Consolidate MLK_USE_NATIVE_XXX macros by mkannwischer · Pull Request #1541 · pq-code-package/mlkem-native

mkannwischer · 2026-02-02T07:00:19Z

In the arithmetic backend, we signal support for certain functions using macros of the form MLK_USE_NATIVE_XXX (e.g., MLK_USE_NATIVE_NTT), while in the FIPS202 backend, we use MLK_USE_XXX_NATIVE (e.g., MLK_USE_FIPS202_X1_NATIVE). This commit consolidates it to MLK_USE_NATIVE_XXX.

Resolves Consolidate backend MLK_USE_XXX flags indicating function support #1523

In the arithmetic backend, we signal support for certain functions using macros of the form MLK_USE_NATIVE_XXX (e.g., MLK_USE_NATIVE_NTT), while in the FIPS202 backend, we use MLK_USE_XXX_NATIVE (e.g., MLK_USE_FIPS202_X1_NATIVE). This commit consolidates it to MLK_USE_NATIVE_XXX. - Resolves #1523 Signed-off-by: Matthias J. Kannwischer <matthias@kannwischer.eu>

oqs-bot · 2026-02-02T07:19:32Z

CBMC Results (ML-KEM-512)

Full Results (151 proofs)

Proof	Status	Current	Previous	Change
`TOTAL`	✅	1146s	1215s	-5.7%
`mlk_indcpa_keypair_derand`	✅	165s	171s	-4%
`mlk_indcpa_enc`	✅	155s	162s	-4%
`mlk_keccak_squeezeblocks_x4`	✅	132s	135s	-2%
`mlk_rej_uniform_c`	✅	69s	82s	-16%
`mlk_polyvec_basemul_acc_montgomery_cached_c`	✅	43s	45s	-4%
`mlk_poly_rej_uniform`	✅	31s	34s	-9%
`mlk_polyvec_add`	✅	26s	26s	+0%
`poly_ntt_native`	✅	26s	35s	-26%
`mlk_ntt_layer`	✅	23s	22s	+5%
`keccakf1600x4_permute_native_x4`	✅	19s	18s	+6%
`polyvec_basemul_acc_montgomery_cached_native`	✅	19s	22s	-14%
`mlk_poly_reduce_native`	✅	15s	14s	+7%
`mlk_indcpa_dec`	✅	12s	11s	+9%
`mlk_keccak_absorb_once_x4`	✅	11s	11s	+0%
`mlk_poly_sub`	✅	9s	8s	+12%
`mlk_keccak_squeeze_once`	✅	8s	9s	-11%
`mlk_keccak_squeezeblocks`	✅	7s	6s	+17%
`mlk_ntt_butterfly_block`	✅	7s	9s	-22%
`mlk_poly_frombytes_native`	✅	7s	10s	-30%
`mlk_poly_frommsg`	✅	7s	9s	-22%
`mlk_poly_rej_uniform_x4`	✅	7s	7s	+0%
`keccakf1600_permute_native`	✅	6s	9s	-33%
`mlk_polymat_permute_bitrev_to_custom`	✅	6s	4s	+50%
`mlk_shake256x4`	✅	6s	4s	+50%
`kem_dec`	✅	5s	5s	+0%
`mlk_fqmul`	✅	5s	6s	-17%
`mlk_poly_mulcache_compute_c`	✅	5s	2s	+150%
`mlk_shake256`	✅	5s	3s	+67%
`kem_enc_derand`	✅	4s	2s	+100%
`kem_keypair`	✅	4s	4s	+0%
`mlk_check_pct`	✅	4s	3s	+33%
`mlk_ct_cmov_zero`	✅	4s	1s	+300%
`mlk_ct_memcmp`	✅	4s	3s	+33%
`mlk_ct_sel_uint8`	✅	4s	4s	+0%
`mlk_gen_matrix`	✅	4s	5s	-20%
`mlk_invntt_layer`	✅	4s	7s	-43%
`mlk_keccak_absorb_once`	✅	4s	5s	-20%
`mlk_poly_cbd_eta1`	✅	4s	4s	+0%
`mlk_poly_mulcache_compute`	✅	4s	5s	-20%
`mlk_poly_reduce_c`	✅	4s	2s	+100%
`mlk_polyvec_mulcache_compute`	✅	4s	3s	+33%
`mlk_scalar_compress_d4`	✅	4s	3s	+33%
`mlk_scalar_compress_d5`	✅	4s	2s	+100%
`mlk_scalar_decompress_d4`	✅	4s	2s	+100%
`ntt_native_x86_64`	✅	4s	4s	+0%
`poly_getnoise_eta1122_4x_native`	✅	4s	4s	+0%
`poly_mulcache_compute_native_aarch64`	✅	4s	3s	+33%
`poly_reduce_native_x86_64`	✅	4s	3s	+33%
`polyvec_basemul_acc_montgomery_cached_k3_native_aarch64`	✅	4s	1s	+300%
`polyvec_basemul_acc_montgomery_cached_k3_native_x86_64`	✅	4s	4s	+0%
`polyvec_basemul_acc_montgomery_cached_k4_native_aarch64`	✅	4s	2s	+100%
`rej_uniform_native`	✅	4s	3s	+33%
`intt_native_aarch64`	✅	3s	2s	+50%
`keccak_f1600_x1_native_aarch64_v84a`	✅	3s	3s	+0%
`kem_check_pk`	✅	3s	4s	-25%
`kem_check_sk`	✅	3s	3s	+0%
`kem_enc`	✅	3s	2s	+50%
`mlk_ct_get_optblocker_i32`	✅	3s	1s	+200%
`mlk_ct_sel_int16`	✅	3s	5s	-40%
`mlk_keccakf1600_permute`	✅	3s	3s	+0%
`mlk_keccakf1600x4_xor_bytes`	✅	3s	2s	+50%
`mlk_poly_add`	✅	3s	1s	+200%
`mlk_poly_compress_du`	✅	3s	5s	-40%
`mlk_poly_decompress_du`	✅	3s	2s	+50%
`mlk_poly_getnoise_eta1122_4x`	✅	3s	3s	+0%
`mlk_poly_getnoise_eta1_4x_native`	✅	3s	2s	+50%
`mlk_poly_tomont_c`	✅	3s	2s	+50%
`mlk_polyvec_basemul_acc_montgomery_cached`	✅	3s	4s	-25%
`mlk_polyvec_compress_du`	✅	3s	3s	+0%
`mlk_polyvec_invntt_tomont`	✅	3s	2s	+50%
`mlk_polyvec_ntt`	✅	3s	2s	+50%
`mlk_rej_uniform`	✅	3s	2s	+50%
`mlk_scalar_compress_d1`	✅	3s	4s	-25%
`mlk_scalar_decompress_d10`	✅	3s	2s	+50%
`mlk_shake128_squeezeblocks`	✅	3s	3s	+0%
`ntt_native_aarch64`	✅	3s	3s	+0%
`nttunpack_native_x86_64`	✅	3s	4s	-25%
`poly_frombytes_native_x86_64`	✅	3s	4s	-25%
`poly_invntt_tomont_native`	✅	3s	2s	+50%
`poly_tobytes_native_aarch64`	✅	3s	1s	+200%
`poly_tomont_native_aarch64`	✅	3s	1s	+200%
`polyvec_basemul_acc_montgomery_cached_k2_native_x86_64`	✅	3s	4s	-25%
`polyvec_basemul_acc_montgomery_cached_k4_native_x86_64`	✅	3s	3s	+0%
`rej_uniform_native_x86_64`	✅	3s	4s	-25%
`sys_check_capability`	✅	3s	1s	+200%
`intt_native_x86_64`	✅	2s	2s	+0%
`keccak_f1600_x4_native_aarch64_v84a`	✅	2s	3s	-33%
`keccak_f1600_x4_native_aarch64_v8a_scalar_hybrid`	✅	2s	2s	+0%
`kem_keypair_derand`	✅	2s	6s	-67%
`mlk_barrett_reduce`	✅	2s	4s	-50%
`mlk_ct_cmask_neg_i16`	✅	2s	2s	+0%
`mlk_ct_get_optblocker_u32`	✅	2s	5s	-60%
`mlk_gen_matrix_serial`	✅	2s	3s	-33%
`mlk_keccakf1600_extract_bytes`	✅	2s	3s	-33%
`mlk_keccakf1600_extract_bytes (big endian)`	✅	2s	1s	+100%
`mlk_keccakf1600_xor_bytes`	✅	2s	2s	+0%
`mlk_keccakf1600_xor_bytes (big endian)`	✅	2s	2s	+0%
`mlk_keccakf1600x4_extract_bytes`	✅	2s	4s	-50%
`mlk_matvec_mul`	✅	2s	2s	+0%
`mlk_montgomery_reduce`	✅	2s	4s	-50%
`mlk_poly_cbd_eta2`	✅	2s	4s	-50%
`mlk_poly_compress_dv`	✅	2s	2s	+0%
`mlk_poly_frombytes`	✅	2s	2s	+0%
`mlk_poly_frombytes_c`	✅	2s	5s	-60%
`mlk_poly_getnoise_eta1_4x`	✅	2s	3s	-33%
`mlk_poly_getnoise_eta2`	✅	2s	5s	-60%
`mlk_poly_invntt_tomont`	✅	2s	2s	+0%
`mlk_poly_invntt_tomont_c`	✅	2s	2s	+0%
`mlk_poly_mulcache_compute_native`	✅	2s	3s	-33%
`mlk_poly_ntt`	✅	2s	2s	+0%
`mlk_poly_ntt_c`	✅	2s	3s	-33%
`mlk_poly_reduce`	✅	2s	1s	+100%
`mlk_poly_tobytes`	✅	2s	2s	+0%
`mlk_poly_tobytes_c`	✅	2s	4s	-50%
`mlk_poly_tomsg`	✅	2s	3s	-33%
`mlk_polyvec_decompress_du`	✅	2s	2s	+0%
`mlk_polyvec_frombytes`	✅	2s	2s	+0%
`mlk_polyvec_permute_bitrev_to_custom`	✅	2s	4s	-50%
`mlk_polyvec_permute_bitrev_to_custom_native`	✅	2s	4s	-50%
`mlk_polyvec_reduce`	✅	2s	4s	-50%
`mlk_polyvec_tobytes`	✅	2s	2s	+0%
`mlk_polyvec_tomont`	✅	2s	2s	+0%
`mlk_scalar_compress_d10`	✅	2s	2s	+0%
`mlk_scalar_decompress_d11`	✅	2s	4s	-50%
`mlk_scalar_decompress_d5`	✅	2s	1s	+100%
`mlk_scalar_signed_to_unsigned_q`	✅	2s	1s	+100%
`mlk_sha3_256`	✅	2s	1s	+100%
`mlk_sha3_512`	✅	2s	2s	+0%
`mlk_shake128x4_squeezeblocks`	✅	2s	3s	-33%
`poly_mulcache_compute_native_x86_64`	✅	2s	1s	+100%
`poly_tobytes_native_x86_64`	✅	2s	1s	+100%
`poly_tomont_native_x86_64`	✅	2s	2s	+0%
`polyvec_basemul_acc_montgomery_cached_k2_native_aarch64`	✅	2s	2s	+0%
`rej_uniform_native_aarch64`	✅	2s	5s	-60%
`keccak_f1600_x1_native_aarch64`	✅	1s	1s	+0%
`keccak_f1600_x4_native_aarch64_v8a_v84a_scalar_hybrid`	✅	1s	1s	+0%
`mlk_ct_cmask_nonzero_u16`	✅	1s	4s	-75%
`mlk_ct_cmask_nonzero_u8`	✅	1s	1s	+0%
`mlk_ct_get_optblocker_u8`	✅	1s	4s	-75%
`mlk_keccakf1600x4_permute`	✅	1s	5s	-80%
`mlk_poly_decompress_dv`	✅	1s	2s	-50%
`mlk_poly_tobytes_native`	✅	1s	2s	-50%
`mlk_poly_tomont`	✅	1s	2s	-50%
`mlk_poly_tomont_native`	✅	1s	2s	-50%
`mlk_scalar_compress_d11`	✅	1s	2s	-50%
`mlk_shake128_absorb_once`	✅	1s	1s	+0%
`mlk_shake128x4_absorb_once`	✅	1s	2s	-50%
`mlk_value_barrier_i32`	✅	1s	3s	-67%
`mlk_value_barrier_u32`	✅	1s	1s	+0%
`mlk_value_barrier_u8`	✅	1s	1s	+0%
`poly_reduce_native_aarch64`	✅	1s	1s	+0%

oqs-bot · 2026-02-02T07:20:39Z

CBMC Results (ML-KEM-768)

Full Results (151 proofs)

Proof	Status	Current	Previous	Change
`TOTAL`	✅	1355s	1299s	+4.3%
`mlk_indcpa_enc`	✅	190s	174s	+9%
`mlk_keccak_squeezeblocks_x4`	✅	155s	142s	+9%
`mlk_indcpa_keypair_derand`	✅	154s	155s	-1%
`mlk_rej_uniform_c`	✅	94s	81s	+16%
`mlk_polyvec_basemul_acc_montgomery_cached_c`	✅	91s	81s	+12%
`polyvec_basemul_acc_montgomery_cached_native`	✅	56s	59s	-5%
`mlk_poly_rej_uniform`	✅	39s	39s	+0%
`mlk_ntt_layer`	✅	36s	39s	-8%
`poly_ntt_native`	✅	30s	27s	+11%
`mlk_polyvec_add`	✅	28s	30s	-7%
`keccakf1600x4_permute_native_x4`	✅	19s	17s	+12%
`mlk_indcpa_dec`	✅	16s	16s	+0%
`mlk_poly_reduce_native`	✅	15s	16s	-6%
`mlk_poly_frombytes_native`	✅	11s	11s	+0%
`mlk_keccak_absorb_once_x4`	✅	10s	11s	-9%
`mlk_poly_frommsg`	✅	10s	7s	+43%
`mlk_poly_sub`	✅	10s	11s	-9%
`mlk_ntt_butterfly_block`	✅	9s	9s	+0%
`poly_frombytes_native_x86_64`	✅	9s	6s	+50%
`mlk_polymat_permute_bitrev_to_custom`	✅	8s	6s	+33%
`mlk_poly_rej_uniform_x4`	✅	7s	7s	+0%
`kem_dec`	✅	6s	7s	-14%
`mlk_fqmul`	✅	6s	7s	-14%
`mlk_invntt_layer`	✅	6s	6s	+0%
`mlk_keccak_squeeze_once`	✅	6s	6s	+0%
`mlk_keccak_squeezeblocks`	✅	6s	8s	-25%
`intt_native_x86_64`	✅	5s	2s	+150%
`keccakf1600_permute_native`	✅	5s	5s	+0%
`kem_enc`	✅	5s	3s	+67%
`kem_keypair`	✅	5s	2s	+150%
`mlk_keccak_absorb_once`	✅	5s	4s	+25%
`mlk_keccakf1600_extract_bytes (big endian)`	✅	5s	4s	+25%
`mlk_poly_frombytes`	✅	5s	1s	+400%
`intt_native_aarch64`	✅	4s	3s	+33%
`mlk_gen_matrix_serial`	✅	4s	3s	+33%
`mlk_keccakf1600_permute`	✅	4s	4s	+0%
`mlk_poly_compress_du`	✅	4s	4s	+0%
`mlk_poly_compress_dv`	✅	4s	2s	+100%
`mlk_poly_frombytes_c`	✅	4s	4s	+0%
`mlk_poly_getnoise_eta1_4x`	✅	4s	4s	+0%
`mlk_poly_getnoise_eta2`	✅	4s	4s	+0%
`mlk_poly_mulcache_compute_native`	✅	4s	3s	+33%
`mlk_poly_tobytes_native`	✅	4s	1s	+300%
`mlk_poly_tomont`	✅	4s	1s	+300%
`mlk_poly_tomsg`	✅	4s	4s	+0%
`mlk_polyvec_basemul_acc_montgomery_cached`	✅	4s	3s	+33%
`mlk_scalar_compress_d10`	✅	4s	2s	+100%
`mlk_scalar_compress_d11`	✅	4s	2s	+100%
`mlk_scalar_decompress_d4`	✅	4s	4s	+0%
`mlk_shake128_absorb_once`	✅	4s	2s	+100%
`mlk_value_barrier_u8`	✅	4s	1s	+300%
`ntt_native_aarch64`	✅	4s	3s	+33%
`keccak_f1600_x1_native_aarch64`	✅	3s	2s	+50%
`keccak_f1600_x1_native_aarch64_v84a`	✅	3s	1s	+200%
`keccak_f1600_x4_native_aarch64_v84a`	✅	3s	1s	+200%
`keccak_f1600_x4_native_aarch64_v8a_v84a_scalar_hybrid`	✅	3s	2s	+50%
`kem_check_pk`	✅	3s	3s	+0%
`kem_enc_derand`	✅	3s	3s	+0%
`mlk_barrett_reduce`	✅	3s	2s	+50%
`mlk_check_pct`	✅	3s	4s	-25%
`mlk_ct_cmask_neg_i16`	✅	3s	2s	+50%
`mlk_ct_get_optblocker_i32`	✅	3s	3s	+0%
`mlk_gen_matrix`	✅	3s	3s	+0%
`mlk_keccakf1600_xor_bytes`	✅	3s	2s	+50%
`mlk_matvec_mul`	✅	3s	2s	+50%
`mlk_poly_decompress_dv`	✅	3s	2s	+50%
`mlk_poly_getnoise_eta1_4x_native`	✅	3s	5s	-40%
`mlk_poly_invntt_tomont`	✅	3s	2s	+50%
`mlk_poly_mulcache_compute_c`	✅	3s	1s	+200%
`mlk_poly_ntt`	✅	3s	3s	+0%
`mlk_poly_reduce`	✅	3s	2s	+50%
`mlk_poly_tobytes_c`	✅	3s	3s	+0%
`mlk_poly_tomont_c`	✅	3s	6s	-50%
`mlk_polyvec_invntt_tomont`	✅	3s	2s	+50%
`mlk_polyvec_ntt`	✅	3s	2s	+50%
`mlk_polyvec_permute_bitrev_to_custom_native`	✅	3s	3s	+0%
`mlk_polyvec_reduce`	✅	3s	1s	+200%
`mlk_polyvec_tobytes`	✅	3s	2s	+50%
`mlk_rej_uniform`	✅	3s	1s	+200%
`mlk_sha3_256`	✅	3s	4s	-25%
`mlk_shake128x4_absorb_once`	✅	3s	1s	+200%
`mlk_shake256x4`	✅	3s	5s	-40%
`ntt_native_x86_64`	✅	3s	3s	+0%
`nttunpack_native_x86_64`	✅	3s	2s	+50%
`poly_getnoise_eta1122_4x_native`	✅	3s	2s	+50%
`poly_mulcache_compute_native_x86_64`	✅	3s	2s	+50%
`poly_reduce_native_x86_64`	✅	3s	2s	+50%
`poly_tobytes_native_x86_64`	✅	3s	2s	+50%
`poly_tomont_native_aarch64`	✅	3s	3s	+0%
`polyvec_basemul_acc_montgomery_cached_k2_native_x86_64`	✅	3s	3s	+0%
`polyvec_basemul_acc_montgomery_cached_k3_native_aarch64`	✅	3s	3s	+0%
`rej_uniform_native_aarch64`	✅	3s	3s	+0%
`keccak_f1600_x4_native_aarch64_v8a_scalar_hybrid`	✅	2s	1s	+100%
`kem_check_sk`	✅	2s	2s	+0%
`kem_keypair_derand`	✅	2s	3s	-33%
`mlk_ct_cmask_nonzero_u16`	✅	2s	2s	+0%
`mlk_ct_cmov_zero`	✅	2s	3s	-33%
`mlk_ct_memcmp`	✅	2s	3s	-33%
`mlk_ct_sel_int16`	✅	2s	4s	-50%
`mlk_keccakf1600_extract_bytes`	✅	2s	1s	+100%
`mlk_keccakf1600_xor_bytes (big endian)`	✅	2s	1s	+100%
`mlk_keccakf1600x4_extract_bytes`	✅	2s	3s	-33%
`mlk_keccakf1600x4_permute`	✅	2s	3s	-33%
`mlk_keccakf1600x4_xor_bytes`	✅	2s	2s	+0%
`mlk_poly_cbd_eta1`	✅	2s	1s	+100%
`mlk_poly_decompress_du`	✅	2s	2s	+0%
`mlk_poly_getnoise_eta1122_4x`	✅	2s	2s	+0%
`mlk_poly_invntt_tomont_c`	✅	2s	1s	+100%
`mlk_poly_mulcache_compute`	✅	2s	1s	+100%
`mlk_poly_ntt_c`	✅	2s	2s	+0%
`mlk_poly_tomont_native`	✅	2s	2s	+0%
`mlk_polyvec_compress_du`	✅	2s	5s	-60%
`mlk_polyvec_decompress_du`	✅	2s	2s	+0%
`mlk_polyvec_mulcache_compute`	✅	2s	2s	+0%
`mlk_polyvec_permute_bitrev_to_custom`	✅	2s	2s	+0%
`mlk_scalar_compress_d1`	✅	2s	1s	+100%
`mlk_scalar_compress_d5`	✅	2s	2s	+0%
`mlk_scalar_decompress_d10`	✅	2s	2s	+0%
`mlk_scalar_signed_to_unsigned_q`	✅	2s	1s	+100%
`mlk_sha3_512`	✅	2s	1s	+100%
`mlk_shake128_squeezeblocks`	✅	2s	1s	+100%
`mlk_value_barrier_i32`	✅	2s	4s	-50%
`poly_mulcache_compute_native_aarch64`	✅	2s	5s	-60%
`poly_reduce_native_aarch64`	✅	2s	3s	-33%
`poly_tobytes_native_aarch64`	✅	2s	4s	-50%
`poly_tomont_native_x86_64`	✅	2s	4s	-50%
`polyvec_basemul_acc_montgomery_cached_k3_native_x86_64`	✅	2s	2s	+0%
`polyvec_basemul_acc_montgomery_cached_k4_native_aarch64`	✅	2s	2s	+0%
`rej_uniform_native`	✅	2s	4s	-50%
`rej_uniform_native_x86_64`	✅	2s	4s	-50%
`mlk_ct_cmask_nonzero_u8`	✅	1s	3s	-67%
`mlk_ct_get_optblocker_u32`	✅	1s	3s	-67%
`mlk_ct_get_optblocker_u8`	✅	1s	2s	-50%
`mlk_ct_sel_uint8`	✅	1s	2s	-50%
`mlk_montgomery_reduce`	✅	1s	3s	-67%
`mlk_poly_add`	✅	1s	4s	-75%
`mlk_poly_cbd_eta2`	✅	1s	1s	+0%
`mlk_poly_reduce_c`	✅	1s	3s	-67%
`mlk_poly_tobytes`	✅	1s	3s	-67%
`mlk_polyvec_frombytes`	✅	1s	1s	+0%
`mlk_polyvec_tomont`	✅	1s	1s	+0%
`mlk_scalar_compress_d4`	✅	1s	3s	-67%
`mlk_scalar_decompress_d11`	✅	1s	3s	-67%
`mlk_scalar_decompress_d5`	✅	1s	3s	-67%
`mlk_shake128x4_squeezeblocks`	✅	1s	2s	-50%
`mlk_shake256`	✅	1s	5s	-80%
`mlk_value_barrier_u32`	✅	1s	3s	-67%
`poly_invntt_tomont_native`	✅	1s	1s	+0%
`polyvec_basemul_acc_montgomery_cached_k2_native_aarch64`	✅	1s	2s	-50%
`polyvec_basemul_acc_montgomery_cached_k4_native_x86_64`	✅	1s	4s	-75%
`sys_check_capability`	✅	1s	4s	-75%

oqs-bot

Mac Mini (M1, 2020) benchmarks

Details

Benchmark suite	Current: `caa0cd1`	Previous: `4f13e12`	Ratio
`ML-KEM-512 keypair`	`12327` cycles	`12327` cycles	`1`
`ML-KEM-512 encaps`	`15033` cycles	`15029` cycles	`1.00`
`ML-KEM-512 decaps`	`19609` cycles	`19607` cycles	`1.00`
`ML-KEM-768 keypair`	`21092` cycles	`21092` cycles	`1`
`ML-KEM-768 encaps`	`23862` cycles	`23863` cycles	`1.00`
`ML-KEM-768 decaps`	`30443` cycles	`30443` cycles	`1`
`ML-KEM-1024 keypair`	`30375` cycles	`30376` cycles	`1.00`
`ML-KEM-1024 encaps`	`34642` cycles	`34642` cycles	`1`
`ML-KEM-1024 decaps`	`44267` cycles	`44279` cycles	`1.00`

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

oqs-bot

ppc64le (POWER10) benchmarks

Details

Benchmark suite	Current: `caa0cd1`	Previous: `4f13e12`	Ratio
`ML-KEM-512 keypair`	`59312` cycles	`61729` cycles	`0.96`
`ML-KEM-512 encaps`	`71978` cycles	`76321` cycles	`0.94`
`ML-KEM-512 decaps`	`91662` cycles	`96900` cycles	`0.95`
`ML-KEM-768 keypair`	`98528` cycles	`101843` cycles	`0.97`
`ML-KEM-768 encaps`	`114864` cycles	`118989` cycles	`0.97`
`ML-KEM-768 decaps`	`140397` cycles	`145386` cycles	`0.97`
`ML-KEM-1024 keypair`	`148781` cycles	`153347` cycles	`0.97`
`ML-KEM-1024 encaps`	`167693` cycles	`172326` cycles	`0.97`
`ML-KEM-1024 decaps`	`198974` cycles	`204380` cycles	`0.97`

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

oqs-bot · 2026-02-02T07:22:28Z

CBMC Results (ML-KEM-1024)

Full Results (151 proofs)

Proof	Status	Current	Previous	Change
`TOTAL`	✅	1594s	1740s	-8.4%
`mlk_indcpa_keypair_derand`	✅	323s	349s	-7%
`mlk_indcpa_enc`	✅	211s	255s	-17%
`mlk_keccak_squeezeblocks_x4`	✅	135s	143s	-6%
`polyvec_basemul_acc_montgomery_cached_native`	✅	127s	138s	-8%
`mlk_rej_uniform_c`	✅	73s	107s	-32%
`mlk_polyvec_basemul_acc_montgomery_cached_c`	✅	65s	58s	+12%
`mlk_poly_rej_uniform`	✅	44s	41s	+7%
`mlk_ntt_layer`	✅	36s	36s	+0%
`mlk_poly_decompress_dv`	✅	35s	43s	-19%
`mlk_indcpa_dec`	✅	29s	25s	+16%
`poly_ntt_native`	✅	25s	28s	-11%
`keccakf1600x4_permute_native_x4`	✅	18s	19s	-5%
`mlk_poly_reduce_native`	✅	16s	17s	-6%
`mlk_polyvec_ntt`	✅	14s	15s	-7%
`mlk_poly_sub`	✅	11s	8s	+38%
`mlk_fqmul`	✅	10s	6s	+67%
`mlk_keccak_absorb_once_x4`	✅	10s	13s	-23%
`mlk_ntt_butterfly_block`	✅	10s	13s	-23%
`mlk_polyvec_add`	✅	10s	9s	+11%
`mlk_poly_compress_du`	✅	9s	9s	+0%
`mlk_poly_frommsg`	✅	9s	8s	+12%
`mlk_keccak_squeeze_once`	✅	8s	10s	-20%
`mlk_keccak_squeezeblocks`	✅	7s	7s	+0%
`mlk_poly_frombytes_native`	✅	7s	9s	-22%
`mlk_poly_rej_uniform_x4`	✅	7s	7s	+0%
`keccakf1600_permute_native`	✅	6s	5s	+20%
`kem_dec`	✅	6s	7s	-14%
`mlk_gen_matrix`	✅	6s	8s	-25%
`mlk_invntt_layer`	✅	6s	6s	+0%
`mlk_shake256x4`	✅	6s	6s	+0%
`mlk_ct_cmask_nonzero_u16`	✅	5s	3s	+67%
`mlk_keccak_absorb_once`	✅	5s	5s	+0%
`mlk_keccakf1600x4_xor_bytes`	✅	5s	3s	+67%
`mlk_montgomery_reduce`	✅	5s	2s	+150%
`mlk_poly_ntt_c`	✅	5s	3s	+67%
`mlk_poly_reduce_c`	✅	5s	2s	+150%
`mlk_polyvec_decompress_du`	✅	5s	3s	+67%
`kem_check_pk`	✅	4s	4s	+0%
`mlk_check_pct`	✅	4s	4s	+0%
`mlk_ct_sel_int16`	✅	4s	2s	+100%
`mlk_gen_matrix_serial`	✅	4s	6s	-33%
`mlk_poly_getnoise_eta1122_4x`	✅	4s	2s	+100%
`mlk_poly_getnoise_eta1_4x`	✅	4s	3s	+33%
`mlk_poly_tomsg`	✅	4s	3s	+33%
`mlk_polymat_permute_bitrev_to_custom`	✅	4s	4s	+0%
`mlk_polyvec_basemul_acc_montgomery_cached`	✅	4s	2s	+100%
`mlk_polyvec_compress_du`	✅	4s	4s	+0%
`mlk_polyvec_mulcache_compute`	✅	4s	3s	+33%
`mlk_polyvec_permute_bitrev_to_custom_native`	✅	4s	5s	-20%
`mlk_scalar_compress_d11`	✅	4s	2s	+100%
`mlk_value_barrier_u8`	✅	4s	2s	+100%
`rej_uniform_native`	✅	4s	3s	+33%
`intt_native_x86_64`	✅	3s	2s	+50%
`keccak_f1600_x4_native_aarch64_v84a`	✅	3s	3s	+0%
`keccak_f1600_x4_native_aarch64_v8a_scalar_hybrid`	✅	3s	2s	+50%
`kem_enc`	✅	3s	3s	+0%
`mlk_barrett_reduce`	✅	3s	2s	+50%
`mlk_ct_get_optblocker_u32`	✅	3s	4s	-25%
`mlk_keccakf1600_extract_bytes (big endian)`	✅	3s	2s	+50%
`mlk_keccakf1600_permute`	✅	3s	4s	-25%
`mlk_keccakf1600_xor_bytes (big endian)`	✅	3s	2s	+50%
`mlk_keccakf1600x4_permute`	✅	3s	3s	+0%
`mlk_poly_add`	✅	3s	3s	+0%
`mlk_poly_cbd_eta1`	✅	3s	3s	+0%
`mlk_poly_frombytes`	✅	3s	3s	+0%
`mlk_poly_getnoise_eta1_4x_native`	✅	3s	3s	+0%
`mlk_poly_invntt_tomont`	✅	3s	3s	+0%
`mlk_poly_mulcache_compute_c`	✅	3s	2s	+50%
`mlk_poly_tobytes`	✅	3s	1s	+200%
`mlk_poly_tobytes_c`	✅	3s	1s	+200%
`mlk_poly_tomont`	✅	3s	2s	+50%
`mlk_poly_tomont_native`	✅	3s	3s	+0%
`mlk_polyvec_frombytes`	✅	3s	2s	+50%
`mlk_polyvec_permute_bitrev_to_custom`	✅	3s	2s	+50%
`mlk_polyvec_tomont`	✅	3s	3s	+0%
`mlk_rej_uniform`	✅	3s	2s	+50%
`mlk_scalar_compress_d5`	✅	3s	1s	+200%
`mlk_sha3_512`	✅	3s	1s	+200%
`mlk_shake128_absorb_once`	✅	3s	2s	+50%
`mlk_shake128_squeezeblocks`	✅	3s	2s	+50%
`mlk_shake256`	✅	3s	4s	-25%
`mlk_value_barrier_i32`	✅	3s	1s	+200%
`mlk_value_barrier_u32`	✅	3s	1s	+200%
`nttunpack_native_x86_64`	✅	3s	1s	+200%
`poly_frombytes_native_x86_64`	✅	3s	6s	-50%
`poly_invntt_tomont_native`	✅	3s	2s	+50%
`poly_mulcache_compute_native_aarch64`	✅	3s	3s	+0%
`poly_reduce_native_x86_64`	✅	3s	3s	+0%
`polyvec_basemul_acc_montgomery_cached_k3_native_x86_64`	✅	3s	3s	+0%
`rej_uniform_native_aarch64`	✅	3s	5s	-40%
`rej_uniform_native_x86_64`	✅	3s	3s	+0%
`intt_native_aarch64`	✅	2s	3s	-33%
`keccak_f1600_x1_native_aarch64`	✅	2s	4s	-50%
`keccak_f1600_x1_native_aarch64_v84a`	✅	2s	4s	-50%
`kem_check_sk`	✅	2s	3s	-33%
`kem_enc_derand`	✅	2s	1s	+100%
`kem_keypair`	✅	2s	1s	+100%
`kem_keypair_derand`	✅	2s	4s	-50%
`mlk_ct_cmov_zero`	✅	2s	1s	+100%
`mlk_ct_get_optblocker_i32`	✅	2s	3s	-33%
`mlk_ct_get_optblocker_u8`	✅	2s	3s	-33%
`mlk_ct_memcmp`	✅	2s	2s	+0%
`mlk_ct_sel_uint8`	✅	2s	5s	-60%
`mlk_keccakf1600_xor_bytes`	✅	2s	3s	-33%
`mlk_matvec_mul`	✅	2s	5s	-60%
`mlk_poly_cbd_eta2`	✅	2s	3s	-33%
`mlk_poly_compress_dv`	✅	2s	2s	+0%
`mlk_poly_decompress_du`	✅	2s	3s	-33%
`mlk_poly_invntt_tomont_c`	✅	2s	2s	+0%
`mlk_poly_mulcache_compute_native`	✅	2s	3s	-33%
`mlk_poly_reduce`	✅	2s	2s	+0%
`mlk_poly_tobytes_native`	✅	2s	7s	-71%
`mlk_polyvec_invntt_tomont`	✅	2s	2s	+0%
`mlk_polyvec_reduce`	✅	2s	3s	-33%
`mlk_polyvec_tobytes`	✅	2s	3s	-33%
`mlk_scalar_compress_d1`	✅	2s	1s	+100%
`mlk_scalar_compress_d10`	✅	2s	2s	+0%
`mlk_scalar_decompress_d11`	✅	2s	3s	-33%
`mlk_scalar_decompress_d4`	✅	2s	3s	-33%
`mlk_scalar_decompress_d5`	✅	2s	3s	-33%
`mlk_scalar_signed_to_unsigned_q`	✅	2s	2s	+0%
`mlk_sha3_256`	✅	2s	3s	-33%
`mlk_shake128x4_absorb_once`	✅	2s	4s	-50%
`mlk_shake128x4_squeezeblocks`	✅	2s	3s	-33%
`ntt_native_aarch64`	✅	2s	3s	-33%
`ntt_native_x86_64`	✅	2s	3s	-33%
`poly_mulcache_compute_native_x86_64`	✅	2s	2s	+0%
`poly_tobytes_native_x86_64`	✅	2s	3s	-33%
`poly_tomont_native_aarch64`	✅	2s	3s	-33%
`poly_tomont_native_x86_64`	✅	2s	4s	-50%
`polyvec_basemul_acc_montgomery_cached_k2_native_x86_64`	✅	2s	3s	-33%
`polyvec_basemul_acc_montgomery_cached_k3_native_aarch64`	✅	2s	2s	+0%
`polyvec_basemul_acc_montgomery_cached_k4_native_x86_64`	✅	2s	2s	+0%
`sys_check_capability`	✅	2s	2s	+0%
`keccak_f1600_x4_native_aarch64_v8a_v84a_scalar_hybrid`	✅	1s	2s	-50%
`mlk_ct_cmask_neg_i16`	✅	1s	4s	-75%
`mlk_ct_cmask_nonzero_u8`	✅	1s	3s	-67%
`mlk_keccakf1600_extract_bytes`	✅	1s	3s	-67%
`mlk_keccakf1600x4_extract_bytes`	✅	1s	2s	-50%
`mlk_poly_frombytes_c`	✅	1s	3s	-67%
`mlk_poly_getnoise_eta2`	✅	1s	2s	-50%
`mlk_poly_mulcache_compute`	✅	1s	1s	+0%
`mlk_poly_ntt`	✅	1s	2s	-50%
`mlk_poly_tomont_c`	✅	1s	2s	-50%
`mlk_scalar_compress_d4`	✅	1s	2s	-50%
`mlk_scalar_decompress_d10`	✅	1s	2s	-50%
`poly_getnoise_eta1122_4x_native`	✅	1s	3s	-67%
`poly_reduce_native_aarch64`	✅	1s	3s	-67%
`poly_tobytes_native_aarch64`	✅	1s	2s	-50%
`polyvec_basemul_acc_montgomery_cached_k2_native_aarch64`	✅	1s	4s	-75%
`polyvec_basemul_acc_montgomery_cached_k4_native_aarch64`	✅	1s	4s	-75%

oqs-bot

Intel Xeon 4th gen (c7i)

Details

Benchmark suite	Current: `caa0cd1`	Previous: `4f13e12`	Ratio
`ML-KEM-512 keypair`	`9682` cycles	`9677` cycles	`1.00`
`ML-KEM-512 encaps`	`11475` cycles	`11476` cycles	`1.00`
`ML-KEM-512 decaps`	`15374` cycles	`15377` cycles	`1.00`
`ML-KEM-768 keypair`	`16409` cycles	`16401` cycles	`1.00`
`ML-KEM-768 encaps`	`17833` cycles	`17841` cycles	`1.00`
`ML-KEM-768 decaps`	`23483` cycles	`23478` cycles	`1.00`
`ML-KEM-1024 keypair`	`22308` cycles	`22387` cycles	`1.00`
`ML-KEM-1024 encaps`	`24608` cycles	`24660` cycles	`1.00`
`ML-KEM-1024 decaps`	`32264` cycles	`32524` cycles	`0.99`

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

oqs-bot

AMD EPYC 3rd gen (c6a)

Details

Benchmark suite	Current: `caa0cd1`	Previous: `4f13e12`	Ratio
`ML-KEM-512 keypair`	`16707` cycles	`16688` cycles	`1.00`
`ML-KEM-512 encaps`	`18360` cycles	`18355` cycles	`1.00`
`ML-KEM-512 decaps`	`23724` cycles	`23724` cycles	`1`
`ML-KEM-768 keypair`	`28440` cycles	`28460` cycles	`1.00`
`ML-KEM-768 encaps`	`29786` cycles	`29796` cycles	`1.00`
`ML-KEM-768 decaps`	`37653` cycles	`37637` cycles	`1.00`
`ML-KEM-1024 keypair`	`41178` cycles	`41221` cycles	`1.00`
`ML-KEM-1024 encaps`	`43517` cycles	`43491` cycles	`1.00`
`ML-KEM-1024 decaps`	`53826` cycles	`53972` cycles	`1.00`

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

oqs-bot

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

Details

Benchmark suite	Current: `caa0cd1`	Previous: `4f13e12`	Ratio
`ML-KEM-512 keypair`	`28489` cycles	`28478` cycles	`1.00`
`ML-KEM-512 encaps`	`35837` cycles	`35791` cycles	`1.00`
`ML-KEM-512 decaps`	`45395` cycles	`45426` cycles	`1.00`
`ML-KEM-768 keypair`	`45944` cycles	`45975` cycles	`1.00`
`ML-KEM-768 encaps`	`56221` cycles	`56333` cycles	`1.00`
`ML-KEM-768 decaps`	`69406` cycles	`69448` cycles	`1.00`
`ML-KEM-1024 keypair`	`71696` cycles	`71673` cycles	`1.00`
`ML-KEM-1024 encaps`	`84467` cycles	`84546` cycles	`1.00`
`ML-KEM-1024 decaps`	`101594` cycles	`101646` 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: `caa0cd1`	Previous: `4f13e12`	Ratio
`ML-KEM-512 keypair`	`12031` cycles	`12068` cycles	`1.00`
`ML-KEM-512 encaps`	`13173` cycles	`13187` cycles	`1.00`
`ML-KEM-512 decaps`	`18051` cycles	`18046` cycles	`1.00`
`ML-KEM-768 keypair`	`20536` cycles	`20539` cycles	`1.00`
`ML-KEM-768 encaps`	`21536` cycles	`21522` cycles	`1.00`
`ML-KEM-768 decaps`	`28659` cycles	`28651` cycles	`1.00`
`ML-KEM-1024 keypair`	`27817` cycles	`27844` cycles	`1.00`
`ML-KEM-1024 encaps`	`30108` cycles	`30078` cycles	`1.00`
`ML-KEM-1024 decaps`	`39365` cycles	`39348` 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: `caa0cd1`	Previous: `4f13e12`	Ratio
`ML-KEM-512 keypair`	`16376` cycles	`16370` cycles	`1.00`
`ML-KEM-512 encaps`	`18694` cycles	`18691` cycles	`1.00`
`ML-KEM-512 decaps`	`25288` cycles	`25287` cycles	`1.00`
`ML-KEM-768 keypair`	`28869` cycles	`28721` cycles	`1.01`
`ML-KEM-768 encaps`	`29812` cycles	`29774` cycles	`1.00`
`ML-KEM-768 decaps`	`39302` cycles	`39296` cycles	`1.00`
`ML-KEM-1024 keypair`	`37698` cycles	`37693` cycles	`1.00`
`ML-KEM-1024 encaps`	`40683` cycles	`40687` cycles	`1.00`
`ML-KEM-1024 decaps`	`54430` cycles	`54408` cycles	`1.00`

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

oqs-bot

Graviton4

Details

Benchmark suite	Current: `caa0cd1`	Previous: `4f13e12`	Ratio
`ML-KEM-512 keypair`	`17688` cycles	`17705` cycles	`1.00`
`ML-KEM-512 encaps`	`20696` cycles	`20703` cycles	`1.00`
`ML-KEM-512 decaps`	`27115` cycles	`27134` cycles	`1.00`
`ML-KEM-768 keypair`	`30011` cycles	`30007` cycles	`1.00`
`ML-KEM-768 encaps`	`32868` cycles	`32800` cycles	`1.00`
`ML-KEM-768 decaps`	`42068` cycles	`42060` cycles	`1.00`
`ML-KEM-1024 keypair`	`43909` cycles	`43921` cycles	`1.00`
`ML-KEM-1024 encaps`	`48937` cycles	`48929` cycles	`1.00`
`ML-KEM-1024 decaps`	`61484` cycles	`61492` 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: `caa0cd1`	Previous: `4f13e12`	Ratio
`ML-KEM-512 keypair`	`50861` cycles	`50678` cycles	`1.00`
`ML-KEM-512 encaps`	`58671` cycles	`58487` cycles	`1.00`
`ML-KEM-512 decaps`	`75019` cycles	`75583` cycles	`0.99`
`ML-KEM-768 keypair`	`86459` cycles	`85899` cycles	`1.01`
`ML-KEM-768 encaps`	`94396` cycles	`94206` cycles	`1.00`
`ML-KEM-768 decaps`	`117789` cycles	`117678` cycles	`1.00`
`ML-KEM-1024 keypair`	`130502` cycles	`130831` cycles	`1.00`
`ML-KEM-1024 encaps`	`142312` cycles	`142700` cycles	`1.00`
`ML-KEM-1024 decaps`	`173742` cycles	`174993` cycles	`0.99`

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

oqs-bot

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

Details

Benchmark suite	Current: `caa0cd1`	Previous: `4f13e12`	Ratio
`ML-KEM-512 keypair`	`40486` cycles	`40460` cycles	`1.00`
`ML-KEM-512 encaps`	`48363` cycles	`48311` cycles	`1.00`
`ML-KEM-512 decaps`	`62480` cycles	`62448` cycles	`1.00`
`ML-KEM-768 keypair`	`63606` cycles	`63563` cycles	`1.00`
`ML-KEM-768 encaps`	`74772` cycles	`74706` cycles	`1.00`
`ML-KEM-768 decaps`	`93207` cycles	`93714` cycles	`0.99`
`ML-KEM-1024 keypair`	`94991` cycles	`94935` cycles	`1.00`
`ML-KEM-1024 encaps`	`109140` cycles	`108977` cycles	`1.00`
`ML-KEM-1024 decaps`	`131959` cycles	`131836` cycles	`1.00`

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

oqs-bot

Graviton3

Details

Benchmark suite	Current: `caa0cd1`	Previous: `4f13e12`	Ratio
`ML-KEM-512 keypair`	`18743` cycles	`18765` cycles	`1.00`
`ML-KEM-512 encaps`	`22035` cycles	`22041` cycles	`1.00`
`ML-KEM-512 decaps`	`29009` cycles	`29039` cycles	`1.00`
`ML-KEM-768 keypair`	`31809` cycles	`31803` cycles	`1.00`
`ML-KEM-768 encaps`	`35008` cycles	`34950` cycles	`1.00`
`ML-KEM-768 decaps`	`45037` cycles	`45049` cycles	`1.00`
`ML-KEM-1024 keypair`	`46364` cycles	`46362` cycles	`1.00`
`ML-KEM-1024 encaps`	`51753` cycles	`51757` cycles	`1.00`
`ML-KEM-1024 decaps`	`65256` cycles	`65267` 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: `caa0cd1`	Previous: `4f13e12`	Ratio
`ML-KEM-512 keypair`	`36517` cycles	`36504` cycles	`1.00`
`ML-KEM-512 encaps`	`43004` cycles	`42998` cycles	`1.00`
`ML-KEM-512 decaps`	`55683` cycles	`55652` cycles	`1.00`
`ML-KEM-768 keypair`	`58467` cycles	`58471` cycles	`1.00`
`ML-KEM-768 encaps`	`67423` cycles	`67393` cycles	`1.00`
`ML-KEM-768 decaps`	`84621` cycles	`84354` cycles	`1.00`
`ML-KEM-1024 keypair`	`88662` cycles	`88681` cycles	`1.00`
`ML-KEM-1024 encaps`	`98888` cycles	`98925` cycles	`1.00`
`ML-KEM-1024 decaps`	`120514` cycles	`120537` cycles	`1.00`

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

oqs-bot

Graviton4 (no-opt)

Details

Benchmark suite	Current: `caa0cd1`	Previous: `4f13e12`	Ratio
`ML-KEM-512 keypair`	`35483` cycles	`35502` cycles	`1.00`
`ML-KEM-512 encaps`	`40240` cycles	`40238` cycles	`1.00`
`ML-KEM-512 decaps`	`51247` cycles	`51243` cycles	`1.00`
`ML-KEM-768 keypair`	`56776` cycles	`56809` cycles	`1.00`
`ML-KEM-768 encaps`	`64650` cycles	`64725` cycles	`1.00`
`ML-KEM-768 decaps`	`78974` cycles	`79037` cycles	`1.00`
`ML-KEM-1024 keypair`	`88039` cycles	`88033` cycles	`1.00`
`ML-KEM-1024 encaps`	`97181` cycles	`97180` cycles	`1.00`
`ML-KEM-1024 decaps`	`116111` cycles	`116099` 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: `caa0cd1`	Previous: `4f13e12`	Ratio
`ML-KEM-512 keypair`	`45830` cycles	`45812` cycles	`1.00`
`ML-KEM-512 encaps`	`54795` cycles	`54782` cycles	`1.00`
`ML-KEM-512 decaps`	`70332` cycles	`70305` cycles	`1.00`
`ML-KEM-768 keypair`	`73824` cycles	`73801` cycles	`1.00`
`ML-KEM-768 encaps`	`85316` cycles	`85255` cycles	`1.00`
`ML-KEM-768 decaps`	`106283` cycles	`106271` cycles	`1.00`
`ML-KEM-1024 keypair`	`111828` cycles	`111786` cycles	`1.00`
`ML-KEM-1024 encaps`	`125970` cycles	`125915` cycles	`1.00`
`ML-KEM-1024 decaps`	`151780` cycles	`151735` cycles	`1.00`

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

oqs-bot

Graviton3 (no-opt)

Details

Benchmark suite	Current: `caa0cd1`	Previous: `4f13e12`	Ratio
`ML-KEM-512 keypair`	`39042` cycles	`39062` cycles	`1.00`
`ML-KEM-512 encaps`	`44640` cycles	`44635` cycles	`1.00`
`ML-KEM-512 decaps`	`56732` cycles	`56712` cycles	`1.00`
`ML-KEM-768 keypair`	`62384` cycles	`62436` cycles	`1.00`
`ML-KEM-768 encaps`	`70928` cycles	`70943` cycles	`1.00`
`ML-KEM-768 decaps`	`86863` cycles	`86875` cycles	`1.00`
`ML-KEM-1024 keypair`	`96270` cycles	`96269` cycles	`1.00`
`ML-KEM-1024 encaps`	`106340` cycles	`106328` cycles	`1.00`
`ML-KEM-1024 decaps`	`126782` cycles	`126795` cycles	`1.00`

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

oqs-bot

Graviton2

Details

Benchmark suite	Current: `caa0cd1`	Previous: `4f13e12`	Ratio
`ML-KEM-512 keypair`	`28338` cycles	`28317` cycles	`1.00`
`ML-KEM-512 encaps`	`34250` cycles	`34300` cycles	`1.00`
`ML-KEM-512 decaps`	`44487` cycles	`44538` cycles	`1.00`
`ML-KEM-768 keypair`	`47887` cycles	`47847` cycles	`1.00`
`ML-KEM-768 encaps`	`54223` cycles	`54179` cycles	`1.00`
`ML-KEM-768 decaps`	`68735` cycles	`68665` cycles	`1.00`
`ML-KEM-1024 keypair`	`70548` cycles	`70577` cycles	`1.00`
`ML-KEM-1024 encaps`	`79147` cycles	`79180` cycles	`1.00`
`ML-KEM-1024 decaps`	`98840` cycles	`98900` 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: `caa0cd1`	Previous: `4f13e12`	Ratio
`ML-KEM-512 keypair`	`155138` cycles	`155127` cycles	`1.00`
`ML-KEM-512 encaps`	`163335` cycles	`163266` cycles	`1.00`
`ML-KEM-512 decaps`	`206594` cycles	`206482` cycles	`1.00`
`ML-KEM-768 keypair`	`249470` cycles	`249530` cycles	`1.00`
`ML-KEM-768 encaps`	`270250` cycles	`270293` cycles	`1.00`
`ML-KEM-768 decaps`	`332037` cycles	`332121` cycles	`1.00`
`ML-KEM-1024 keypair`	`395154` cycles	`395107` cycles	`1.00`
`ML-KEM-1024 encaps`	`422159` cycles	`423803` cycles	`1.00`
`ML-KEM-1024 decaps`	`506961` cycles	`505540` cycles	`1.00`

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

oqs-bot

Graviton2 (no-opt)

Details

Benchmark suite	Current: `caa0cd1`	Previous: `4f13e12`	Ratio
`ML-KEM-512 keypair`	`59163` cycles	`59178` cycles	`1.00`
`ML-KEM-512 encaps`	`68636` cycles	`68656` cycles	`1.00`
`ML-KEM-512 decaps`	`87423` cycles	`87403` cycles	`1.00`
`ML-KEM-768 keypair`	`95844` cycles	`95411` cycles	`1.00`
`ML-KEM-768 encaps`	`109434` cycles	`109348` cycles	`1.00`
`ML-KEM-768 decaps`	`134307` cycles	`134230` cycles	`1.00`
`ML-KEM-1024 keypair`	`148525` cycles	`147922` cycles	`1.00`
`ML-KEM-1024 encaps`	`164450` cycles	`163963` cycles	`1.00`
`ML-KEM-1024 decaps`	`195730` cycles	`195119` 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: `caa0cd1`	Previous: `4f13e12`	Ratio
`ML-KEM-512 keypair`	`28331` cycles	`28402` cycles	`1.00`
`ML-KEM-512 encaps`	`34303` cycles	`34330` cycles	`1.00`
`ML-KEM-512 decaps`	`44529` cycles	`44594` cycles	`1.00`
`ML-KEM-768 keypair`	`47860` cycles	`47799` cycles	`1.00`
`ML-KEM-768 encaps`	`54127` cycles	`54109` cycles	`1.00`
`ML-KEM-768 decaps`	`68679` cycles	`68614` cycles	`1.00`
`ML-KEM-1024 keypair`	`70562` cycles	`70522` cycles	`1.00`
`ML-KEM-1024 encaps`	`79122` cycles	`79092` cycles	`1.00`
`ML-KEM-1024 decaps`	`98818` cycles	`98772` 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: `caa0cd1`	Previous: `4f13e12`	Ratio
`ML-KEM-512 keypair`	`59482` cycles	`59513` cycles	`1.00`
`ML-KEM-512 encaps`	`67276` cycles	`67237` cycles	`1.00`
`ML-KEM-512 decaps`	`85785` cycles	`85783` cycles	`1.00`
`ML-KEM-768 keypair`	`96998` cycles	`96990` cycles	`1.00`
`ML-KEM-768 encaps`	`110458` cycles	`110613` cycles	`1.00`
`ML-KEM-768 decaps`	`137300` cycles	`137269` cycles	`1.00`
`ML-KEM-1024 keypair`	`154015` cycles	`154033` cycles	`1.00`
`ML-KEM-1024 encaps`	`170601` cycles	`170579` cycles	`1.00`
`ML-KEM-1024 decaps`	`206582` cycles	`206470` cycles	`1.00`

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

hanno-becker

I support this, but it is an API break, so we need a fall-back path, ideally with a warning being emitted when the now-deprecated defines are used.

mkannwischer · 2026-04-02T09:51:02Z

Closing this as I don't care enough about it.

mkannwischer added benchmark this PR should be benchmarked in CI needs-mldsa-native-port labels Feb 2, 2026

mkannwischer added benchmark this PR should be benchmarked in CI and removed benchmark this PR should be benchmarked in CI labels Feb 2, 2026

oqs-bot reviewed Feb 2, 2026

View reviewed changes

mkannwischer marked this pull request as ready for review February 2, 2026 09:03

mkannwischer requested a review from a team as a code owner February 2, 2026 09:03

hanno-becker requested changes Feb 26, 2026

View reviewed changes

hanno-becker added the API-BREAK label Feb 26, 2026

mkannwischer closed this Apr 2, 2026

mkannwischer removed the needs-mldsa-native-port label Apr 5, 2026

Conversation

mkannwischer commented Feb 2, 2026

Uh oh!

oqs-bot commented Feb 2, 2026

CBMC Results (ML-KEM-512)

Uh oh!

oqs-bot commented Feb 2, 2026

CBMC Results (ML-KEM-768)

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

Mac Mini (M1, 2020) benchmarks

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

ppc64le (POWER10) benchmarks

Uh oh!

oqs-bot commented Feb 2, 2026

CBMC Results (ML-KEM-1024)

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

Intel Xeon 4th gen (c7i)

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

AMD EPYC 3rd gen (c6a)

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

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

Uh oh!

oqs-bot left a comment

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

Intel Xeon 3rd gen (c6i)

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

Graviton4

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

Arm Cortex-A72 (Raspberry Pi 4) benchmarks

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

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

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

Graviton3

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

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

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

Graviton4 (no-opt)

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

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

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

Graviton3 (no-opt)

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

Graviton2

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

SpacemiT K1 8 (Banana Pi F3) benchmarks

Uh oh!