Optimized AVX2 batched Keccak_x4

[manastasova]

This commit improves the performance of the AVX2 Keccak-F1600x4 implementation by:

• Replace gather-based loading (vpgatherqq) with explicit loads and transpose operations using unpack and permute instructions.
• Replace individual scatter stores with batched 4-lane transpose and store operations (MLK_SCATTER_STORE256_4X).
• Merged prepareTheta into thetaRhoPiChiIota at the start of each round.
• Eliminated separate E state variables by using temporaries (Tba, etc.) and copying back to A at end of each round, removing A/E alternation.
• Changed ROUNDS24 from 24 unrolled macro calls to a loop-based MLK_ROUNDS_x2 processing 2 rounds per iteration.

[oqs-bot]

Mac Mini (M1, 2020) benchmarks

Details

Benchmark suite	Current: 1647e99	Previous: ebed0f3	Ratio
ML-KEM-512 keypair	12327 cycles	12327 cycles	1
ML-KEM-512 encaps	15031 cycles	15030 cycles	1.00
ML-KEM-512 decaps	19607 cycles	19609 cycles	1.00
ML-KEM-768 keypair	21092 cycles	21092 cycles	1
ML-KEM-768 encaps	23863 cycles	23861 cycles	1.00
ML-KEM-768 decaps	30443 cycles	30442 cycles	1.00
ML-KEM-1024 keypair	30376 cycles	30376 cycles	1
ML-KEM-1024 encaps	34642 cycles	34643 cycles	1.00
ML-KEM-1024 decaps	44279 cycles	44268 cycles	1.00

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

[oqs-bot]

ppc64le (POWER10) benchmarks

Details

Benchmark suite	Current: 1647e99	Previous: ebed0f3	Ratio
ML-KEM-512 keypair	59191 cycles	59226 cycles	1.00
ML-KEM-512 encaps	71851 cycles	71908 cycles	1.00
ML-KEM-512 decaps	91623 cycles	91531 cycles	1.00
ML-KEM-768 keypair	98551 cycles	98104 cycles	1.00
ML-KEM-768 encaps	114881 cycles	114532 cycles	1.00
ML-KEM-768 decaps	140316 cycles	140016 cycles	1.00
ML-KEM-1024 keypair	148524 cycles	148872 cycles	1.00
ML-KEM-1024 encaps	167437 cycles	167765 cycles	1.00
ML-KEM-1024 decaps	198455 cycles	199095 cycles	1.00

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

[oqs-bot]

Intel Xeon 4th gen (c7i)

Details

Benchmark suite	Current: 1647e99	Previous: ebed0f3	Ratio
ML-KEM-512 keypair	9367 cycles	9650 cycles	0.97
ML-KEM-512 encaps	11028 cycles	11457 cycles	0.96
ML-KEM-512 decaps	15284 cycles	15335 cycles	1.00
ML-KEM-768 keypair	16012 cycles	16453 cycles	0.97
ML-KEM-768 encaps	17642 cycles	17930 cycles	0.98
ML-KEM-768 decaps	23218 cycles	23627 cycles	0.98
ML-KEM-1024 keypair	22181 cycles	22362 cycles	0.99
ML-KEM-1024 encaps	24116 cycles	24602 cycles	0.98
ML-KEM-1024 decaps	31703 cycles	32362 cycles	0.98

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

[oqs-bot]

AMD EPYC 3rd gen (c6a)

Details

Benchmark suite	Current: 1647e99	Previous: ebed0f3	Ratio
ML-KEM-512 keypair	13858 cycles	16681 cycles	0.83
ML-KEM-512 encaps	15565 cycles	18380 cycles	0.85
ML-KEM-512 decaps	21034 cycles	23712 cycles	0.89
ML-KEM-768 keypair	23415 cycles	28448 cycles	0.82
ML-KEM-768 encaps	24751 cycles	29801 cycles	0.83
ML-KEM-768 decaps	32627 cycles	37656 cycles	0.87
ML-KEM-1024 keypair	32841 cycles	41276 cycles	0.80
ML-KEM-1024 encaps	35167 cycles	43491 cycles	0.81
ML-KEM-1024 decaps	45730 cycles	53885 cycles	0.85

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

[oqs-bot]

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

Details

Benchmark suite	Current: 1647e99	Previous: ebed0f3	Ratio
ML-KEM-512 keypair	28399 cycles	28434 cycles	1.00
ML-KEM-512 encaps	35834 cycles	35766 cycles	1.00
ML-KEM-512 decaps	45554 cycles	45475 cycles	1.00
ML-KEM-768 keypair	45863 cycles	45954 cycles	1.00
ML-KEM-768 encaps	56265 cycles	56116 cycles	1.00
ML-KEM-768 decaps	69388 cycles	69482 cycles	1.00
ML-KEM-1024 keypair	71853 cycles	71559 cycles	1.00
ML-KEM-1024 encaps	84528 cycles	84605 cycles	1.00
ML-KEM-1024 decaps	101544 cycles	101115 cycles	1.00

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

[oqs-bot]

Arm Cortex-A76 (Raspberry Pi 5) benchmarks

Details

Benchmark suite	Current: 1647e99	Previous: ebed0f3	Ratio
ML-KEM-512 keypair	28311 cycles	28387 cycles	1.00
ML-KEM-512 encaps	34295 cycles	34239 cycles	1.00
ML-KEM-512 decaps	44520 cycles	44594 cycles	1.00
ML-KEM-768 keypair	47859 cycles	47826 cycles	1.00
ML-KEM-768 encaps	54136 cycles	54274 cycles	1.00
ML-KEM-768 decaps	68640 cycles	68610 cycles	1.00
ML-KEM-1024 keypair	70519 cycles	70609 cycles	1.00
ML-KEM-1024 encaps	79075 cycles	79023 cycles	1.00
ML-KEM-1024 decaps	98785 cycles	98791 cycles	1.00

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

[oqs-bot]

Intel Xeon 3rd gen (c6i)

Details

Benchmark suite	Current: 1647e99	Previous: ebed0f3	Ratio
ML-KEM-512 keypair	14391 cycles	16399 cycles	0.88
ML-KEM-512 encaps	16918 cycles	18695 cycles	0.90
ML-KEM-512 decaps	23408 cycles	25297 cycles	0.93
ML-KEM-768 keypair	25118 cycles	27938 cycles	0.90
ML-KEM-768 encaps	27078 cycles	29785 cycles	0.91
ML-KEM-768 decaps	36486 cycles	41180 cycles	0.89
ML-KEM-1024 keypair	33822 cycles	37708 cycles	0.90
ML-KEM-1024 encaps	36159 cycles	40685 cycles	0.89
ML-KEM-1024 decaps	49453 cycles	54424 cycles	0.91

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

[oqs-bot]

Graviton4

Details

Benchmark suite	Current: 1647e99	Previous: ebed0f3	Ratio
ML-KEM-512 keypair	17688 cycles	17702 cycles	1.00
ML-KEM-512 encaps	20670 cycles	20702 cycles	1.00
ML-KEM-512 decaps	27133 cycles	27133 cycles	1
ML-KEM-768 keypair	29987 cycles	30013 cycles	1.00
ML-KEM-768 encaps	32854 cycles	32811 cycles	1.00
ML-KEM-768 decaps	42013 cycles	42061 cycles	1.00
ML-KEM-1024 keypair	43899 cycles	43914 cycles	1.00
ML-KEM-1024 encaps	48921 cycles	48930 cycles	1.00
ML-KEM-1024 decaps	61602 cycles	61496 cycles	1.00

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

[oqs-bot]

AMD EPYC 4th gen (c7a)

Details

Benchmark suite	Current: 1647e99	Previous: ebed0f3	Ratio
ML-KEM-512 keypair	10521 cycles	12005 cycles	0.88
ML-KEM-512 encaps	12061 cycles	13291 cycles	0.91
ML-KEM-512 decaps	17019 cycles	18051 cycles	0.94
ML-KEM-768 keypair	18414 cycles	20559 cycles	0.90
ML-KEM-768 encaps	19515 cycles	21546 cycles	0.91
ML-KEM-768 decaps	26626 cycles	28661 cycles	0.93
ML-KEM-1024 keypair	24623 cycles	27867 cycles	0.88
ML-KEM-1024 encaps	26733 cycles	29966 cycles	0.89
ML-KEM-1024 decaps	36391 cycles	39500 cycles	0.92

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

[oqs-bot]

Graviton3

Details

Benchmark suite	Current: 1647e99	Previous: ebed0f3	Ratio
ML-KEM-512 keypair	18737 cycles	18762 cycles	1.00
ML-KEM-512 encaps	22002 cycles	22034 cycles	1.00
ML-KEM-512 decaps	29037 cycles	29048 cycles	1.00
ML-KEM-768 keypair	31782 cycles	31800 cycles	1.00
ML-KEM-768 encaps	35006 cycles	34950 cycles	1.00
ML-KEM-768 decaps	45010 cycles	45042 cycles	1.00
ML-KEM-1024 keypair	46334 cycles	46355 cycles	1.00
ML-KEM-1024 encaps	51700 cycles	51750 cycles	1.00
ML-KEM-1024 decaps	65265 cycles	65261 cycles	1.00

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

[oqs-bot]

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

Details

Benchmark suite	Current: 1647e99	Previous: ebed0f3	Ratio
ML-KEM-512 keypair	40116 cycles	40315 cycles	1.00
ML-KEM-512 encaps	48288 cycles	48372 cycles	1.00
ML-KEM-512 decaps	62456 cycles	62465 cycles	1.00
ML-KEM-768 keypair	63657 cycles	63637 cycles	1.00
ML-KEM-768 encaps	74664 cycles	74830 cycles	1.00
ML-KEM-768 decaps	93254 cycles	93236 cycles	1.00
ML-KEM-1024 keypair	95000 cycles	95054 cycles	1.00
ML-KEM-1024 encaps	108988 cycles	109053 cycles	1.00
ML-KEM-1024 decaps	131759 cycles	131949 cycles	1.00

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

[oqs-bot]

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

Details

Benchmark suite	Current: 1647e99	Previous: ebed0f3	Ratio
ML-KEM-512 keypair	45781 cycles	45812 cycles	1.00
ML-KEM-512 encaps	54744 cycles	54778 cycles	1.00
ML-KEM-512 decaps	70275 cycles	70387 cycles	1.00
ML-KEM-768 keypair	73830 cycles	73966 cycles	1.00
ML-KEM-768 encaps	85352 cycles	85383 cycles	1.00
ML-KEM-768 decaps	106339 cycles	106459 cycles	1.00
ML-KEM-1024 keypair	111726 cycles	111805 cycles	1.00
ML-KEM-1024 encaps	125852 cycles	125952 cycles	1.00
ML-KEM-1024 decaps	151675 cycles	151837 cycles	1.00

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

[oqs-bot]

Graviton4 (no-opt)

Details

Benchmark suite	Current: 1647e99	Previous: ebed0f3	Ratio
ML-KEM-512 keypair	35451 cycles	35858 cycles	0.99
ML-KEM-512 encaps	40205 cycles	40235 cycles	1.00
ML-KEM-512 decaps	51249 cycles	51231 cycles	1.00
ML-KEM-768 keypair	56829 cycles	56807 cycles	1.00
ML-KEM-768 encaps	64602 cycles	65391 cycles	0.99
ML-KEM-768 decaps	78906 cycles	79291 cycles	1.00
ML-KEM-1024 keypair	88081 cycles	88036 cycles	1.00
ML-KEM-1024 encaps	97248 cycles	97186 cycles	1.00
ML-KEM-1024 decaps	116231 cycles	116104 cycles	1.00

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

[oqs-bot]

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

Details

Benchmark suite	Current: 1647e99	Previous: ebed0f3	Ratio
ML-KEM-512 keypair	36534 cycles	36687 cycles	1.00
ML-KEM-512 encaps	43003 cycles	43011 cycles	1.00
ML-KEM-512 decaps	55675 cycles	55666 cycles	1.00
ML-KEM-768 keypair	58415 cycles	58457 cycles	1.00
ML-KEM-768 encaps	67402 cycles	67409 cycles	1.00
ML-KEM-768 decaps	84350 cycles	84377 cycles	1.00
ML-KEM-1024 keypair	88631 cycles	88658 cycles	1.00
ML-KEM-1024 encaps	98864 cycles	98909 cycles	1.00
ML-KEM-1024 decaps	120369 cycles	120440 cycles	1.00

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

[oqs-bot]

Graviton3 (no-opt)

Details

Benchmark suite	Current: 1647e99	Previous: ebed0f3	Ratio
ML-KEM-512 keypair	39000 cycles	39834 cycles	0.98
ML-KEM-512 encaps	44609 cycles	44640 cycles	1.00
ML-KEM-512 decaps	56711 cycles	56703 cycles	1.00
ML-KEM-768 keypair	62433 cycles	62431 cycles	1.00
ML-KEM-768 encaps	70885 cycles	71780 cycles	0.99
ML-KEM-768 decaps	86781 cycles	87166 cycles	1.00
ML-KEM-1024 keypair	96335 cycles	96262 cycles	1.00
ML-KEM-1024 encaps	106377 cycles	106330 cycles	1.00
ML-KEM-1024 decaps	126937 cycles	126801 cycles	1.00

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

[oqs-bot]

Arm Cortex-A55 (Snapdragon 888) benchmarks

Details

Benchmark suite	Current: 1647e99	Previous: ebed0f3	Ratio
ML-KEM-512 keypair	59531 cycles	59477 cycles	1.00
ML-KEM-512 encaps	67162 cycles	67251 cycles	1.00
ML-KEM-512 decaps	85830 cycles	85738 cycles	1.00
ML-KEM-768 keypair	96983 cycles	97018 cycles	1.00
ML-KEM-768 encaps	110439 cycles	110424 cycles	1.00
ML-KEM-768 decaps	137219 cycles	137136 cycles	1.00
ML-KEM-1024 keypair	154182 cycles	154167 cycles	1.00
ML-KEM-1024 encaps	170792 cycles	170457 cycles	1.00
ML-KEM-1024 decaps	206619 cycles	206792 cycles	1.00

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

[oqs-bot]

Graviton2

Details

Benchmark suite	Current: 1647e99	Previous: ebed0f3	Ratio
ML-KEM-512 keypair	28394 cycles	28314 cycles	1.00
ML-KEM-512 encaps	34245 cycles	34303 cycles	1.00
ML-KEM-512 decaps	44588 cycles	44520 cycles	1.00
ML-KEM-768 keypair	47894 cycles	47846 cycles	1.00
ML-KEM-768 encaps	54377 cycles	54137 cycles	1.00
ML-KEM-768 decaps	68808 cycles	68665 cycles	1.00
ML-KEM-1024 keypair	70500 cycles	70549 cycles	1.00
ML-KEM-1024 encaps	79000 cycles	79141 cycles	1.00
ML-KEM-1024 decaps	98806 cycles	98835 cycles	1.00

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

[oqs-bot]

Graviton2 (no-opt)

Details

Benchmark suite	Current: 1647e99	Previous: ebed0f3	Ratio
ML-KEM-512 keypair	59023 cycles	59570 cycles	0.99
ML-KEM-512 encaps	68543 cycles	68575 cycles	1.00
ML-KEM-512 decaps	87352 cycles	87326 cycles	1.00
ML-KEM-768 keypair	95685 cycles	95742 cycles	1.00
ML-KEM-768 encaps	109493 cycles	109660 cycles	1.00
ML-KEM-768 decaps	134366 cycles	134532 cycles	1.00
ML-KEM-1024 keypair	146719 cycles	148351 cycles	0.99
ML-KEM-1024 encaps	162524 cycles	164301 cycles	0.99
ML-KEM-1024 decaps	194686 cycles	195562 cycles	1.00

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

[oqs-bot]

SpacemiT K1 8 (Banana Pi F3) benchmarks

Details

Benchmark suite	Current: 1647e99	Previous: ebed0f3	Ratio
ML-KEM-512 keypair	155156 cycles	155116 cycles	1.00
ML-KEM-512 encaps	163334 cycles	163295 cycles	1.00
ML-KEM-512 decaps	206536 cycles	206550 cycles	1.00
ML-KEM-768 keypair	249502 cycles	249522 cycles	1.00
ML-KEM-768 encaps	270309 cycles	270296 cycles	1.00
ML-KEM-768 decaps	332165 cycles	332114 cycles	1.00
ML-KEM-1024 keypair	395238 cycles	395146 cycles	1.00
ML-KEM-1024 encaps	423919 cycles	423791 cycles	1.00
ML-KEM-1024 decaps	505639 cycles	505524 cycles	1.00

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

[oqs-bot]

Arm Cortex-A72 (Raspberry Pi 4) benchmarks

Details

Benchmark suite	Current: 1647e99	Previous: ebed0f3	Ratio
ML-KEM-512 keypair	50737 cycles	50672 cycles	1.00
ML-KEM-512 encaps	58468 cycles	58540 cycles	1.00
ML-KEM-512 decaps	74124 cycles	74115 cycles	1.00
ML-KEM-768 keypair	86957 cycles	86481 cycles	1.01
ML-KEM-768 encaps	95630 cycles	94380 cycles	1.01
ML-KEM-768 decaps	118875 cycles	117403 cycles	1.01
ML-KEM-1024 keypair	131396 cycles	129771 cycles	1.01
ML-KEM-1024 encaps	143409 cycles	142171 cycles	1.01
ML-KEM-1024 decaps	174419 cycles	173614 cycles	1.00

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

[oqs-bot]

CBMC Results (ML-KEM-512)

Full Results (139 proofs)

Proof	Status	Current	Previous	Change
**TOTAL**	✅	1428s	1185s	+20.5%
mlk_indcpa_keypair_derand	✅	236s	205s	+15%
mlk_indcpa_enc	✅	224s	181s	+24%
mlk_keccak_squeezeblocks_x4	✅	170s	133s	+28%
mlk_rej_uniform_c	✅	108s	73s	+48%
mlk_polyvec_basemul_acc_montgomery_cached_c	✅	60s	43s	+40%
poly_ntt_native	✅	43s	34s	+26%
mlk_poly_rej_uniform	✅	39s	32s	+22%
mlk_ntt_layer	✅	29s	23s	+26%
polyvec_basemul_acc_montgomery_cached_native	✅	25s	21s	+19%
keccakf1600x4_permute_native_x4	✅	18s	18s	+0%
mlk_poly_reduce_native	✅	18s	14s	+29%
mlk_poly_sub	✅	13s	9s	+44%
mlk_keccak_absorb_once_x4	✅	12s	9s	+33%
mlk_ntt_butterfly_block	✅	12s	10s	+20%
mlk_poly_frommsg	✅	12s	9s	+33%
mlk_poly_frombytes_native	✅	11s	11s	+0%
mlk_polyvec_add	✅	11s	10s	+10%
mlk_poly_rej_uniform_x4	✅	10s	6s	+67%
mlk_fqmul	✅	8s	6s	+33%
mlk_indcpa_dec	✅	8s	10s	-20%
mlk_keccak_squeeze_once	✅	8s	8s	+0%
mlk_keccak_squeezeblocks	✅	8s	7s	+14%
mlk_invntt_layer	✅	7s	6s	+17%
mlk_poly_tomsg	✅	7s	1s	+600%
mlk_polymat_permute_bitrev_to_custom	✅	7s	2s	+250%
kem_dec	✅	6s	5s	+20%
mlk_keccakf1600_permute	✅	6s	4s	+50%
mlk_poly_tomont	✅	6s	3s	+100%
keccak_f1600_x4_native_aarch64_v8a_scalar_hybrid	✅	5s	3s	+67%
keccakf1600_permute_native	✅	5s	5s	+0%
mlk_gen_matrix_serial	✅	5s	3s	+67%
mlk_keccak_absorb_once	✅	5s	5s	+0%
mlk_poly_cbd_eta2	✅	5s	4s	+25%
mlk_poly_compress_du	✅	5s	4s	+25%
mlk_poly_getnoise_eta1122_4x	✅	5s	5s	+0%
mlk_poly_getnoise_eta1_4x	✅	5s	3s	+67%
mlk_poly_getnoise_eta1_4x_native	✅	5s	2s	+150%
mlk_poly_mulcache_compute_c	✅	5s	2s	+150%
mlk_scalar_decompress_d10	✅	5s	4s	+25%
mlk_shake256x4	✅	5s	6s	-17%
keccak_f1600_x4_native_aarch64_v84a	✅	4s	2s	+100%
mlk_keccakf1600_extract_bytes (big endian)	✅	4s	1s	+300%
mlk_keccakf1600_xor_bytes (big endian)	✅	4s	3s	+33%
mlk_poly_add	✅	4s	2s	+100%
mlk_poly_cbd_eta1	✅	4s	2s	+100%
mlk_poly_decompress_du	✅	4s	2s	+100%
mlk_poly_getnoise_eta2	✅	4s	2s	+100%
mlk_poly_mulcache_compute	✅	4s	2s	+100%
mlk_poly_reduce_c	✅	4s	2s	+100%
mlk_polyvec_permute_bitrev_to_custom_native	✅	4s	4s	+0%
mlk_polyvec_reduce	✅	4s	2s	+100%
mlk_polyvec_tomont	✅	4s	2s	+100%
mlk_scalar_compress_d5	✅	4s	2s	+100%
poly_getnoise_eta1122_4x_native	✅	4s	1s	+300%
poly_tomont_native_aarch64	✅	4s	3s	+33%
rej_uniform_native	✅	4s	3s	+33%
rej_uniform_native_aarch64	✅	4s	3s	+33%
keccak_f1600_x1_native_aarch64_v84a	✅	3s	3s	+0%
kem_check_pk	✅	3s	3s	+0%
kem_check_sk	✅	3s	1s	+200%
kem_enc_derand	✅	3s	4s	-25%
kem_keypair	✅	3s	4s	-25%
mlk_ct_cmask_neg_i16	✅	3s	2s	+50%
mlk_ct_cmask_nonzero_u16	✅	3s	3s	+0%
mlk_ct_cmask_nonzero_u8	✅	3s	2s	+50%
mlk_ct_memcmp	✅	3s	2s	+50%
mlk_gen_matrix	✅	3s	4s	-25%
mlk_poly_frombytes	✅	3s	3s	+0%
mlk_poly_invntt_tomont_c	✅	3s	4s	-25%
mlk_poly_ntt	✅	3s	2s	+50%
mlk_poly_tobytes	✅	3s	2s	+50%
mlk_poly_tomont_native	✅	3s	2s	+50%
mlk_polyvec_basemul_acc_montgomery_cached	✅	3s	2s	+50%
mlk_polyvec_compress_du	✅	3s	3s	+0%
mlk_polyvec_decompress_du	✅	3s	5s	-40%
mlk_polyvec_frombytes	✅	3s	2s	+50%
mlk_scalar_compress_d11	✅	3s	2s	+50%
mlk_scalar_decompress_d11	✅	3s	3s	+0%
mlk_scalar_decompress_d4	✅	3s	3s	+0%
mlk_scalar_decompress_d5	✅	3s	1s	+200%
poly_invntt_tomont_native	✅	3s	5s	-40%
poly_mulcache_compute_native_aarch64	✅	3s	3s	+0%
poly_reduce_native_aarch64	✅	3s	2s	+50%
polyvec_basemul_acc_montgomery_cached_k3_native_aarch64	✅	3s	3s	+0%
sys_check_capability	✅	3s	3s	+0%
intt_native_aarch64	✅	2s	3s	-33%
keccak_f1600_x4_native_aarch64_v8a_v84a_scalar_hybrid	✅	2s	3s	-33%
kem_keypair_derand	✅	2s	5s	-60%
mlk_barrett_reduce	✅	2s	2s	+0%
mlk_check_pct	✅	2s	1s	+100%
mlk_ct_cmov_zero	✅	2s	6s	-67%
mlk_ct_get_optblocker_i32	✅	2s	2s	+0%
mlk_ct_get_optblocker_u32	✅	2s	1s	+100%
mlk_ct_sel_int16	✅	2s	1s	+100%
mlk_ct_sel_uint8	✅	2s	2s	+0%
mlk_keccakf1600_extract_bytes	✅	2s	2s	+0%
mlk_keccakf1600_xor_bytes	✅	2s	2s	+0%
mlk_keccakf1600x4_permute	✅	2s	3s	-33%
mlk_montgomery_reduce	✅	2s	2s	+0%
mlk_poly_compress_dv	✅	2s	3s	-33%
mlk_poly_decompress_dv	✅	2s	2s	+0%
mlk_poly_frombytes_c	✅	2s	1s	+100%
mlk_poly_mulcache_compute_native	✅	2s	3s	-33%
mlk_poly_ntt_c	✅	2s	1s	+100%
mlk_poly_reduce	✅	2s	2s	+0%
mlk_poly_tobytes_c	✅	2s	2s	+0%
mlk_poly_tobytes_native	✅	2s	3s	-33%
mlk_poly_tomont_c	✅	2s	4s	-50%
mlk_polyvec_mulcache_compute	✅	2s	2s	+0%
mlk_polyvec_ntt	✅	2s	2s	+0%
mlk_polyvec_tobytes	✅	2s	2s	+0%
mlk_rej_uniform	✅	2s	4s	-50%
mlk_scalar_compress_d1	✅	2s	2s	+0%
mlk_scalar_compress_d10	✅	2s	4s	-50%
mlk_scalar_compress_d4	✅	2s	3s	-33%
mlk_scalar_signed_to_unsigned_q	✅	2s	1s	+100%
mlk_sha3_512	✅	2s	3s	-33%
mlk_shake128_absorb_once	✅	2s	3s	-33%
mlk_shake128_squeezeblocks	✅	2s	1s	+100%
mlk_shake128x4_absorb_once	✅	2s	4s	-50%
mlk_shake256	✅	2s	1s	+100%
mlk_value_barrier_i32	✅	2s	2s	+0%
mlk_value_barrier_u32	✅	2s	4s	-50%
polyvec_basemul_acc_montgomery_cached_k2_native_aarch64	✅	2s	2s	+0%
polyvec_basemul_acc_montgomery_cached_k4_native_aarch64	✅	2s	3s	-33%
keccak_f1600_x1_native_aarch64	✅	1s	3s	-67%
kem_enc	✅	1s	2s	-50%
mlk_ct_get_optblocker_u8	✅	1s	2s	-50%
mlk_keccakf1600x4_extract_bytes	✅	1s	2s	-50%
mlk_keccakf1600x4_xor_bytes	✅	1s	2s	-50%
mlk_matvec_mul	✅	1s	2s	-50%
mlk_poly_invntt_tomont	✅	1s	4s	-75%
mlk_polyvec_invntt_tomont	✅	1s	3s	-67%
mlk_polyvec_permute_bitrev_to_custom	✅	1s	1s	+0%
mlk_sha3_256	✅	1s	2s	-50%
mlk_shake128x4_squeezeblocks	✅	1s	1s	+0%
mlk_value_barrier_u8	✅	1s	2s	-50%
ntt_native_aarch64	✅	1s	3s	-67%
poly_tobytes_native_aarch64	✅	1s	2s	-50%

[oqs-bot]

CBMC Results (ML-KEM-1024)

⚠️ Attention Required

Proof	Status	Current	Previous	Change
mlk_ntt_layer	⚠️	47s	26s	+81%

Full Results (139 proofs)

Proof	Status	Current	Previous	Change
**TOTAL**	✅	1571s	1431s	+9.8%
mlk_indcpa_enc	✅	245s	254s	-4%
mlk_indcpa_keypair_derand	✅	176s	163s	+8%
mlk_keccak_squeezeblocks_x4	✅	144s	122s	+18%
polyvec_basemul_acc_montgomery_cached_native	✅	137s	107s	+28%
mlk_polyvec_add	✅	131s	120s	+9%
mlk_rej_uniform_c	✅	74s	68s	+9%
mlk_polyvec_basemul_acc_montgomery_cached_c	✅	48s	44s	+9%
mlk_ntt_layer	⚠️	47s	26s	+81%
mlk_poly_rej_uniform	✅	47s	35s	+34%
mlk_poly_decompress_dv	✅	27s	25s	+8%
poly_ntt_native	✅	23s	22s	+5%
keccakf1600x4_permute_native_x4	✅	19s	19s	+0%
mlk_poly_reduce_native	✅	17s	13s	+31%
mlk_polyvec_ntt	✅	17s	12s	+42%
mlk_indcpa_dec	✅	16s	16s	+0%
mlk_poly_compress_du	✅	12s	9s	+33%
mlk_poly_frommsg	✅	11s	9s	+22%
mlk_ntt_butterfly_block	✅	10s	10s	+0%
mlk_poly_frombytes_native	✅	10s	7s	+43%
mlk_poly_sub	✅	10s	9s	+11%
kem_dec	✅	9s	8s	+12%
mlk_fqmul	✅	9s	6s	+50%
mlk_keccak_absorb_once_x4	✅	9s	8s	+12%
mlk_poly_rej_uniform_x4	✅	9s	9s	+0%
mlk_gen_matrix	✅	7s	6s	+17%
mlk_gen_matrix_serial	✅	7s	6s	+17%
mlk_keccak_squeeze_once	✅	7s	8s	-12%
mlk_keccak_squeezeblocks	✅	7s	8s	-12%
mlk_polyvec_permute_bitrev_to_custom_native	✅	6s	5s	+20%
keccakf1600_permute_native	✅	5s	8s	-38%
mlk_invntt_layer	✅	5s	4s	+25%
mlk_keccak_absorb_once	✅	5s	3s	+67%
mlk_poly_tomsg	✅	5s	3s	+67%
mlk_polymat_permute_bitrev_to_custom	✅	5s	5s	+0%
mlk_polyvec_tobytes	✅	5s	3s	+67%
keccak_f1600_x1_native_aarch64	✅	4s	3s	+33%
mlk_check_pct	✅	4s	5s	-20%
mlk_ct_memcmp	✅	4s	4s	+0%
mlk_keccakf1600_permute	✅	4s	3s	+33%
mlk_keccakf1600x4_extract_bytes	✅	4s	3s	+33%
mlk_poly_frombytes	✅	4s	4s	+0%
mlk_poly_getnoise_eta1_4x	✅	4s	3s	+33%
mlk_poly_getnoise_eta1_4x_native	✅	4s	2s	+100%
mlk_poly_tobytes_native	✅	4s	2s	+100%
mlk_poly_tomont	✅	4s	1s	+300%
mlk_scalar_compress_d5	✅	4s	3s	+33%
mlk_scalar_decompress_d11	✅	4s	2s	+100%
mlk_shake256x4	✅	4s	6s	-33%
poly_tobytes_native_aarch64	✅	4s	3s	+33%
intt_native_aarch64	✅	3s	4s	-25%
kem_keypair	✅	3s	2s	+50%
mlk_ct_cmask_nonzero_u8	✅	3s	3s	+0%
mlk_ct_get_optblocker_i32	✅	3s	2s	+50%
mlk_keccakf1600_xor_bytes (big endian)	✅	3s	2s	+50%
mlk_matvec_mul	✅	3s	4s	-25%
mlk_montgomery_reduce	✅	3s	2s	+50%
mlk_poly_cbd_eta1	✅	3s	2s	+50%
mlk_poly_cbd_eta2	✅	3s	2s	+50%
mlk_poly_compress_dv	✅	3s	1s	+200%
mlk_poly_getnoise_eta1122_4x	✅	3s	4s	-25%
mlk_poly_mulcache_compute_c	✅	3s	2s	+50%
mlk_poly_ntt_c	✅	3s	3s	+0%
mlk_poly_reduce	✅	3s	2s	+50%
mlk_poly_tobytes	✅	3s	1s	+200%
mlk_poly_tomont_c	✅	3s	1s	+200%
mlk_poly_tomont_native	✅	3s	3s	+0%
mlk_polyvec_compress_du	✅	3s	3s	+0%
mlk_polyvec_invntt_tomont	✅	3s	2s	+50%
mlk_polyvec_tomont	✅	3s	1s	+200%
mlk_scalar_decompress_d10	✅	3s	1s	+200%
mlk_scalar_decompress_d4	✅	3s	2s	+50%
mlk_scalar_decompress_d5	✅	3s	3s	+0%
mlk_shake128_absorb_once	✅	3s	3s	+0%
mlk_shake128x4_absorb_once	✅	3s	2s	+50%
mlk_shake256	✅	3s	2s	+50%
mlk_value_barrier_i32	✅	3s	2s	+50%
ntt_native_aarch64	✅	3s	2s	+50%
poly_invntt_tomont_native	✅	3s	3s	+0%
poly_tomont_native_aarch64	✅	3s	2s	+50%
polyvec_basemul_acc_montgomery_cached_k3_native_aarch64	✅	3s	2s	+50%
polyvec_basemul_acc_montgomery_cached_k4_native_aarch64	✅	3s	3s	+0%
rej_uniform_native_aarch64	✅	3s	4s	-25%
keccak_f1600_x4_native_aarch64_v8a_scalar_hybrid	✅	2s	1s	+100%
kem_check_pk	✅	2s	2s	+0%
kem_enc	✅	2s	2s	+0%
kem_enc_derand	✅	2s	2s	+0%
kem_keypair_derand	✅	2s	3s	-33%
mlk_barrett_reduce	✅	2s	2s	+0%
mlk_ct_cmask_nonzero_u16	✅	2s	2s	+0%
mlk_ct_get_optblocker_u32	✅	2s	2s	+0%
mlk_ct_sel_int16	✅	2s	2s	+0%
mlk_keccakf1600_extract_bytes (big endian)	✅	2s	1s	+100%
mlk_poly_add	✅	2s	3s	-33%
mlk_poly_getnoise_eta2	✅	2s	5s	-60%
mlk_poly_invntt_tomont	✅	2s	1s	+100%
mlk_poly_invntt_tomont_c	✅	2s	2s	+0%
mlk_poly_mulcache_compute	✅	2s	3s	-33%
mlk_poly_mulcache_compute_native	✅	2s	4s	-50%
mlk_poly_ntt	✅	2s	3s	-33%
mlk_poly_reduce_c	✅	2s	3s	-33%
mlk_polyvec_decompress_du	✅	2s	1s	+100%
mlk_polyvec_mulcache_compute	✅	2s	3s	-33%
mlk_polyvec_permute_bitrev_to_custom	✅	2s	1s	+100%
mlk_rej_uniform	✅	2s	1s	+100%
mlk_scalar_compress_d1	✅	2s	2s	+0%
mlk_scalar_compress_d10	✅	2s	2s	+0%
mlk_scalar_compress_d11	✅	2s	2s	+0%
mlk_scalar_compress_d4	✅	2s	2s	+0%
mlk_scalar_signed_to_unsigned_q	✅	2s	3s	-33%
mlk_sha3_256	✅	2s	3s	-33%
mlk_sha3_512	✅	2s	1s	+100%
mlk_shake128_squeezeblocks	✅	2s	3s	-33%
mlk_shake128x4_squeezeblocks	✅	2s	2s	+0%
mlk_value_barrier_u32	✅	2s	2s	+0%
mlk_value_barrier_u8	✅	2s	3s	-33%
poly_getnoise_eta1122_4x_native	✅	2s	1s	+100%
poly_mulcache_compute_native_aarch64	✅	2s	2s	+0%
poly_reduce_native_aarch64	✅	2s	4s	-50%
polyvec_basemul_acc_montgomery_cached_k2_native_aarch64	✅	2s	2s	+0%
rej_uniform_native	✅	2s	3s	-33%
keccak_f1600_x1_native_aarch64_v84a	✅	1s	1s	+0%
keccak_f1600_x4_native_aarch64_v84a	✅	1s	2s	-50%
keccak_f1600_x4_native_aarch64_v8a_v84a_scalar_hybrid	✅	1s	2s	-50%
kem_check_sk	✅	1s	2s	-50%
mlk_ct_cmask_neg_i16	✅	1s	1s	+0%
mlk_ct_cmov_zero	✅	1s	2s	-50%
mlk_ct_get_optblocker_u8	✅	1s	3s	-67%
mlk_ct_sel_uint8	✅	1s	3s	-67%
mlk_keccakf1600_extract_bytes	✅	1s	5s	-80%
mlk_keccakf1600_xor_bytes	✅	1s	1s	+0%
mlk_keccakf1600x4_permute	✅	1s	3s	-67%
mlk_keccakf1600x4_xor_bytes	✅	1s	2s	-50%
mlk_poly_decompress_du	✅	1s	1s	+0%
mlk_poly_frombytes_c	✅	1s	1s	+0%
mlk_poly_tobytes_c	✅	1s	2s	-50%
mlk_polyvec_basemul_acc_montgomery_cached	✅	1s	3s	-67%
mlk_polyvec_frombytes	✅	1s	3s	-67%
mlk_polyvec_reduce	✅	1s	3s	-67%
sys_check_capability	✅	1s	4s	-75%

[oqs-bot]

CBMC Results (ML-KEM-768)

Full Results (139 proofs)

Proof	Status	Current	Previous	Change
**TOTAL**	✅	1580s	1681s	-6.0%
mlk_indcpa_keypair_derand	✅	474s	498s	-5%
mlk_indcpa_enc	✅	207s	232s	-11%
mlk_keccak_squeezeblocks_x4	✅	138s	141s	-2%
mlk_rej_uniform_c	✅	85s	101s	-16%
polyvec_basemul_acc_montgomery_cached_native	✅	63s	65s	-3%
mlk_polyvec_basemul_acc_montgomery_cached_c	✅	51s	64s	-20%
mlk_poly_rej_uniform	✅	42s	45s	-7%
mlk_ntt_layer	✅	35s	45s	-22%
poly_ntt_native	✅	30s	33s	-9%
keccakf1600x4_permute_native_x4	✅	20s	18s	+11%
mlk_indcpa_dec	✅	15s	13s	+15%
mlk_poly_reduce_native	✅	15s	17s	-12%
mlk_polyvec_add	✅	13s	13s	+0%
mlk_keccak_absorb_once_x4	✅	11s	9s	+22%
mlk_keccak_squeeze_once	✅	10s	9s	+11%
mlk_poly_frombytes_native	✅	9s	10s	-10%
mlk_poly_frommsg	✅	9s	10s	-10%
keccakf1600_permute_native	✅	8s	5s	+60%
kem_dec	✅	8s	6s	+33%
mlk_ntt_butterfly_block	✅	8s	9s	-11%
mlk_poly_rej_uniform_x4	✅	8s	7s	+14%
mlk_poly_sub	✅	8s	9s	-11%
mlk_polymat_permute_bitrev_to_custom	✅	8s	6s	+33%
mlk_invntt_layer	✅	7s	6s	+17%
mlk_keccak_squeezeblocks	✅	7s	9s	-22%
mlk_fqmul	✅	6s	5s	+20%
poly_mulcache_compute_native_aarch64	✅	6s	4s	+50%
mlk_poly_cbd_eta2	✅	5s	2s	+150%
poly_invntt_tomont_native	✅	5s	3s	+67%
mlk_check_pct	✅	4s	3s	+33%
mlk_ct_sel_uint8	✅	4s	2s	+100%
mlk_gen_matrix	✅	4s	4s	+0%
mlk_keccak_absorb_once	✅	4s	5s	-20%
mlk_keccakf1600_permute	✅	4s	4s	+0%
mlk_poly_compress_du	✅	4s	4s	+0%
mlk_poly_getnoise_eta1_4x	✅	4s	4s	+0%
mlk_poly_getnoise_eta1_4x_native	✅	4s	3s	+33%
mlk_poly_reduce_c	✅	4s	2s	+100%
mlk_poly_tobytes	✅	4s	1s	+300%
mlk_polyvec_basemul_acc_montgomery_cached	✅	4s	5s	-20%
mlk_polyvec_invntt_tomont	✅	4s	2s	+100%
mlk_shake256x4	✅	4s	5s	-20%
mlk_value_barrier_i32	✅	4s	3s	+33%
polyvec_basemul_acc_montgomery_cached_k4_native_aarch64	✅	4s	3s	+33%
intt_native_aarch64	✅	3s	1s	+200%
keccak_f1600_x1_native_aarch64_v84a	✅	3s	1s	+200%
kem_check_pk	✅	3s	3s	+0%
kem_check_sk	✅	3s	1s	+200%
kem_enc	✅	3s	1s	+200%
mlk_ct_cmov_zero	✅	3s	2s	+50%
mlk_ct_memcmp	✅	3s	1s	+200%
mlk_gen_matrix_serial	✅	3s	4s	-25%
mlk_keccakf1600_xor_bytes	✅	3s	2s	+50%
mlk_keccakf1600_xor_bytes (big endian)	✅	3s	1s	+200%
mlk_keccakf1600x4_extract_bytes	✅	3s	2s	+50%
mlk_matvec_mul	✅	3s	2s	+50%
mlk_poly_add	✅	3s	4s	-25%
mlk_poly_decompress_du	✅	3s	3s	+0%
mlk_poly_decompress_dv	✅	3s	3s	+0%
mlk_poly_frombytes	✅	3s	2s	+50%
mlk_poly_getnoise_eta2	✅	3s	2s	+50%
mlk_poly_invntt_tomont_c	✅	3s	3s	+0%
mlk_poly_mulcache_compute	✅	3s	2s	+50%
mlk_poly_tomont	✅	3s	2s	+50%
mlk_poly_tomsg	✅	3s	3s	+0%
mlk_polyvec_compress_du	✅	3s	3s	+0%
mlk_polyvec_permute_bitrev_to_custom_native	✅	3s	2s	+50%
mlk_scalar_compress_d1	✅	3s	3s	+0%
mlk_scalar_compress_d10	✅	3s	2s	+50%
mlk_scalar_decompress_d11	✅	3s	3s	+0%
mlk_shake128_absorb_once	✅	3s	5s	-40%
mlk_shake128_squeezeblocks	✅	3s	2s	+50%
mlk_shake128x4_absorb_once	✅	3s	2s	+50%
mlk_value_barrier_u8	✅	3s	2s	+50%
poly_reduce_native_aarch64	✅	3s	3s	+0%
polyvec_basemul_acc_montgomery_cached_k2_native_aarch64	✅	3s	2s	+50%
rej_uniform_native	✅	3s	5s	-40%
keccak_f1600_x4_native_aarch64_v84a	✅	2s	1s	+100%
keccak_f1600_x4_native_aarch64_v8a_scalar_hybrid	✅	2s	1s	+100%
kem_enc_derand	✅	2s	5s	-60%
kem_keypair_derand	✅	2s	3s	-33%
mlk_ct_cmask_neg_i16	✅	2s	2s	+0%
mlk_ct_cmask_nonzero_u16	✅	2s	2s	+0%
mlk_ct_cmask_nonzero_u8	✅	2s	2s	+0%
mlk_ct_get_optblocker_i32	✅	2s	2s	+0%
mlk_ct_get_optblocker_u32	✅	2s	3s	-33%
mlk_ct_get_optblocker_u8	✅	2s	4s	-50%
mlk_ct_sel_int16	✅	2s	4s	-50%
mlk_keccakf1600_extract_bytes	✅	2s	3s	-33%
mlk_keccakf1600_extract_bytes (big endian)	✅	2s	2s	+0%
mlk_montgomery_reduce	✅	2s	3s	-33%
mlk_poly_cbd_eta1	✅	2s	2s	+0%
mlk_poly_frombytes_c	✅	2s	2s	+0%
mlk_poly_getnoise_eta1122_4x	✅	2s	2s	+0%
mlk_poly_invntt_tomont	✅	2s	2s	+0%
mlk_poly_mulcache_compute_c	✅	2s	3s	-33%
mlk_poly_mulcache_compute_native	✅	2s	3s	-33%
mlk_poly_ntt	✅	2s	3s	-33%
mlk_poly_tobytes_native	✅	2s	2s	+0%
mlk_poly_tomont_c	✅	2s	1s	+100%
mlk_polyvec_decompress_du	✅	2s	3s	-33%
mlk_polyvec_frombytes	✅	2s	2s	+0%
mlk_polyvec_mulcache_compute	✅	2s	2s	+0%
mlk_polyvec_ntt	✅	2s	2s	+0%
mlk_polyvec_permute_bitrev_to_custom	✅	2s	4s	-50%
mlk_polyvec_reduce	✅	2s	2s	+0%
mlk_polyvec_tobytes	✅	2s	5s	-60%
mlk_polyvec_tomont	✅	2s	2s	+0%
mlk_rej_uniform	✅	2s	3s	-33%
mlk_scalar_compress_d11	✅	2s	2s	+0%
mlk_scalar_compress_d4	✅	2s	3s	-33%
mlk_scalar_compress_d5	✅	2s	3s	-33%
mlk_scalar_decompress_d10	✅	2s	4s	-50%
mlk_scalar_decompress_d5	✅	2s	1s	+100%
mlk_scalar_signed_to_unsigned_q	✅	2s	4s	-50%
mlk_sha3_256	✅	2s	1s	+100%
mlk_sha3_512	✅	2s	2s	+0%
mlk_shake256	✅	2s	1s	+100%
mlk_value_barrier_u32	✅	2s	2s	+0%
ntt_native_aarch64	✅	2s	3s	-33%
poly_getnoise_eta1122_4x_native	✅	2s	2s	+0%
poly_tobytes_native_aarch64	✅	2s	2s	+0%
polyvec_basemul_acc_montgomery_cached_k3_native_aarch64	✅	2s	4s	-50%
rej_uniform_native_aarch64	✅	2s	5s	-60%
sys_check_capability	✅	2s	2s	+0%
keccak_f1600_x1_native_aarch64	✅	1s	2s	-50%
keccak_f1600_x4_native_aarch64_v8a_v84a_scalar_hybrid	✅	1s	4s	-75%
kem_keypair	✅	1s	4s	-75%
mlk_barrett_reduce	✅	1s	3s	-67%
mlk_keccakf1600x4_permute	✅	1s	1s	+0%
mlk_keccakf1600x4_xor_bytes	✅	1s	1s	+0%
mlk_poly_compress_dv	✅	1s	2s	-50%
mlk_poly_ntt_c	✅	1s	3s	-67%
mlk_poly_reduce	✅	1s	2s	-50%
mlk_poly_tobytes_c	✅	1s	2s	-50%
mlk_poly_tomont_native	✅	1s	3s	-67%
mlk_scalar_decompress_d4	✅	1s	3s	-67%
mlk_shake128x4_squeezeblocks	✅	1s	2s	-50%
poly_tomont_native_aarch64	✅	1s	3s	-67%