Performance measurements › arm_1 › sign-26 · system: x86_1 · arm_1

sign-26 SQIsign2D-push1/2 — 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: SQIsign2D-push1/2
Implementation versions measured: reference
Parameter sets: SQIsign2D-lvl1, SQIsign2D-lvl2, SQIsign2D-lvl3, SQIsign2D-lvl4
Security evaluation: sign-26 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-26/kat.sha256 in the harness) before timing.

instance	build flags	KAT
`SQIsign2D-lvl1`	harness-default	PASS
`SQIsign2D-lvl2`	harness-default	PASS
`SQIsign2D-lvl3`	harness-default	MISMATCH [1]
`SQIsign2D-lvl4`	harness-default	PASS

[1] The submitted Test_Vectors/KAT_SIG_SQIsign2D-lvl3.txt is a splice of level-2 and level-3 records (10 of 12 records have the level-2 secret-key length of 676 bytes instead of 900), so no level-3 build can reproduce it; the public keys agree (sign-26/pseudocode.md, discrepancy 2). 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)
`SQIsign2D-lvl1`	keygen	524.97 M	195 ms	5.13	195 ms	100 (5 × 20)
`SQIsign2D-lvl1`	sign	559.85 M	208 ms	4.81	208 ms	100 (5 × 20)
`SQIsign2D-lvl1`	verify	131.84 M	49 ms	20.4	49 ms	100 (5 × 20)
`SQIsign2D-lvl2`	keygen	1.69 G	627 ms	1.59	627 ms	100 (5 × 20)
`SQIsign2D-lvl2`	sign	2.89 G	1.07 s	0.933	1.07 s	100 (5 × 20)
`SQIsign2D-lvl2`	verify	400.18 M	149 ms	6.72	149 ms	100 (5 × 20)
`SQIsign2D-lvl3`	keygen	3.96 G	1.47 s	0.681	1.47 s	100 (5 × 20)
`SQIsign2D-lvl3`	sign	3.87 G	1.43 s	0.697	1.43 s	100 (5 × 20)
`SQIsign2D-lvl3`	verify	941.11 M	350 ms	2.86	350 ms	100 (5 × 20)
`SQIsign2D-lvl4`	keygen	45.65 G	16.9 s	0.0591	17 s	20 (5 × 4)
`SQIsign2D-lvl4`	sign	40.22 G	14.9 s	0.067	14.9 s	20 (5 × 4)
`SQIsign2D-lvl4`	verify	10.17 G	3.78 s	0.265	3.8 s	95 (5 × 19)

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
`SQIsign2D-lvl1`	keygen	64445160	1640 KiB	2560 KiB
`SQIsign2D-lvl1`	sign	64445160	2428 KiB	9948 KiB
`SQIsign2D-lvl1`	verify	64445160	4628 KiB	9740 KiB
`SQIsign2D-lvl2`	keygen	96521832	1664 KiB	2764 KiB
`SQIsign2D-lvl2`	sign	96521832	2604 KiB	18008 KiB
`SQIsign2D-lvl2`	verify	96521832	3600 KiB	16956 KiB
`SQIsign2D-lvl3`	keygen	128660584	1664 KiB	2996 KiB
`SQIsign2D-lvl3`	sign	128660584	2756 KiB	28808 KiB
`SQIsign2D-lvl3`	verify	128660584	5192 KiB	29804 KiB
`SQIsign2D-lvl4`	keygen	273510704	1668 KiB	3888 KiB
`SQIsign2D-lvl4`	sign	273510704	5756 KiB	26504 KiB
`SQIsign2D-lvl4`	verify	273510704	11380 KiB	111620 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
`SQIsign2D-lvl1`	64	456	150
`SQIsign2D-lvl2`	96	676	218
`SQIsign2D-lvl3`	128	900	293
`SQIsign2D-lvl4`	262	1838	593

Symmetric primitives

Survey verdict: ICCS-only — iccs/xof_iccs.c glue; NIST SHAKE/AES leftovers unused

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
`SQIsign2D-lvl1`	keygen	0.0%	0.6%	drng 727
`SQIsign2D-lvl1`	sign	0.0%	1.6%	drng 2.24e+03, pseudoXOF 1
`SQIsign2D-lvl1`	verify	0.0%	0.0%	pseudoXOF 1
`SQIsign2D-lvl2`	keygen	0.0%	0.3%	drng 1.27e+03
`SQIsign2D-lvl2`	sign	0.0%	0.5%	drng 6.11e+03, pseudoXOF 1.67
`SQIsign2D-lvl2`	verify	0.0%	0.0%	pseudoXOF 1
`SQIsign2D-lvl3`	keygen	0.0%	0.2%	drng 1.54e+03
`SQIsign2D-lvl3`	sign	0.0%	1.4%	drng 1.17e+04, pseudoXOF 1
`SQIsign2D-lvl3`	verify	0.0%	0.0%	pseudoXOF 1
`SQIsign2D-lvl4`	keygen	0.0%	0.1%	drng 4.22e+03
`SQIsign2D-lvl4`	sign	0.0%	0.1%	drng 9.46e+03, pseudoXOF 1
`SQIsign2D-lvl4`	verify	0.0%	0.0%	pseudoXOF 1

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
`SQIsign2D-lvl1`	KAT log (sha256 `1789a9c338e6ff90…`)	`kat/sign-26/SQIsign2D-lvl1.log`
`SQIsign2D-lvl1`	timing keygen	`records/sign-26/SQIsign2D-lvl1__keygen.json`
`SQIsign2D-lvl1`	timing sign	`records/sign-26/SQIsign2D-lvl1__sign.json`
`SQIsign2D-lvl1`	timing verify	`records/sign-26/SQIsign2D-lvl1__verify.json`
`SQIsign2D-lvl1`	hash profile keygen	`profile/sign-26/SQIsign2D-lvl1__keygen.json`
`SQIsign2D-lvl1`	hash profile sign	`profile/sign-26/SQIsign2D-lvl1__sign.json`
`SQIsign2D-lvl1`	hash profile verify	`profile/sign-26/SQIsign2D-lvl1__verify.json`
`SQIsign2D-lvl2`	KAT log (sha256 `42c9cd5053278b0d…`)	`kat/sign-26/SQIsign2D-lvl2.log`
`SQIsign2D-lvl2`	timing keygen	`records/sign-26/SQIsign2D-lvl2__keygen.json`
`SQIsign2D-lvl2`	timing sign	`records/sign-26/SQIsign2D-lvl2__sign.json`
`SQIsign2D-lvl2`	timing verify	`records/sign-26/SQIsign2D-lvl2__verify.json`
`SQIsign2D-lvl2`	hash profile keygen	`profile/sign-26/SQIsign2D-lvl2__keygen.json`
`SQIsign2D-lvl2`	hash profile sign	`profile/sign-26/SQIsign2D-lvl2__sign.json`
`SQIsign2D-lvl2`	hash profile verify	`profile/sign-26/SQIsign2D-lvl2__verify.json`
`SQIsign2D-lvl3`	KAT log (sha256 `69ed27e663bac327…`)	`kat/sign-26/SQIsign2D-lvl3.log`
`SQIsign2D-lvl3`	timing keygen	`records/sign-26/SQIsign2D-lvl3__keygen.json`
`SQIsign2D-lvl3`	timing sign	`records/sign-26/SQIsign2D-lvl3__sign.json`
`SQIsign2D-lvl3`	timing verify	`records/sign-26/SQIsign2D-lvl3__verify.json`
`SQIsign2D-lvl3`	hash profile keygen	`profile/sign-26/SQIsign2D-lvl3__keygen.json`
`SQIsign2D-lvl3`	hash profile sign	`profile/sign-26/SQIsign2D-lvl3__sign.json`
`SQIsign2D-lvl3`	hash profile verify	`profile/sign-26/SQIsign2D-lvl3__verify.json`
`SQIsign2D-lvl4`	KAT log (sha256 `25c5893c06d3664f…`)	`kat/sign-26/SQIsign2D-lvl4.log`
`SQIsign2D-lvl4`	timing keygen	`records/sign-26/SQIsign2D-lvl4__keygen.json`
`SQIsign2D-lvl4`	timing sign	`records/sign-26/SQIsign2D-lvl4__sign.json`
`SQIsign2D-lvl4`	timing verify	`records/sign-26/SQIsign2D-lvl4__verify.json`
`SQIsign2D-lvl4`	hash profile keygen	`profile/sign-26/SQIsign2D-lvl4__keygen.json`
`SQIsign2D-lvl4`	hash profile sign	`profile/sign-26/SQIsign2D-lvl4__sign.json`
`SQIsign2D-lvl4`	hash profile verify	`profile/sign-26/SQIsign2D-lvl4__verify.json`

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