Performance measurements › arm_1 › kem-36 · system: x86_1 · arm_1

kem-36 TRIKE — 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: key encapsulation
Algorithm: TRIKE
Implementation versions measured: reference
Parameter sets: TRIKE-2, TRIKE-5, TRIKE-7, TRIKE-9
Security evaluation: kem-36 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 kem-36/kat.sha256 in the harness) before timing.

instance	build flags	KAT
`TRIKE-2`	guide	PASS
`TRIKE-5`	guide	PASS
`TRIKE-7`	guide	PASS
`TRIKE-9`	guide	PASS

4. Performance

instance	operation	mean cycles	mean time	ops/s	median trial	n (trials × iters)
`TRIKE-2`	keygen	63.90 M	23.7 ms	42.2	23.5 ms	135 (5 × 27)
`TRIKE-2`	enc	6.65 M	2.47 ms	405	2.46 ms	1285 (5 × 257)
`TRIKE-2`	dec	35.18 M	13.1 ms	76.6	13.1 ms	245 (5 × 49)
`TRIKE-5`	keygen	474.51 M	176 ms	5.68	176 ms	100 (5 × 20)
`TRIKE-5`	enc	45.93 M	17 ms	58.7	17 ms	190 (5 × 38)
`TRIKE-5`	dec	133.26 M	49.4 ms	20.2	49.4 ms	100 (5 × 20)
`TRIKE-7`	keygen	1.54 G	572 ms	1.75	572 ms	100 (5 × 20)
`TRIKE-7`	enc	133.31 M	49.5 ms	20.2	49.5 ms	100 (5 × 20)
`TRIKE-7`	dec	405.06 M	150 ms	6.65	150 ms	100 (5 × 20)
`TRIKE-9`	keygen	1.94 G	720 ms	1.39	720 ms	100 (5 × 20)
`TRIKE-9`	enc	136.51 M	50.6 ms	19.7	50.6 ms	100 (5 × 20)
`TRIKE-9`	dec	796.85 M	296 ms	3.38	296 ms	100 (5 × 20)

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
`TRIKE-2`	keygen	27408	1424 KiB	1548 KiB
`TRIKE-2`	enc	27408	1492 KiB	1560 KiB
`TRIKE-2`	dec	27408	1660 KiB	3840 KiB
`TRIKE-5`	keygen	27840	1444 KiB	3740 KiB
`TRIKE-5`	enc	27840	3556 KiB	3620 KiB
`TRIKE-5`	dec	27840	3656 KiB	3996 KiB
`TRIKE-7`	keygen	27736	3468 KiB	3848 KiB
`TRIKE-7`	enc	27736	3648 KiB	3712 KiB
`TRIKE-7`	dec	27736	3600 KiB	4028 KiB
`TRIKE-9`	keygen	27576	1504 KiB	3772 KiB
`TRIKE-9`	enc	27576	3848 KiB	3912 KiB
`TRIKE-9`	dec	27576	3900 KiB	5592 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	ciphertext	shared secret
`TRIKE-2`	1980	6328	3928	32
`TRIKE-5`	4453	13988	8874	32
`TRIKE-7`	8776	27260	17488	64
`TRIKE-9`	14320	44228	28576	64

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
`TRIKE-2`	keygen	0.0%	1.2%	drng 113
`TRIKE-2`	enc	6.3%	21%	drng 267, pseudohash 2
`TRIKE-2`	dec	1.2%	3.9%	drng 266, pseudohash 2
`TRIKE-5`	keygen	0.0%	0.3%	drng 171
`TRIKE-5`	enc	2.0%	5.3%	drng 433, pseudohash 2
`TRIKE-5`	dec	0.7%	1.8%	drng 432, pseudohash 2
`TRIKE-7`	keygen	0.0%	0.1%	drng 255
`TRIKE-7`	enc	1.3%	3.0%	drng 663, pseudohash 2
`TRIKE-7`	dec	0.4%	1.0%	drng 662, pseudohash 2
`TRIKE-9`	keygen	0.0%	0.2%	drng 339
`TRIKE-9`	enc	2.1%	4.1%	drng 881, pseudohash 2
`TRIKE-9`	dec	0.4%	0.7%	drng 880, pseudohash 2

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
`TRIKE-2`	KAT log (sha256 `45ce672140b72918…`)	`kat/kem-36/TRIKE-2.log`
`TRIKE-2`	timing dec	`records/kem-36/TRIKE-2__dec.json`
`TRIKE-2`	timing enc	`records/kem-36/TRIKE-2__enc.json`
`TRIKE-2`	timing keygen	`records/kem-36/TRIKE-2__keygen.json`
`TRIKE-2`	hash profile dec	`profile/kem-36/TRIKE-2__dec.json`
`TRIKE-2`	hash profile enc	`profile/kem-36/TRIKE-2__enc.json`
`TRIKE-2`	hash profile keygen	`profile/kem-36/TRIKE-2__keygen.json`
`TRIKE-5`	KAT log (sha256 `f03ad615d441c702…`)	`kat/kem-36/TRIKE-5.log`
`TRIKE-5`	timing dec	`records/kem-36/TRIKE-5__dec.json`
`TRIKE-5`	timing enc	`records/kem-36/TRIKE-5__enc.json`
`TRIKE-5`	timing keygen	`records/kem-36/TRIKE-5__keygen.json`
`TRIKE-5`	hash profile dec	`profile/kem-36/TRIKE-5__dec.json`
`TRIKE-5`	hash profile enc	`profile/kem-36/TRIKE-5__enc.json`
`TRIKE-5`	hash profile keygen	`profile/kem-36/TRIKE-5__keygen.json`
`TRIKE-7`	KAT log (sha256 `62cc4313b653022e…`)	`kat/kem-36/TRIKE-7.log`
`TRIKE-7`	timing dec	`records/kem-36/TRIKE-7__dec.json`
`TRIKE-7`	timing enc	`records/kem-36/TRIKE-7__enc.json`
`TRIKE-7`	timing keygen	`records/kem-36/TRIKE-7__keygen.json`
`TRIKE-7`	hash profile dec	`profile/kem-36/TRIKE-7__dec.json`
`TRIKE-7`	hash profile enc	`profile/kem-36/TRIKE-7__enc.json`
`TRIKE-7`	hash profile keygen	`profile/kem-36/TRIKE-7__keygen.json`
`TRIKE-9`	KAT log (sha256 `ca06ad9161d1d573…`)	`kat/kem-36/TRIKE-9.log`
`TRIKE-9`	timing dec	`records/kem-36/TRIKE-9__dec.json`
`TRIKE-9`	timing enc	`records/kem-36/TRIKE-9__enc.json`
`TRIKE-9`	timing keygen	`records/kem-36/TRIKE-9__keygen.json`
`TRIKE-9`	hash profile dec	`profile/kem-36/TRIKE-9__dec.json`
`TRIKE-9`	hash profile enc	`profile/kem-36/TRIKE-9__enc.json`
`TRIKE-9`	hash profile keygen	`profile/kem-36/TRIKE-9__keygen.json`

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