <[email protected]> Signed-off-by: jammychiou1 <[email protected]>

This commit adds native implementations of polyvecl_pointwise_acc_montgomery written from scratch. Co-authored-by: Matthias J. Kannwischer <[email protected]> Signed-off-by: jammychiou1 <[email protected]>

CBMC: polyvec_matrix-expand OBJECT_BITS 10 -> 11 CBMC: polyvecl_pointwise_acc_montgomery OBJECT_BITS 10->12 CBMC: polyvecl_chknorm OBJECT_BITS 9->12 Signed-off-by: Matthias J. Kannwischer <[email protected]>

mkannwischer

Thanks for all the work @jakemas and @jammychiou1!
Looks good to me!

jakemas added the benchmark label Sep 19, 2025

github-actions bot reviewed Sep 19, 2025

View reviewed changes

oqs-bot reviewed Sep 19, 2025

View reviewed changes

github-actions bot reviewed Sep 19, 2025

View reviewed changes

oqs-bot reviewed Sep 19, 2025

View reviewed changes

github-actions bot reviewed Sep 19, 2025

View reviewed changes

mkannwischer force-pushed the avx2-pointwise-acc branch from 816f665 to 1fda1e9 Compare September 22, 2025 04:48

jakemas marked this pull request as ready for review September 22, 2025 17:09

jakemas requested a review from a team as a code owner September 22, 2025 17:09

jammychiou1 force-pushed the avx2-pointwise-acc branch from 38156be to cdf39ee Compare October 12, 2025 17:23

jammychiou1 added benchmark and removed benchmark labels Oct 12, 2025

jammychiou1 force-pushed the avx2-pointwise-acc branch from cdf39ee to a2e40c1 Compare October 13, 2025 02:05

jammychiou1 added benchmark and removed benchmark labels Oct 13, 2025

jammychiou1 mentioned this pull request Oct 20, 2025

Achieve competitive performance on x86_64 #303

Closed

mkannwischer force-pushed the avx2-pointwise-acc branch 2 times, most recently from 5276218 to 601f653 Compare October 22, 2025 04:02

mkannwischer requested changes Oct 22, 2025

View reviewed changes

mldsa/native/api.h Outdated Show resolved Hide resolved

mkannwischer marked this pull request as ready for review October 22, 2025 04:50

jakemas and others added 5 commits October 22, 2025 14:38

AVX2: Add native implementation of poly_pointwise_montgomery

d7bb741

Co-authored-by: jammychiou1 <[email protected]> Signed-off-by: Jake Massimo <[email protected]> Signed-off-by: jammychiou1 <[email protected]>

AVX2: Add native implementations of polyvecl_pointwise_acc_montgomery

640f30b

Co-authored-by: jammychiou1 <[email protected]> Signed-off-by: Jake Massimo <[email protected]> Signed-off-by: jammychiou1 <[email protected]>

added pointwise to bench_components

88bd6d5

Signed-off-by: Jake Massimo <[email protected]>

AArch64: Add native implementation of poly_pointwise_montgomery

73833d6

This commit adds a native implementation of poly_pointwise_montgomery written from scratch. Co-authored-by: Matthias J. Kannwischer <[email protected]> Signed-off-by: jammychiou1 <[email protected]>

AArch64: Add native implementations of polyvecl_pointwise_acc_montgomery

a61b5e0

This commit adds native implementations of polyvecl_pointwise_acc_montgomery written from scratch. Co-authored-by: Matthias J. Kannwischer <[email protected]> Signed-off-by: jammychiou1 <[email protected]>

jammychiou1 force-pushed the avx2-pointwise-acc branch from 601f653 to 78a1305 Compare October 22, 2025 06:41

mkannwischer changed the title ~~AVX2: Add pointwise and pointwise_acc native backend~~ AVX2/Neon: Add native pointwise and pointwise_acc native Oct 22, 2025

CBMC: Increase OBJECT_BITS for various proofs

216176f

CBMC: polyvec_matrix-expand OBJECT_BITS 10 -> 11 CBMC: polyvecl_pointwise_acc_montgomery OBJECT_BITS 10->12 CBMC: polyvecl_chknorm OBJECT_BITS 9->12 Signed-off-by: Matthias J. Kannwischer <[email protected]>

mkannwischer force-pushed the avx2-pointwise-acc branch from 78a1305 to 216176f Compare October 22, 2025 07:02

mkannwischer approved these changes Oct 22, 2025

View reviewed changes

mkannwischer added benchmark and removed benchmark labels Oct 22, 2025

mkannwischer merged commit a74d2db into main Oct 23, 2025
276 checks passed

mkannwischer deleted the avx2-pointwise-acc branch October 23, 2025 01:30

This was referenced Oct 23, 2025

Add Neon mld_polyvecl_pointwise_acc_montgomery_l{4,5,7}_native #281

Closed

AVX2/AArch64: Add native poly_pointwise_montgomery #346

Closed

CBMC proof performance regression #547

Open

Uh oh!

AVX2/Neon: Add native pointwise and pointwise_acc native #468

AVX2/Neon: Add native pointwise and pointwise_acc native #468

Uh oh!

Conversation

jakemas commented Sep 19, 2025 • edited by mkannwischer Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

github-actions bot left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Mac Mini (M1, 2020) benchmarks (opt)

Uh oh!

github-actions bot left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Mac Mini (M1, 2020) benchmarks (no-opt)

Uh oh!

github-actions bot left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Arm Cortex-A76 (Raspberry Pi 5) benchmarks (opt)

Uh oh!

github-actions bot left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Arm Cortex-A76 (Raspberry Pi 5) benchmarks (no-opt)

Uh oh!

oqs-bot left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Intel Xeon 4th gen (c7i)

Uh oh!

oqs-bot left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Intel Xeon 4th gen (c7i) (no-opt)

Uh oh!

oqs-bot left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Intel Xeon 3rd gen (c6i)

Uh oh!

github-actions bot left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Arm Cortex-A55 (Snapdragon 888) benchmarks (opt)

Uh oh!

oqs-bot left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

AMD EPYC 3rd gen (c6a)

Uh oh!

oqs-bot left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Graviton4

Uh oh!

oqs-bot left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Intel Xeon 3rd gen (c6i) (no-opt)

Uh oh!

oqs-bot left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

AMD EPYC 3rd gen (c6a) (no-opt)

Uh oh!

oqs-bot left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

AMD EPYC 4th gen (c7a)

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

⚠️ Performance Alert ⚠️

Uh oh!

oqs-bot left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

AVX2/Neon: Add native `pointwise` and `pointwise_acc native` #468

AVX2/Neon: Add native `pointwise` and `pointwise_acc native` #468

jakemas commented Sep 19, 2025 •

edited by mkannwischer

Loading

github-actions bot left a comment •

edited

Loading

github-actions bot left a comment •

edited

Loading

github-actions bot left a comment •

edited

Loading

github-actions bot left a comment •

edited

Loading

oqs-bot left a comment •

edited

Loading

oqs-bot left a comment •

edited

Loading

oqs-bot left a comment •

edited

Loading

github-actions bot left a comment •

edited

Loading

oqs-bot left a comment •

edited

Loading

oqs-bot left a comment •

edited

Loading

oqs-bot left a comment •

edited

Loading

oqs-bot left a comment •

edited

Loading

oqs-bot left a comment •

edited

Loading

oqs-bot left a comment •

edited

Loading

oqs-bot left a comment •

edited

Loading

oqs-bot left a comment •

edited

Loading

oqs-bot left a comment •

edited

Loading

oqs-bot left a comment •

edited

Loading

oqs-bot left a comment •

edited

Loading

github-actions bot left a comment •

edited

Loading

github-actions bot left a comment •

edited

Loading

github-actions bot left a comment •

edited

Loading

github-actions bot left a comment •

edited

Loading

github-actions bot left a comment •

edited

Loading

github-actions bot left a comment •

edited

Loading