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

mkannwischer · 2025-05-23T15:33:58Z

These are basically written from scratch inspired by the same functions in mlkem-native.
Resolves #257

These are basically written from scratch inspired by the same functions in mlkem-native. Resolves #257 Signed-off-by: Matthias J. Kannwischer <[email protected]>

github-actions

Mac Mini (M1, 2020) benchmarks (opt)

Benchmark suite	Current: `d42cabe`	Previous: `a1ad592`	Ratio
`ML-DSA-44 keypair`	`98316` cycles	`99918` cycles	`0.98`
`ML-DSA-44 sign`	`220684` cycles	`132053` cycles	`1.67`
`ML-DSA-44 verify`	`100497` cycles	`102131` cycles	`0.98`
`ML-DSA-65 keypair`	`173974` cycles	`175471` cycles	`0.99`
`ML-DSA-65 sign`	`353943` cycles	`363936` cycles	`0.97`
`ML-DSA-65 verify`	`164775` cycles	`167106` cycles	`0.99`
`ML-DSA-87 keypair`	`290957` cycles	`293910` cycles	`0.99`
`ML-DSA-87 sign`	`493610` cycles	`502464` cycles	`0.98`
`ML-DSA-87 verify`	`287075` cycles	`289463` 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 'Mac Mini (M1, 2020) benchmarks (opt)'.
Benchmark result of this commit is worse than the previous benchmark result exceeding threshold 1.03.

Benchmark suite	Current: `d42cabe`	Previous: `a1ad592`	Ratio
`ML-DSA-44 sign`	`220684` cycles	`132053` cycles	`1.67`

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

github-actions

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

Benchmark suite	Current: `d42cabe`	Previous: `a1ad592`	Ratio
`ML-DSA-44 keypair`	`116509` cycles	`116518` cycles	`1.00`
`ML-DSA-44 sign`	`369476` cycles	`369399` cycles	`1.00`
`ML-DSA-44 verify`	`127116` cycles	`127125` cycles	`1.00`
`ML-DSA-65 keypair`	`200043` cycles	`200041` cycles	`1.00`
`ML-DSA-65 sign`	`585254` cycles	`585208` cycles	`1.00`
`ML-DSA-65 verify`	`202091` cycles	`202110` cycles	`1.00`
`ML-DSA-87 keypair`	`325420` cycles	`325439` cycles	`1.00`
`ML-DSA-87 sign`	`751622` cycles	`751554` cycles	`1.00`
`ML-DSA-87 verify`	`334450` cycles	`334484` cycles	`1.00`

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

oqs-bot

Intel Xeon 4th gen (c7i)

Benchmark suite	Current: `d42cabe`	Previous: `a1ad592`	Ratio
`ML-DSA-44 keypair`	`98576` cycles	`98448` cycles	`1.00`
`ML-DSA-44 sign`	`282551` cycles	`283034` cycles	`1.00`
`ML-DSA-44 verify`	`103578` cycles	`103610` cycles	`1.00`
`ML-DSA-65 keypair`	`165288` cycles	`166302` cycles	`0.99`
`ML-DSA-65 sign`	`444392` cycles	`450712` cycles	`0.99`
`ML-DSA-65 verify`	`164041` cycles	`163487` cycles	`1.00`
`ML-DSA-87 keypair`	`274913` cycles	`275107` cycles	`1.00`
`ML-DSA-87 sign`	`585105` cycles	`588169` cycles	`0.99`
`ML-DSA-87 verify`	`272172` cycles	`272413` cycles	`1.00`

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

oqs-bot

Intel Xeon 3rd gen (c6i)

Benchmark suite	Current: `d42cabe`	Previous: `a1ad592`	Ratio
`ML-DSA-44 keypair`	`152473` cycles	`152817` cycles	`1.00`
`ML-DSA-44 sign`	`444867` cycles	`444730` cycles	`1.00`
`ML-DSA-44 verify`	`161573` cycles	`162011` cycles	`1.00`
`ML-DSA-65 keypair`	`254712` cycles	`254947` cycles	`1.00`
`ML-DSA-65 sign`	`691717` cycles	`691685` cycles	`1.00`
`ML-DSA-65 verify`	`254827` cycles	`254964` cycles	`1.00`
`ML-DSA-87 keypair`	`425968` cycles	`426582` cycles	`1.00`
`ML-DSA-87 sign`	`918484` cycles	`918576` cycles	`1.00`
`ML-DSA-87 verify`	`429052` cycles	`429321` cycles	`1.00`

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

oqs-bot

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

Benchmark suite	Current: `d42cabe`	Previous: `a1ad592`	Ratio
`ML-DSA-44 keypair`	`98668` cycles	`98328` cycles	`1.00`
`ML-DSA-44 sign`	`282320` cycles	`284172` cycles	`0.99`
`ML-DSA-44 verify`	`103613` cycles	`103502` cycles	`1.00`
`ML-DSA-65 keypair`	`165240` cycles	`165342` cycles	`1.00`
`ML-DSA-65 sign`	`448146` cycles	`450605` cycles	`0.99`
`ML-DSA-65 verify`	`164191` cycles	`163019` cycles	`1.01`
`ML-DSA-87 keypair`	`274350` cycles	`274337` cycles	`1.00`
`ML-DSA-87 sign`	`586899` cycles	`588869` cycles	`1.00`
`ML-DSA-87 verify`	`272324` cycles	`272426` 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: `d42cabe`	Previous: `a1ad592`	Ratio
`ML-DSA-44 keypair`	`152436` cycles	`152797` cycles	`1.00`
`ML-DSA-44 sign`	`445311` cycles	`444938` cycles	`1.00`
`ML-DSA-44 verify`	`161554` cycles	`161573` cycles	`1.00`
`ML-DSA-65 keypair`	`254784` cycles	`254877` cycles	`1.00`
`ML-DSA-65 sign`	`691153` cycles	`692396` cycles	`1.00`
`ML-DSA-65 verify`	`254889` cycles	`254956` cycles	`1.00`
`ML-DSA-87 keypair`	`426072` cycles	`426162` cycles	`1.00`
`ML-DSA-87 sign`	`915460` cycles	`918959` cycles	`1.00`
`ML-DSA-87 verify`	`429089` cycles	`429545` cycles	`1.00`

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

oqs-bot

AMD EPYC 3rd gen (c6a)

Benchmark suite	Current: `d42cabe`	Previous: `a1ad592`	Ratio
`ML-DSA-44 keypair`	`135719` cycles	`135595` cycles	`1.00`
`ML-DSA-44 sign`	`439383` cycles	`460077` cycles	`0.96`
`ML-DSA-44 verify`	`146823` cycles	`146923` cycles	`1.00`
`ML-DSA-65 keypair`	`226936` cycles	`226813` cycles	`1.00`
`ML-DSA-65 sign`	`676569` cycles	`676423` cycles	`1.00`
`ML-DSA-65 verify`	`229617` cycles	`229912` cycles	`1.00`
`ML-DSA-87 keypair`	`374443` cycles	`374438` cycles	`1.00`
`ML-DSA-87 sign`	`885712` cycles	`886683` cycles	`1.00`
`ML-DSA-87 verify`	`382329` cycles	`382591` cycles	`1.00`

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

oqs-bot

AMD EPYC 4th gen (c7a)

Benchmark suite	Current: `d42cabe`	Previous: `a1ad592`	Ratio
`ML-DSA-44 keypair`	`119953` cycles	`119880` cycles	`1.00`
`ML-DSA-44 sign`	`369392` cycles	`370516` cycles	`1.00`
`ML-DSA-44 verify`	`128403` cycles	`128235` cycles	`1.00`
`ML-DSA-65 keypair`	`200526` cycles	`200258` cycles	`1.00`
`ML-DSA-65 sign`	`561111` cycles	`561922` cycles	`1.00`
`ML-DSA-65 verify`	`200839` cycles	`200810` cycles	`1.00`
`ML-DSA-87 keypair`	`332065` cycles	`331563` cycles	`1.00`
`ML-DSA-87 sign`	`734386` cycles	`733607` cycles	`1.00`
`ML-DSA-87 verify`	`334472` cycles	`334886` 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: `d42cabe`	Previous: `a1ad592`	Ratio
`ML-DSA-44 keypair`	`135613` cycles	`135523` cycles	`1.00`
`ML-DSA-44 sign`	`436774` cycles	`437522` cycles	`1.00`
`ML-DSA-44 verify`	`146942` cycles	`146988` cycles	`1.00`
`ML-DSA-65 keypair`	`227620` cycles	`226865` cycles	`1.00`
`ML-DSA-65 sign`	`676887` cycles	`676891` cycles	`1.00`
`ML-DSA-65 verify`	`229663` cycles	`229530` cycles	`1.00`
`ML-DSA-87 keypair`	`374740` cycles	`374111` cycles	`1.00`
`ML-DSA-87 sign`	`886940` cycles	`885569` cycles	`1.00`
`ML-DSA-87 verify`	`382618` cycles	`382125` 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: `d42cabe`	Previous: `a1ad592`	Ratio
`ML-DSA-44 keypair`	`195624` cycles	`195964` cycles	`1.00`
`ML-DSA-44 sign`	`467758` cycles	`468523` cycles	`1.00`
`ML-DSA-44 verify`	`197830` cycles	`198404` cycles	`1.00`
`ML-DSA-65 keypair`	`349008` cycles	`349221` cycles	`1.00`
`ML-DSA-65 sign`	`765607` cycles	`767136` cycles	`1.00`
`ML-DSA-65 verify`	`328521` cycles	`330684` cycles	`0.99`
`ML-DSA-87 keypair`	`573483` cycles	`573746` cycles	`1.00`
`ML-DSA-87 sign`	`1040563` cycles	`1042716` cycles	`1.00`
`ML-DSA-87 verify`	`560522` cycles	`560487` cycles	`1.00`

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

oqs-bot

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

Benchmark suite	Current: `d42cabe`	Previous: `a1ad592`	Ratio
`ML-DSA-44 keypair`	`119903` cycles	`119744` cycles	`1.00`
`ML-DSA-44 sign`	`370928` cycles	`369650` cycles	`1.00`
`ML-DSA-44 verify`	`128201` cycles	`128008` cycles	`1.00`
`ML-DSA-65 keypair`	`200352` cycles	`200092` cycles	`1.00`
`ML-DSA-65 sign`	`560976` cycles	`561317` cycles	`1.00`
`ML-DSA-65 verify`	`200521` cycles	`200864` cycles	`1.00`
`ML-DSA-87 keypair`	`332274` cycles	`331646` cycles	`1.00`
`ML-DSA-87 sign`	`734263` cycles	`734232` cycles	`1.00`
`ML-DSA-87 verify`	`334732` cycles	`334633` cycles	`1.00`

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

oqs-bot

Graviton3

Benchmark suite	Current: `d42cabe`	Previous: `a1ad592`	Ratio
`ML-DSA-44 keypair`	`126828` cycles	`126641` cycles	`1.00`
`ML-DSA-44 sign`	`285023` cycles	`285808` cycles	`1.00`
`ML-DSA-44 verify`	`127622` cycles	`127747` cycles	`1.00`
`ML-DSA-65 keypair`	`219954` cycles	`220285` cycles	`1.00`
`ML-DSA-65 sign`	`466102` cycles	`467165` cycles	`1.00`
`ML-DSA-65 verify`	`210602` cycles	`211176` cycles	`1.00`
`ML-DSA-87 keypair`	`374094` cycles	`373971` cycles	`1.00`
`ML-DSA-87 sign`	`644273` cycles	`643614` cycles	`1.00`
`ML-DSA-87 verify`	`361401` cycles	`362774` cycles	`1.00`

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

oqs-bot

Graviton4

Benchmark suite	Current: `d42cabe`	Previous: `a1ad592`	Ratio
`ML-DSA-44 keypair`	`120167` cycles	`119681` cycles	`1.00`
`ML-DSA-44 sign`	`268656` cycles	`270559` cycles	`0.99`
`ML-DSA-44 verify`	`120239` cycles	`120433` cycles	`1.00`
`ML-DSA-65 keypair`	`206536` cycles	`207140` cycles	`1.00`
`ML-DSA-65 sign`	`431821` cycles	`431516` cycles	`1.00`
`ML-DSA-65 verify`	`197845` cycles	`197678` cycles	`1.00`
`ML-DSA-87 keypair`	`350679` cycles	`350590` cycles	`1.00`
`ML-DSA-87 sign`	`596343` cycles	`595201` cycles	`1.00`
`ML-DSA-87 verify`	`341605` cycles	`337811` cycles	`1.01`

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: `d42cabe`	Previous: `a1ad592`	Ratio
`ML-DSA-44 keypair`	`214247` cycles	`214664` cycles	`1.00`
`ML-DSA-44 sign`	`629037` cycles	`629503` cycles	`1.00`
`ML-DSA-44 verify`	`228850` cycles	`228902` cycles	`1.00`
`ML-DSA-65 keypair`	`375563` cycles	`375527` cycles	`1.00`
`ML-DSA-65 sign`	`1010364` cycles	`1010760` cycles	`1.00`
`ML-DSA-65 verify`	`370746` cycles	`372563` cycles	`1.00`
`ML-DSA-87 keypair`	`610482` cycles	`610537` cycles	`1.00`
`ML-DSA-87 sign`	`1327394` cycles	`1327026` cycles	`1.00`
`ML-DSA-87 verify`	`618772` cycles	`618897` cycles	`1.00`

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

oqs-bot

Graviton2

Benchmark suite	Current: `d42cabe`	Previous: `a1ad592`	Ratio
`ML-DSA-44 keypair`	`195861` cycles	`195697` cycles	`1.00`
`ML-DSA-44 sign`	`468539` cycles	`468646` cycles	`1.00`
`ML-DSA-44 verify`	`197822` cycles	`198261` cycles	`1.00`
`ML-DSA-65 keypair`	`349056` cycles	`349282` cycles	`1.00`
`ML-DSA-65 sign`	`765040` cycles	`767664` cycles	`1.00`
`ML-DSA-65 verify`	`328682` cycles	`329028` cycles	`1.00`
`ML-DSA-87 keypair`	`573927` cycles	`573783` cycles	`1.00`
`ML-DSA-87 sign`	`1042643` cycles	`1043660` cycles	`1.00`
`ML-DSA-87 verify`	`563636` cycles	`561129` cycles	`1.00`

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

oqs-bot

Graviton3 (no-opt)

Benchmark suite	Current: `d42cabe`	Previous: `a1ad592`	Ratio
`ML-DSA-44 keypair`	`139030` cycles	`138903` cycles	`1.00`
`ML-DSA-44 sign`	`392499` cycles	`391978` cycles	`1.00`
`ML-DSA-44 verify`	`146701` cycles	`146832` cycles	`1.00`
`ML-DSA-65 keypair`	`236534` cycles	`236812` cycles	`1.00`
`ML-DSA-65 sign`	`626069` cycles	`626252` cycles	`1.00`
`ML-DSA-65 verify`	`237097` cycles	`236981` cycles	`1.00`
`ML-DSA-87 keypair`	`398380` cycles	`398261` cycles	`1.00`
`ML-DSA-87 sign`	`830509` cycles	`828830` cycles	`1.00`
`ML-DSA-87 verify`	`397772` cycles	`397726` cycles	`1.00`

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

oqs-bot

Graviton4 (no-opt)

Benchmark suite	Current: `d42cabe`	Previous: `a1ad592`	Ratio
`ML-DSA-44 keypair`	`132410` cycles	`132412` cycles	`1.00`
`ML-DSA-44 sign`	`387166` cycles	`386507` cycles	`1.00`
`ML-DSA-44 verify`	`141170` cycles	`141130` cycles	`1.00`
`ML-DSA-65 keypair`	`225474` cycles	`225456` cycles	`1.00`
`ML-DSA-65 sign`	`624225` cycles	`624356` cycles	`1.00`
`ML-DSA-65 verify`	`227147` cycles	`227004` cycles	`1.00`
`ML-DSA-87 keypair`	`375628` cycles	`375457` cycles	`1.00`
`ML-DSA-87 sign`	`812950` cycles	`813146` cycles	`1.00`
`ML-DSA-87 verify`	`375666` cycles	`383996` cycles	`0.98`

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

oqs-bot

Graviton2 (no-opt)

Benchmark suite	Current: `d42cabe`	Previous: `a1ad592`	Ratio
`ML-DSA-44 keypair`	`214446` cycles	`214424` cycles	`1.00`
`ML-DSA-44 sign`	`629408` cycles	`629455` cycles	`1.00`
`ML-DSA-44 verify`	`228855` cycles	`229049` cycles	`1.00`
`ML-DSA-65 keypair`	`375509` cycles	`375590` cycles	`1.00`
`ML-DSA-65 sign`	`1011501` cycles	`1011336` cycles	`1.00`
`ML-DSA-65 verify`	`370933` cycles	`370873` cycles	`1.00`
`ML-DSA-87 keypair`	`610847` cycles	`611172` cycles	`1.00`
`ML-DSA-87 sign`	`1329471` cycles	`1330825` cycles	`1.00`
`ML-DSA-87 verify`	`619342` cycles	`619191` cycles	`1.00`

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

github-actions

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

Benchmark suite	Current: `d42cabe`	Previous: `a1ad592`	Ratio
`ML-DSA-44 keypair`	`380077` cycles	`379448` cycles	`1.00`
`ML-DSA-44 sign`	`999662` cycles	`1001117` cycles	`1.00`
`ML-DSA-44 verify`	`397396` cycles	`397880` cycles	`1.00`
`ML-DSA-65 keypair`	`657229` cycles	`658158` cycles	`1.00`
`ML-DSA-65 sign`	`1628091` cycles	`1625426` cycles	`1.00`
`ML-DSA-65 verify`	`638198` cycles	`638217` cycles	`1.00`
`ML-DSA-87 keypair`	`1099547` cycles	`1094113` cycles	`1.00`
`ML-DSA-87 sign`	`2211915` cycles	`2189695` cycles	`1.01`
`ML-DSA-87 verify`	`1089585` cycles	`1079077` cycles	`1.01`

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

github-actions

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

Benchmark suite	Current: `d42cabe`	Previous: `a1ad592`	Ratio
`ML-DSA-44 keypair`	`265297` cycles	`262456` cycles	`1.01`
`ML-DSA-44 sign`	`732281` cycles	`662663` cycles	`1.11`
`ML-DSA-44 verify`	`266153` cycles	`264971` cycles	`1.00`
`ML-DSA-65 keypair`	`494385` cycles	`492613` cycles	`1.00`
`ML-DSA-65 sign`	`1083336` cycles	`1067660` cycles	`1.01`
`ML-DSA-65 verify`	`442359` cycles	`440033` cycles	`1.01`
`ML-DSA-87 keypair`	`768058` cycles	`766459` cycles	`1.00`
`ML-DSA-87 sign`	`1462867` cycles	`1473228` cycles	`0.99`
`ML-DSA-87 verify`	`753437` cycles	`749770` 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: `d42cabe`	Previous: `a1ad592`	Ratio
`ML-DSA-44 sign`	`732281` cycles	`662663` cycles	`1.11`

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

github-actions

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

Benchmark suite	Current: `d42cabe`	Previous: `a1ad592`	Ratio
`ML-DSA-44 keypair`	`469937` cycles	`470053` cycles	`1.00`
`ML-DSA-44 sign`	`1781834` cycles	`1776230` cycles	`1.00`
`ML-DSA-44 verify`	`539638` cycles	`539372` cycles	`1.00`
`ML-DSA-65 keypair`	`784427` cycles	`784497` cycles	`1.00`
`ML-DSA-65 sign`	`2811250` cycles	`2815347` cycles	`1.00`
`ML-DSA-65 verify`	`835543` cycles	`835790` cycles	`1.00`
`ML-DSA-87 keypair`	`1271776` cycles	`1273300` cycles	`1.00`
`ML-DSA-87 sign`	`3553794` cycles	`3552014` cycles	`1.00`
`ML-DSA-87 verify`	`1348729` cycles	`1349518` cycles	`1.00`

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: `d42cabe`	Previous: `a1ad592`	Ratio
`ML-DSA-44 keypair`	`296529` cycles	`296615` cycles	`1.00`
`ML-DSA-44 sign`	`944539` cycles	`966468` cycles	`0.98`
`ML-DSA-44 verify`	`317852` cycles	`319290` cycles	`1.00`
`ML-DSA-65 keypair`	`537688` cycles	`537927` cycles	`1.00`
`ML-DSA-65 sign`	`1511337` cycles	`1510193` cycles	`1.00`
`ML-DSA-65 verify`	`514411` cycles	`514869` cycles	`1.00`
`ML-DSA-87 keypair`	`830727` cycles	`833700` cycles	`1.00`
`ML-DSA-87 sign`	`1961244` cycles	`1939495` cycles	`1.01`
`ML-DSA-87 verify`	`849787` cycles	`847955` 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 (no-opt)'.
Benchmark result of this commit is worse than the previous benchmark result exceeding threshold 1.03.

Benchmark suite	Current: `816ff86`	Previous: `a1ad592`	Ratio
`ML-DSA-44 sign`	`1041577` cycles	`966468` cycles	`1.08`

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

Signed-off-by: Matthias J. Kannwischer <[email protected]>

github-actions

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

Benchmark suite	Current: `d42cabe`	Previous: `a1ad592`	Ratio
`ML-DSA-44 keypair`	`948528` cycles	`948544` cycles	`1.00`
`ML-DSA-44 sign`	`3671235` cycles	`3669784` cycles	`1.00`
`ML-DSA-44 verify`	`1075781` cycles	`1075755` cycles	`1.00`
`ML-DSA-65 keypair`	`1563674` cycles	`1563457` cycles	`1.00`
`ML-DSA-65 sign`	`5846558` cycles	`5839467` cycles	`1.00`
`ML-DSA-65 verify`	`1699603` cycles	`1698569` cycles	`1.00`
`ML-DSA-87 keypair`	`2538731` cycles	`2539499` cycles	`1.00`
`ML-DSA-87 sign`	`7265323` cycles	`7264583` cycles	`1.00`
`ML-DSA-87 verify`	`2700701` cycles	`2701398` cycles	`1.00`

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

Add Neon mld_polyvecl_pointwise_acc_montgomery_l{4,5,7}_native

c2e3863

These are basically written from scratch inspired by the same functions in mlkem-native. Resolves #257 Signed-off-by: Matthias J. Kannwischer <[email protected]>

mkannwischer force-pushed the neon-basemul branch from aa0efb9 to c2e3863 Compare May 23, 2025 15:34

mkannwischer changed the title ~~Add Neon mld_polyvecl_pointwise_acc_montgomery_l{4,5,7}_native~~ Add Neon mld_polyvecl_pointwise_acc_montgomery_l{4,5,7}_native May 23, 2025

mkannwischer added the benchmark label May 23, 2025