Remove redundant reduction from `unpack_pk()` and `unpack_sk()` #505

hanno-becker · 2024-12-10T06:52:53Z

Resolves #480

The specifications for the polynomial base multiplication currently require one input to be bound by MLKEM_Q in absolute value. To ensure this, the unpacking routines unpack_pk() and unpack_sk() currently explicitly reduce their output via poly_reduce().

Functionally, this reduction is not needed:

For the PK, the modulus-check already ensures that the unpacked PK will be unsigned canonical.
For the SK, there is no requirement in the standard to check that the SK is reduced, or reduce it after unpacking.
There is no risk of overflow in the base multiplication even if the inputs are bound by 4096 rather than MLKEM_Q.

This commit strengthens the specifications for all polynomial base multiplication routines

basemul_cached
poly_basemul_montgomery_cached
polyvec_basemul_acc_montgomery_cached to work with the input bound 4096 instead of MLKEM_Q. Accordingly, the output bound needs weakening from 3/2 * Q to 2*Q; this could be much tighter if need be, but it's unnecessary.

The polynomial reductions in check_pk and check_sk can then be safely removed.

oqs-bot

Arm Cortex-A76 (Raspberry Pi 5) benchmarks

Benchmark suite	Current: `dab62af`	Previous: `f981699`	Ratio
`ML-KEM-512 keypair`	`29195` cycles	`29175` cycles	`1.00`
`ML-KEM-512 encaps`	`35496` cycles	`35852` cycles	`0.99`
`ML-KEM-512 decaps`	`46064` cycles	`46653` cycles	`0.99`
`ML-KEM-768 keypair`	`49187` cycles	`49176` cycles	`1.00`
`ML-KEM-768 encaps`	`55325` cycles	`55842` cycles	`0.99`
`ML-KEM-768 decaps`	`70046` cycles	`71009` cycles	`0.99`
`ML-KEM-1024 keypair`	`72112` cycles	`72213` cycles	`1.00`
`ML-KEM-1024 encaps`	`81038` cycles	`81522` cycles	`0.99`
`ML-KEM-1024 decaps`	`100783` cycles	`102032` cycles	`0.99`

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

oqs-bot

Intel Xeon 4th gen (c7i)

Benchmark suite	Current: `dab62af`	Previous: `f981699`	Ratio
`ML-KEM-512 keypair`	`13870` cycles	`13835` cycles	`1.00`
`ML-KEM-512 encaps`	`18054` cycles	`18205` cycles	`0.99`
`ML-KEM-512 decaps`	`23951` cycles	`24104` cycles	`0.99`
`ML-KEM-768 keypair`	`22450` cycles	`22551` cycles	`1.00`
`ML-KEM-768 encaps`	`24406` cycles	`24623` cycles	`0.99`
`ML-KEM-768 decaps`	`32402` cycles	`32838` cycles	`0.99`
`ML-KEM-1024 keypair`	`32043` cycles	`30856` cycles	`1.04`
`ML-KEM-1024 encaps`	`35433` cycles	`34532` cycles	`1.03`
`ML-KEM-1024 decaps`	`47063` cycles	`46135` cycles	`1.02`

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

oqs-bot

⚠️ Performance Alert ⚠️

Possible performance regression was detected for benchmark 'Intel Xeon 4th gen (c7i)'.
Benchmark result of this commit is worse than the previous benchmark result exceeding threshold 1.03.

Benchmark suite	Current: `dab62af`	Previous: `f981699`	Ratio
`ML-KEM-1024 keypair`	`32043` cycles	`30856` cycles	`1.04`

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

oqs-bot

AMD EPYC 3rd gen (c6a)

Benchmark suite	Current: `dab62af`	Previous: `f981699`	Ratio
`ML-KEM-512 keypair`	`18139` cycles	`18122` cycles	`1.00`
`ML-KEM-512 encaps`	`22898` cycles	`23102` cycles	`0.99`
`ML-KEM-512 decaps`	`30131` cycles	`30423` cycles	`0.99`
`ML-KEM-768 keypair`	`31053` cycles	`31078` cycles	`1.00`
`ML-KEM-768 encaps`	`33981` cycles	`33973` cycles	`1.00`
`ML-KEM-768 decaps`	`44635` cycles	`44794` cycles	`1.00`
`ML-KEM-1024 keypair`	`44528` cycles	`44464` cycles	`1.00`
`ML-KEM-1024 encaps`	`49885` cycles	`49908` cycles	`1.00`
`ML-KEM-1024 decaps`	`64218` cycles	`64600` cycles	`0.99`

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

oqs-bot

Intel Xeon 3rd gen (c6i)

Benchmark suite	Current: `dab62af`	Previous: `f981699`	Ratio
`ML-KEM-512 keypair`	`20311` cycles	`20339` cycles	`1.00`
`ML-KEM-512 encaps`	`26992` cycles	`27100` cycles	`1.00`
`ML-KEM-512 decaps`	`35985` cycles	`36228` cycles	`0.99`
`ML-KEM-768 keypair`	`34794` cycles	`34857` cycles	`1.00`
`ML-KEM-768 encaps`	`38042` cycles	`38334` cycles	`0.99`
`ML-KEM-768 decaps`	`51015` cycles	`51398` cycles	`0.99`
`ML-KEM-1024 keypair`	`47888` cycles	`47925` cycles	`1.00`
`ML-KEM-1024 encaps`	`53897` cycles	`54096` cycles	`1.00`
`ML-KEM-1024 decaps`	`71557` cycles	`72059` cycles	`0.99`

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

oqs-bot

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

Benchmark suite	Current: `dab62af`	Previous: `f981699`	Ratio
`ML-KEM-512 keypair`	`36013` cycles	`36184` cycles	`1.00`
`ML-KEM-512 encaps`	`44779` cycles	`46464` cycles	`0.96`
`ML-KEM-512 decaps`	`58465` cycles	`61924` cycles	`0.94`
`ML-KEM-768 keypair`	`59472` cycles	`59104` cycles	`1.01`
`ML-KEM-768 encaps`	`70549` cycles	`72980` cycles	`0.97`
`ML-KEM-768 decaps`	`88420` cycles	`91716` cycles	`0.96`
`ML-KEM-1024 keypair`	`88094` cycles	`88370` cycles	`1.00`
`ML-KEM-1024 encaps`	`104847` cycles	`109377` cycles	`0.96`
`ML-KEM-1024 decaps`	`127666` cycles	`133626` cycles	`0.96`

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

oqs-bot

Arm Cortex-A55 (Snapdragon 888) benchmarks

Benchmark suite	Current: `dab62af`	Previous: `f981699`	Ratio
`ML-KEM-512 keypair`	`58032` cycles	`58001` cycles	`1.00`
`ML-KEM-512 encaps`	`65079` cycles	`65710` cycles	`0.99`
`ML-KEM-512 decaps`	`83098` cycles	`84327` cycles	`0.99`
`ML-KEM-768 keypair`	`98170` cycles	`98213` cycles	`1.00`
`ML-KEM-768 encaps`	`109841` cycles	`110478` cycles	`0.99`
`ML-KEM-768 decaps`	`134544` cycles	`136579` cycles	`0.99`
`ML-KEM-1024 keypair`	`148719` cycles	`149118` cycles	`1.00`
`ML-KEM-1024 encaps`	`164662` cycles	`166023` cycles	`0.99`
`ML-KEM-1024 decaps`	`198996` cycles	`201670` cycles	`0.99`

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

oqs-bot

Graviton3

Benchmark suite	Current: `dab62af`	Previous: `f981699`	Ratio
`ML-KEM-512 keypair`	`19050` cycles	`19029` cycles	`1.00`
`ML-KEM-512 encaps`	`23675` cycles	`23842` cycles	`0.99`
`ML-KEM-512 decaps`	`30823` cycles	`31230` cycles	`0.99`
`ML-KEM-768 keypair`	`32341` cycles	`32329` cycles	`1.00`
`ML-KEM-768 encaps`	`35814` cycles	`36085` cycles	`0.99`
`ML-KEM-768 decaps`	`45938` cycles	`46490` cycles	`0.99`
`ML-KEM-1024 keypair`	`46947` cycles	`46967` cycles	`1.00`
`ML-KEM-1024 encaps`	`52711` cycles	`53087` cycles	`0.99`
`ML-KEM-1024 decaps`	`66631` cycles	`67402` cycles	`0.99`

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

oqs-bot

Graviton2

Benchmark suite	Current: `dab62af`	Previous: `f981699`	Ratio
`ML-KEM-512 keypair`	`29200` cycles	`29174` cycles	`1.00`
`ML-KEM-512 encaps`	`35508` cycles	`35857` cycles	`0.99`
`ML-KEM-512 decaps`	`46073` cycles	`46664` cycles	`0.99`
`ML-KEM-768 keypair`	`49210` cycles	`49189` cycles	`1.00`
`ML-KEM-768 encaps`	`55349` cycles	`55864` cycles	`0.99`
`ML-KEM-768 decaps`	`70072` cycles	`70980` cycles	`0.99`
`ML-KEM-1024 keypair`	`72184` cycles	`72212` cycles	`1.00`
`ML-KEM-1024 encaps`	`81115` cycles	`81653` cycles	`0.99`
`ML-KEM-1024 decaps`	`100910` cycles	`102097` cycles	`0.99`

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

oqs-bot

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

Benchmark suite	Current: `dab62af`	Previous: `f981699`	Ratio
`ML-KEM-512 keypair`	`52600` cycles	`52351` cycles	`1.00`
`ML-KEM-512 encaps`	`65406` cycles	`67804` cycles	`0.96`
`ML-KEM-512 decaps`	`88422` cycles	`92660` cycles	`0.95`
`ML-KEM-768 keypair`	`85134` cycles	`84684` cycles	`1.01`
`ML-KEM-768 encaps`	`100891` cycles	`104911` cycles	`0.96`
`ML-KEM-768 decaps`	`131042` cycles	`137837` cycles	`0.95`
`ML-KEM-1024 keypair`	`125029` cycles	`125494` cycles	`1.00`
`ML-KEM-1024 encaps`	`145900` cycles	`149742` cycles	`0.97`
`ML-KEM-1024 decaps`	`183516` cycles	`192272` cycles	`0.95`

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

oqs-bot

AMD EPYC 4th gen (c7a)

Benchmark suite	Current: `dab62af`	Previous: `f981699`	Ratio
`ML-KEM-512 keypair`	`15077` cycles	`15195` cycles	`0.99`
`ML-KEM-512 encaps`	`19786` cycles	`19753` cycles	`1.00`
`ML-KEM-512 decaps`	`26383` cycles	`26409` cycles	`1.00`
`ML-KEM-768 keypair`	`25601` cycles	`25693` cycles	`1.00`
`ML-KEM-768 encaps`	`28205` cycles	`28282` cycles	`1.00`
`ML-KEM-768 decaps`	`38068` cycles	`38236` cycles	`1.00`
`ML-KEM-1024 keypair`	`35525` cycles	`35523` cycles	`1.00`
`ML-KEM-1024 encaps`	`40582` cycles	`40677` cycles	`1.00`
`ML-KEM-1024 decaps`	`54288` cycles	`54488` 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: `dab62af`	Previous: `f981699`	Ratio
`ML-KEM-512 keypair`	`56574` cycles	`56728` cycles	`1.00`
`ML-KEM-512 encaps`	`69246` cycles	`71823` cycles	`0.96`
`ML-KEM-512 decaps`	`91230` cycles	`96355` cycles	`0.95`
`ML-KEM-768 keypair`	`91807` cycles	`91849` cycles	`1.00`
`ML-KEM-768 encaps`	`107745` cycles	`111503` cycles	`0.97`
`ML-KEM-768 decaps`	`136743` cycles	`144538` cycles	`0.95`
`ML-KEM-1024 keypair`	`134418` cycles	`134562` cycles	`1.00`
`ML-KEM-1024 encaps`	`155046` cycles	`159971` cycles	`0.97`
`ML-KEM-1024 decaps`	`191432` cycles	`201413` cycles	`0.95`

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

oqs-bot

Graviton3 (no-opt)

Benchmark suite	Current: `dab62af`	Previous: `f981699`	Ratio
`ML-KEM-512 keypair`	`45363` cycles	`45367` cycles	`1.00`
`ML-KEM-512 encaps`	`54218` cycles	`56174` cycles	`0.97`
`ML-KEM-512 decaps`	`71251` cycles	`75096` cycles	`0.95`
`ML-KEM-768 keypair`	`74834` cycles	`74864` cycles	`1.00`
`ML-KEM-768 encaps`	`86211` cycles	`89191` cycles	`0.97`
`ML-KEM-768 decaps`	`108714` cycles	`114550` cycles	`0.95`
`ML-KEM-1024 keypair`	`111061` cycles	`111084` cycles	`1.00`
`ML-KEM-1024 encaps`	`125973` cycles	`130004` cycles	`0.97`
`ML-KEM-1024 decaps`	`154732` cycles	`162583` cycles	`0.95`

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

oqs-bot

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

Benchmark suite	Current: `dab62af`	Previous: `f981699`	Ratio
`ML-KEM-512 keypair`	`45973` cycles	`45729` cycles	`1.01`
`ML-KEM-512 encaps`	`56925` cycles	`58795` cycles	`0.97`
`ML-KEM-512 decaps`	`76223` cycles	`80001` cycles	`0.95`
`ML-KEM-768 keypair`	`74643` cycles	`74644` cycles	`1.00`
`ML-KEM-768 encaps`	`88674` cycles	`91486` cycles	`0.97`
`ML-KEM-768 decaps`	`114563` cycles	`120140` cycles	`0.95`
`ML-KEM-1024 keypair`	`110155` cycles	`110191` cycles	`1.00`
`ML-KEM-1024 encaps`	`127240` cycles	`131275` cycles	`0.97`
`ML-KEM-1024 decaps`	`159951` cycles	`167645` cycles	`0.95`

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

oqs-bot

Graviton4

Benchmark suite	Current: `dab62af`	Previous: `f981699`	Ratio
`ML-KEM-512 keypair`	`18268` cycles	`18264` cycles	`1.00`
`ML-KEM-512 encaps`	`22311` cycles	`22478` cycles	`0.99`
`ML-KEM-512 decaps`	`29090` cycles	`29397` cycles	`0.99`
`ML-KEM-768 keypair`	`30783` cycles	`30770` cycles	`1.00`
`ML-KEM-768 encaps`	`33800` cycles	`34058` cycles	`0.99`
`ML-KEM-768 decaps`	`43358` cycles	`43907` cycles	`0.99`
`ML-KEM-1024 keypair`	`44506` cycles	`44487` cycles	`1.00`
`ML-KEM-1024 encaps`	`49856` cycles	`50207` cycles	`0.99`
`ML-KEM-1024 decaps`	`63002` cycles	`63634` cycles	`0.99`

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

oqs-bot

Graviton2 (no-opt)

Benchmark suite	Current: `dab62af`	Previous: `f981699`	Ratio
`ML-KEM-512 keypair`	`71218` cycles	`71079` cycles	`1.00`
`ML-KEM-512 encaps`	`85054` cycles	`87838` cycles	`0.97`
`ML-KEM-512 decaps`	`112664` cycles	`118157` cycles	`0.95`
`ML-KEM-768 keypair`	`117473` cycles	`117329` cycles	`1.00`
`ML-KEM-768 encaps`	`135129` cycles	`139224` cycles	`0.97`
`ML-KEM-768 decaps`	`171645` cycles	`179987` cycles	`0.95`
`ML-KEM-1024 keypair`	`174987` cycles	`175295` cycles	`1.00`
`ML-KEM-1024 encaps`	`197508` cycles	`202760` cycles	`0.97`
`ML-KEM-1024 decaps`	`243514` cycles	`254085` cycles	`0.96`

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

oqs-bot

Graviton4 (no-opt)

Benchmark suite	Current: `dab62af`	Previous: `f981699`	Ratio
`ML-KEM-512 keypair`	`41965` cycles	`41917` cycles	`1.00`
`ML-KEM-512 encaps`	`50130` cycles	`51762` cycles	`0.97`
`ML-KEM-512 decaps`	`66009` cycles	`69326` cycles	`0.95`
`ML-KEM-768 keypair`	`69094` cycles	`69060` cycles	`1.00`
`ML-KEM-768 encaps`	`79800` cycles	`82680` cycles	`0.97`
`ML-KEM-768 decaps`	`100993` cycles	`106478` cycles	`0.95`
`ML-KEM-1024 keypair`	`102224` cycles	`102236` cycles	`1.00`
`ML-KEM-1024 encaps`	`117184` cycles	`120656` cycles	`0.97`
`ML-KEM-1024 decaps`	`143349` cycles	`150644` cycles	`0.95`

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

oqs-bot

Bananapi bpi-f3 benchmarks

Benchmark suite	Current: `dab62af`	Previous: `f981699`	Ratio
`ML-KEM-512 keypair`	`368994` cycles	`369202` cycles	`1.00`
`ML-KEM-512 encaps`	`508057` cycles	`521758` cycles	`0.97`
`ML-KEM-512 decaps`	`682264` cycles	`709553` cycles	`0.96`
`ML-KEM-768 keypair`	`619098` cycles	`625015` cycles	`0.99`
`ML-KEM-768 encaps`	`798949` cycles	`826476` cycles	`0.97`
`ML-KEM-768 decaps`	`1023862` cycles	`1071745` cycles	`0.96`
`ML-KEM-1024 keypair`	`921776` cycles	`921796` cycles	`1.00`
`ML-KEM-1024 encaps`	`1145354` cycles	`1172924` cycles	`0.98`
`ML-KEM-1024 decaps`	`1420208` cycles	`1474585` cycles	`0.96`

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

oqs-bot

Arm Cortex-A72 (Raspberry Pi 4) benchmarks

Benchmark suite	Current: `dab62af`	Previous: `f981699`	Ratio
`ML-KEM-512 keypair`	`51310` cycles	`52197` cycles	`0.98`
`ML-KEM-512 encaps`	`58091` cycles	`59431` cycles	`0.98`
`ML-KEM-512 decaps`	`73729` cycles	`75460` cycles	`0.98`
`ML-KEM-768 keypair`	`86975` cycles	`87682` cycles	`0.99`
`ML-KEM-768 encaps`	`94905` cycles	`98211` cycles	`0.97`
`ML-KEM-768 decaps`	`118337` cycles	`120510` cycles	`0.98`
`ML-KEM-1024 keypair`	`131009` cycles	`131711` cycles	`0.99`
`ML-KEM-1024 encaps`	`144294` cycles	`146371` cycles	`0.99`
`ML-KEM-1024 decaps`	`175486` cycles	`178985` cycles	`0.98`

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

mlkem/polyvec.h

hanno-becker · 2024-12-10T14:17:10Z

mlkem/reduce.h

+  requires(a > -(2 * 4096 * 32768))
+  requires(a <  (2 * 4096 * 32768))


No, it's a <

mkannwischer

Thanks @hanno-becker - this is great and results in a surprisingly large performance gain on some platforms of up to 5%.

One nit.

The specifications for the polynomial base multiplication currently require one input to be bound by q in absolute value. To ensure this, the unpacking routines `unpack_pk()` and `unpack_sk()` currently explicitly reduce their output via `poly_reduce()`. Functionally, this reduction is not needed: - For the PK, the modulus-check already ensures that the unpacked PK will be unsigned canonical. - For the SK, there is no requirement in the standard to check that the SK is reduced, or reduce it after unpacking. - There is no risk of overflow in the base multiplication even if the inputs are bound by 4096 rather than q: - For the C base multiplication, the input bound before modular reduction is now 2 * 4096 * 2^15, where before it was 2 * q * 2^15 -- both is below 2^31 so we are fine. After reduction, the 3/2 * q bound is weakened to 2 * q (a much better bound would be possible, but is unneeded), which is still sufficient for up to K=4-fold accumulation since 8 * q < INT16_MAX. - For the AArch64 base multiplication, we are accumulating in 32-bit up to a single final reduction. As in the C case, each 32-bit polynomial multiplication is bound by 2 * 4096 * 2^15 = 2^28, so the accumulated value before reduction is < 2^30 < INT32_MAX. - For the x86_64 base multiplication, the reasoning is unchanged: We don't accumulate in 32-bit but conduct a Montgomery multiplication for every monomial in the product, giving an output bound by q per monomial. Up to 8 such are accumulated, giving a bound by 8 * q < INT16_MAX. This commit strengthens the specifications for all polynomial base multiplication routines - basemul_cached - poly_basemul_montgomery_cached - polyvec_basemul_acc_montgomery - polyvec_basemul_acc_montgomery_cached - matvec_mul The polynomial reductions in `check_pk` and `check_sk` can then be safely removed. Signed-off-by: Hanno Becker <[email protected]>

mkannwischer · 2024-12-10T14:51:14Z

Thanks for the change!

hanno-becker marked this pull request as ready for review December 10, 2024 06:53

hanno-becker requested a review from a team as a code owner December 10, 2024 06:53

hanno-becker force-pushed the redundant_reduce branch 5 times, most recently from 84dbfaa to dab62af Compare December 10, 2024 07:49

hanno-becker added the benchmark this PR should be benchmarked in CI label Dec 10, 2024

oqs-bot reviewed Dec 10, 2024

View reviewed changes

hanno-becker force-pushed the redundant_reduce branch 3 times, most recently from 1ea6feb to be01930 Compare December 10, 2024 10:21

mkannwischer reviewed Dec 10, 2024

View reviewed changes

mlkem/polyvec.h Outdated Show resolved Hide resolved

hanno-becker commented Dec 10, 2024

View reviewed changes

mkannwischer approved these changes Dec 10, 2024

View reviewed changes

hanno-becker force-pushed the redundant_reduce branch from be01930 to 6fc7eb4 Compare December 10, 2024 14:17

mkannwischer merged commit b0d5553 into main Dec 10, 2024
41 checks passed

mkannwischer deleted the redundant_reduce branch December 10, 2024 14:51

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Remove redundant reduction from `unpack_pk()` and `unpack_sk()` #505

Remove redundant reduction from `unpack_pk()` and `unpack_sk()` #505

hanno-becker commented Dec 10, 2024 •

edited

Loading

oqs-bot left a comment

oqs-bot left a comment

oqs-bot left a comment

oqs-bot left a comment

oqs-bot left a comment

oqs-bot left a comment

oqs-bot left a comment

oqs-bot left a comment

oqs-bot left a comment

oqs-bot left a comment

oqs-bot left a comment

oqs-bot left a comment

oqs-bot left a comment

oqs-bot left a comment

oqs-bot left a comment

oqs-bot left a comment

oqs-bot left a comment

oqs-bot left a comment

oqs-bot left a comment

hanno-becker Dec 10, 2024

mkannwischer left a comment

mkannwischer commented Dec 10, 2024

		requires(a > -(2 * 4096 * 32768))
		requires(a < (2 * 4096 * 32768))

Remove redundant reduction from unpack_pk() and unpack_sk() #505

Remove redundant reduction from unpack_pk() and unpack_sk() #505

Conversation

hanno-becker commented Dec 10, 2024 • edited Loading

oqs-bot left a comment

Choose a reason for hiding this comment

Arm Cortex-A76 (Raspberry Pi 5) benchmarks

oqs-bot left a comment

Choose a reason for hiding this comment

Intel Xeon 4th gen (c7i)

oqs-bot left a comment

Choose a reason for hiding this comment

⚠️ Performance Alert ⚠️

oqs-bot left a comment

Choose a reason for hiding this comment

AMD EPYC 3rd gen (c6a)

oqs-bot left a comment

Choose a reason for hiding this comment

Intel Xeon 3rd gen (c6i)

oqs-bot left a comment

Choose a reason for hiding this comment

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

oqs-bot left a comment

Choose a reason for hiding this comment

Arm Cortex-A55 (Snapdragon 888) benchmarks

oqs-bot left a comment

Choose a reason for hiding this comment

Graviton3

oqs-bot left a comment

Choose a reason for hiding this comment

Graviton2

oqs-bot left a comment

Choose a reason for hiding this comment

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

oqs-bot left a comment

Choose a reason for hiding this comment

AMD EPYC 4th gen (c7a)

oqs-bot left a comment

Choose a reason for hiding this comment

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

oqs-bot left a comment

Choose a reason for hiding this comment

Graviton3 (no-opt)

oqs-bot left a comment

Choose a reason for hiding this comment

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

oqs-bot left a comment

Choose a reason for hiding this comment

Graviton4

oqs-bot left a comment

Choose a reason for hiding this comment

Graviton2 (no-opt)

oqs-bot left a comment

Choose a reason for hiding this comment

Graviton4 (no-opt)

oqs-bot left a comment

Choose a reason for hiding this comment

Bananapi bpi-f3 benchmarks

oqs-bot left a comment

Choose a reason for hiding this comment

Arm Cortex-A72 (Raspberry Pi 4) benchmarks

hanno-becker Dec 10, 2024

Choose a reason for hiding this comment

mkannwischer left a comment

Choose a reason for hiding this comment

mkannwischer commented Dec 10, 2024

Remove redundant reduction from `unpack_pk()` and `unpack_sk()` #505

Remove redundant reduction from `unpack_pk()` and `unpack_sk()` #505

hanno-becker commented Dec 10, 2024 •

edited

Loading