Performance measurements › arm_1 › sign-12 · system: x86_1 · arm_1

sign-12 Galas Signature Scheme — performance on AArch64 (system arm_1)

Independent measurement following the structure of the NICCS ARM self-assessment guide, §3.5 (1)–(7). Not a submitter self-assessment.

1. Basic information

Category: public-key algorithm; function: digital signature
Algorithm: Galas Signature Scheme
Implementation versions measured: reference
Parameter sets: Galas-160F, Galas-160S, Galas-256F, Galas-256S, Galas-384F, Galas-384S, Galas-512F, Galas-512S
Security evaluation: sign-12 report
Measurement method: arm_1 method
Submission: NICCS page

2. Assessment environment

item	value
processor	Qualcomm Oryon (CPU 2, one core)
machine	ASUS Vivobook S 15
clock	fixed 2.71 GHz (governor performance, minimum = maximum), boost off, SMT none
memory	30562 MiB
OS / kernel	Ubuntu 26.04.1 LTS / 7.0.0-34-generic
compiler / build tool	gcc (Ubuntu 15.2.0-16ubuntu1) 15.2.0 / cmake version 4.2.3
campaign start / end (UTC)	2026-09-28T15:05:46 / 2026-09-29T08:43:52

3. Functional testing (KAT)

Each library was checked against the submitted KAT vectors (SHA-256 manifest sign-12/kat.sha256 in the harness) before timing.

instance	build flags	KAT
`Galas-160F`	harness-default	MISMATCH [1]
`Galas-160S`	harness-default	MISMATCH [1]
`Galas-256F`	harness-default	MISMATCH [1]
`Galas-256S`	harness-default	MISMATCH [1]
`Galas-384F`	harness-default	MISMATCH [1]
`Galas-384S`	harness-default	MISMATCH [1]
`Galas-512F`	harness-default	MISMATCH [1]
`Galas-512S`	harness-default	MISMATCH [1]

[1] API deviation: sig_keygen does not draw from drng_algorithm but seeds a private DRNG from 32 zero bytes (or from a seed set by a non-API helper used only by the submitters' own KAT generator), so the official KAT flow yields different keys; with keygen drawing from the DRNG the first KAT records reproduce exactly (sign-12/Makefile). Galas-512S additionally exceeds the 900 s KAT time limit. Timed anyway; note that key generation always produces the same key pair. These instances are timed anyway; their output is not validated.

4. Performance

instance	operation	mean cycles	mean time	ops/s	median trial	n (trials × iters)
`Galas-160F`	keygen	17.40 M	6.45 ms	155	6.46 ms	485 (5 × 97)
`Galas-160F`	sign	1.58 G	585 ms	1.71	589 ms	100 (5 × 20)
`Galas-160F`	verify	968.28 M	359 ms	2.78	360 ms	100 (5 × 20)
`Galas-160S`	keygen	17.39 M	6.45 ms	155	6.46 ms	490 (5 × 98)
`Galas-160S`	sign	2.04 G	758 ms	1.32	758 ms	100 (5 × 20)
`Galas-160S`	verify	1.40 G	519 ms	1.92	519 ms	100 (5 × 20)
`Galas-256F`	keygen	45.48 M	16.9 ms	59.3	16.9 ms	190 (5 × 38)
`Galas-256F`	sign	6.53 G	2.42 s	0.413	2.44 s	100 (5 × 20)
`Galas-256F`	verify	3.90 G	1.45 s	0.692	1.47 s	100 (5 × 20)
`Galas-256S`	keygen	45.24 M	16.8 ms	59.6	16.9 ms	190 (5 × 38)
`Galas-256S`	sign	10.48 G	3.89 s	0.257	3.89 s	80 (5 × 16)
`Galas-256S`	verify	7.81 G	2.9 s	0.345	2.9 s	100 (5 × 20)
`Galas-384F`	keygen	106.98 M	39.7 ms	25.2	39.7 ms	100 (5 × 20)
`Galas-384F`	sign	24.98 G	9.27 s	0.108	9.27 s	30 (5 × 6)
`Galas-384F`	verify	14.31 G	5.31 s	0.188	5.31 s	65 (5 × 13)
`Galas-384S`	keygen	107.01 M	39.7 ms	25.2	39.7 ms	100 (5 × 20)
`Galas-384S`	sign	33.17 G	12.3 s	0.0813	12.3 s	25 (5 × 5)
`Galas-384S`	verify	22.25 G	8.26 s	0.121	8.26 s	40 (5 × 8)
`Galas-512F`	keygen	195.39 M	72.5 ms	13.8	73 ms	100 (5 × 20)
`Galas-512F`	sign	66.84 G	24.8 s	0.0403	25 s	10 (5 × 2)
`Galas-512F`	verify	35.30 G	13.1 s	0.0763	13.1 s	30 (5 × 6)
`Galas-512S`	keygen	196.94 M	73.1 ms	13.7	73.1 ms	100 (5 × 20)
`Galas-512S`	sign	83.19 G	30.9 s	0.0324	30.8 s	10 (5 × 2)
`Galas-512S`	verify	50.62 G	18.8 s	0.0532	18.8 s	15 (5 × 3)

5. Resource consumption

Static memory is approximated by the library's loadable ELF segments; peak memory is the process high-water mark (VmHWM) including the benchmark driver and libc, so both are upper-bound proxies rather than isolated algorithm memory.

instance	operation	static: ELF image (bytes)	baseline RSS	peak RSS
`Galas-160F`	keygen	590452	3620 KiB	3684 KiB
`Galas-160F`	sign	590452	1592 KiB	7388 KiB
`Galas-160F`	verify	590452	6844 KiB	7192 KiB
`Galas-160S`	keygen	590452	1592 KiB	1656 KiB
`Galas-160S`	sign	590452	1592 KiB	9596 KiB
`Galas-160S`	verify	590452	7992 KiB	9512 KiB
`Galas-256F`	keygen	590452	1600 KiB	1728 KiB
`Galas-256F`	sign	590452	1664 KiB	13848 KiB
`Galas-256F`	verify	590452	13444 KiB	14012 KiB
`Galas-256S`	keygen	590452	1596 KiB	1724 KiB
`Galas-256S`	sign	590452	1660 KiB	26608 KiB
`Galas-256S`	verify	590452	12280 KiB	26264 KiB
`Galas-384F`	keygen	590452	1620 KiB	1876 KiB
`Galas-384F`	sign	590452	1812 KiB	38896 KiB
`Galas-384F`	verify	590452	35064 KiB	38068 KiB
`Galas-384S`	keygen	590452	1612 KiB	1868 KiB
`Galas-384S`	sign	590452	1804 KiB	64592 KiB
`Galas-384S`	verify	590452	36452 KiB	64100 KiB
`Galas-512F`	keygen	590452	1644 KiB	2028 KiB
`Galas-512F`	sign	590452	1964 KiB	84780 KiB
`Galas-512F`	verify	590452	78904 KiB	83928 KiB
`Galas-512S`	keygen	590452	1632 KiB	2020 KiB
`Galas-512S`	sign	590452	1956 KiB	134496 KiB
`Galas-512S`	verify	590452	80608 KiB	134100 KiB

6. Transmission and storage overhead

External public-key, ciphertext and signature sizes follow the curated performance/external_sizes.csv catalog; secret-key and shared-secret lengths remain API figures. See the size audit for disagreements.

instance	public key	secret key	signature
`Galas-160F`	40	40	5964
`Galas-160S`	40	40	4812
`Galas-256F`	64	64	15950
`Galas-256S`	64	64	12114
`Galas-384F`	96	96	33384
`Galas-384S`	96	96	27784
`Galas-512F`	128	128	59416
`Galas-512S`	128	128	49516

Symmetric primitives

Survey verdict: ICCS-only

Share of each operation spent in the ICCS placeholder hash functions and in the ICCS DRNG (reference build, measured in the same run with link-time wrappers):

instance	operation	pseudohash/XOF/sm3	DRNG	calls per operation
`Galas-160F`	keygen	0.0%	0.1%	drng 6
`Galas-160F`	sign	4.7%	0.0%	pseudoXOF 3.69e+04, pseudohash 102
`Galas-160F`	verify	7.5%	0.0%	pseudoXOF 3.65e+04, pseudohash 28
`Galas-160S`	keygen	0.0%	0.1%	drng 6
`Galas-160S`	sign	25%	0.0%	pseudoXOF 2.49e+05, pseudohash 1.85e+03
`Galas-160S`	verify	34%	0.0%	pseudoXOF 2.49e+05, pseudohash 22
`Galas-256F`	keygen	0.0%	0.1%	drng 6
`Galas-256F`	sign	3.6%	0.0%	pseudoXOF 6.18e+04, pseudohash 297
`Galas-256F`	verify	5.7%	0.0%	pseudoXOF 6.11e+04, pseudohash 44
`Galas-256S`	keygen	0.0%	0.1%	drng 6
`Galas-256S`	sign	38%	0.0%	pseudoXOF 1.12e+06, pseudohash 5.65e+03
`Galas-256S`	verify	50%	0.0%	pseudoXOF 1.12e+06, pseudohash 30
`Galas-384F`	keygen	0.0%	0.0%	drng 6
`Galas-384F`	sign	3.2%	0.0%	pseudoXOF 2.3e+05, pseudohash 266
`Galas-384F`	verify	5.8%	0.0%	pseudoXOF 2.29e+05, pseudohash 57
`Galas-384S`	keygen	0.0%	0.0%	drng 6
`Galas-384S`	sign	25%	0.0%	pseudoXOF 2.38e+06, pseudohash 3.04e+03
`Galas-384S`	verify	39%	0.0%	pseudoXOF 2.37e+06, pseudohash 41
`Galas-512F`	keygen	0.0%	0.0%	drng 6
`Galas-512F`	sign	2.3%	0.0%	pseudoXOF 4.22e+05, pseudohash 178
`Galas-512F`	verify	4.4%	0.0%	pseudoXOF 4.19e+05, pseudohash 73
`Galas-512S`	keygen	0.0%	0.0%	drng 6
`Galas-512S`	sign	21%	0.0%	pseudoXOF 4.64e+06, pseudohash 3.29e+03
`Galas-512S`	verify	34%	0.0%	pseudoXOF 4.64e+06, pseudohash 51

7. Raw evidence index

Paths are relative to performance/data/arm_1/ in the harness; each JSON file records its commands, environment and trials.

instance	item	file
`Galas-160F`	KAT log (sha256 `ce820a23f0eb5a0a…`)	`kat/sign-12/Galas-160F.log`
`Galas-160F`	timing keygen	`records/sign-12/Galas-160F__keygen.json`
`Galas-160F`	timing sign	`records/sign-12/Galas-160F__sign.json`
`Galas-160F`	timing verify	`records/sign-12/Galas-160F__verify.json`
`Galas-160F`	hash profile keygen	`profile/sign-12/Galas-160F__keygen.json`
`Galas-160F`	hash profile sign	`profile/sign-12/Galas-160F__sign.json`
`Galas-160F`	hash profile verify	`profile/sign-12/Galas-160F__verify.json`
`Galas-160S`	KAT log (sha256 `fe7595055ff70d19…`)	`kat/sign-12/Galas-160S.log`
`Galas-160S`	timing keygen	`records/sign-12/Galas-160S__keygen.json`
`Galas-160S`	timing sign	`records/sign-12/Galas-160S__sign.json`
`Galas-160S`	timing verify	`records/sign-12/Galas-160S__verify.json`
`Galas-160S`	hash profile keygen	`profile/sign-12/Galas-160S__keygen.json`
`Galas-160S`	hash profile sign	`profile/sign-12/Galas-160S__sign.json`
`Galas-160S`	hash profile verify	`profile/sign-12/Galas-160S__verify.json`
`Galas-256F`	KAT log (sha256 `047d8cb9dbadb4c6…`)	`kat/sign-12/Galas-256F.log`
`Galas-256F`	timing keygen	`records/sign-12/Galas-256F__keygen.json`
`Galas-256F`	timing sign	`records/sign-12/Galas-256F__sign.json`
`Galas-256F`	timing verify	`records/sign-12/Galas-256F__verify.json`
`Galas-256F`	hash profile keygen	`profile/sign-12/Galas-256F__keygen.json`
`Galas-256F`	hash profile sign	`profile/sign-12/Galas-256F__sign.json`
`Galas-256F`	hash profile verify	`profile/sign-12/Galas-256F__verify.json`
`Galas-256S`	KAT log (sha256 `5b3011221b86b9b0…`)	`kat/sign-12/Galas-256S.log`
`Galas-256S`	timing keygen	`records/sign-12/Galas-256S__keygen.json`
`Galas-256S`	timing sign	`records/sign-12/Galas-256S__sign.json`
`Galas-256S`	timing verify	`records/sign-12/Galas-256S__verify.json`
`Galas-256S`	hash profile keygen	`profile/sign-12/Galas-256S__keygen.json`
`Galas-256S`	hash profile sign	`profile/sign-12/Galas-256S__sign.json`
`Galas-256S`	hash profile verify	`profile/sign-12/Galas-256S__verify.json`
`Galas-384F`	KAT log (sha256 `119f01db3d3f3732…`)	`kat/sign-12/Galas-384F.log`
`Galas-384F`	timing keygen	`records/sign-12/Galas-384F__keygen.json`
`Galas-384F`	timing sign	`records/sign-12/Galas-384F__sign.json`
`Galas-384F`	timing verify	`records/sign-12/Galas-384F__verify.json`
`Galas-384F`	hash profile keygen	`profile/sign-12/Galas-384F__keygen.json`
`Galas-384F`	hash profile sign	`profile/sign-12/Galas-384F__sign.json`
`Galas-384F`	hash profile verify	`profile/sign-12/Galas-384F__verify.json`
`Galas-384S`	KAT log (sha256 `2d1a95e4e5f779f3…`)	`kat/sign-12/Galas-384S.log`
`Galas-384S`	timing keygen	`records/sign-12/Galas-384S__keygen.json`
`Galas-384S`	timing sign	`records/sign-12/Galas-384S__sign.json`
`Galas-384S`	timing verify	`records/sign-12/Galas-384S__verify.json`
`Galas-384S`	hash profile keygen	`profile/sign-12/Galas-384S__keygen.json`
`Galas-384S`	hash profile sign	`profile/sign-12/Galas-384S__sign.json`
`Galas-384S`	hash profile verify	`profile/sign-12/Galas-384S__verify.json`
`Galas-512F`	KAT log (sha256 `046f335400853e85…`)	`kat/sign-12/Galas-512F.log`
`Galas-512F`	timing keygen	`records/sign-12/Galas-512F__keygen.json`
`Galas-512F`	timing sign	`records/sign-12/Galas-512F__sign.json`
`Galas-512F`	timing verify	`records/sign-12/Galas-512F__verify.json`
`Galas-512F`	hash profile keygen	`profile/sign-12/Galas-512F__keygen.json`
`Galas-512F`	hash profile sign	`profile/sign-12/Galas-512F__sign.json`
`Galas-512F`	hash profile verify	`profile/sign-12/Galas-512F__verify.json`
`Galas-512S`	KAT log (sha256 `6fee394bdc94afbe…`)	`kat/sign-12/Galas-512S.log`
`Galas-512S`	timing keygen	`records/sign-12/Galas-512S__keygen.json`
`Galas-512S`	timing sign	`records/sign-12/Galas-512S__sign.json`
`Galas-512S`	timing verify	`records/sign-12/Galas-512S__verify.json`
`Galas-512S`	hash profile keygen	`profile/sign-12/Galas-512S__keygen.json`
`Galas-512S`	hash profile sign	`profile/sign-12/Galas-512S__sign.json`
`Galas-512S`	hash profile verify	`profile/sign-12/Galas-512S__verify.json`

Scripts: performance/campaign.py, performance/ngcc_perf.c, performance/hashprof/ in the harness.