Performance measurements › x86_1 › kex-04 · system: x86_1 · arm_1

kex-04 DKEX (Ding Key Exchange) — performance on x86-64 (system x86_1)

Independent measurement following the structure of the NICCS x86 self-assessment guide, §3.5 (1)–(7). Not a submitter self-assessment.

1. Basic information

Category: public-key algorithm; function: key exchange
Algorithm: DKEX (Ding Key Exchange)
Implementation versions measured: reference
Parameter sets: DKEX-128, DKEX-256, DKEX-512
Security evaluation: kex-04 report
Measurement method: x86_1 method
Submission: NICCS page

2. Assessment environment

item	value
processor	12th Gen Intel(R) Core(TM) i7-12700 (CPU 2, one core)
clock	max 2.10 GHz, governor performance, turbo off, SMT off
memory	31788 MiB
OS / kernel	Debian GNU/Linux 13 (trixie) / 6.12.107+deb13-amd64
compiler / build tool	gcc (Debian 14.2.0-19) 14.2.0 / cmake version 3.31.6
campaign start / end (UTC)	2026-09-25T10:02:21 / 2026-09-28T10:51:21

3. Functional testing (KAT)

Each library was checked against the submitted KAT vectors (SHA-256 manifest kex-04/kat.sha256 in the harness) before timing.

instance	build flags	KAT
`DKEX-128`	guide	PASS
`DKEX-256`	guide	PASS
`DKEX-512`	guide	PASS

4. Performance

instance	operation	mean cycles	mean time	ops/s	median trial	n (trials × iters)
`DKEX-128`	exchange	3.84 M	1.84 ms	544	1.84 ms	3000 (5 × 600)
`DKEX-128`	init_a	267.8 k	132 µs	7.58e+03	128 µs	3000 (5 × 600)
`DKEX-128`	init_b	267.3 k	128 µs	7.84e+03	127 µs	3000 (5 × 600)
`DKEX-128`	pass1	203.3 k	97 µs	1.03e+04	97 µs	3000 (5 × 600)
`DKEX-128`	pass2	1.39 M	662 µs	1.51e+03	662 µs	3000 (5 × 600)
`DKEX-128`	pass3	1.48 M	718 µs	1.39e+03	705 µs	3000 (5 × 600)
`DKEX-128`	derive_a	3243	1.49 µs	6.69e+05	1.44 µs	3000 (5 × 600)
`DKEX-128`	derive_b	287.6 k	137 µs	7.28e+03	137 µs	3000 (5 × 600)
`DKEX-256`	exchange	8.70 M	4.16 ms	240	4.16 ms	1030 (5 × 206)
`DKEX-256`	init_a	717.9 k	343 µs	2.92e+03	343 µs	1030 (5 × 206)
`DKEX-256`	init_b	724.7 k	346 µs	2.89e+03	346 µs	1030 (5 × 206)
`DKEX-256`	pass1	604.8 k	289 µs	3.46e+03	289 µs	1030 (5 × 206)
`DKEX-256`	pass2	2.90 M	1.39 ms	721	1.39 ms	1030 (5 × 206)
`DKEX-256`	pass3	3.02 M	1.44 ms	693	1.44 ms	1030 (5 × 206)
`DKEX-256`	derive_a	3242	1.44 µs	6.95e+05	1.44 µs	1030 (5 × 206)
`DKEX-256`	derive_b	752.3 k	359 µs	2.78e+03	359 µs	1030 (5 × 206)
`DKEX-512`	exchange	12.24 M	5.85 ms	171	5.85 ms	890 (5 × 178)
`DKEX-512`	init_a	730.9 k	349 µs	2.86e+03	349 µs	890 (5 × 178)
`DKEX-512`	init_b	734.3 k	351 µs	2.85e+03	350 µs	890 (5 × 178)
`DKEX-512`	pass1	2.06 M	1.01 ms	991	984 µs	890 (5 × 178)
`DKEX-512`	pass2	4.71 M	2.25 ms	444	2.25 ms	890 (5 × 178)
`DKEX-512`	pass3	3.28 M	1.57 ms	637	1.57 ms	890 (5 × 178)
`DKEX-512`	derive_a	9114	4.24 µs	2.36e+05	4.25 µs	890 (5 × 178)
`DKEX-512`	derive_b	768.0 k	367 µs	2.73e+03	367 µs	890 (5 × 178)

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
`DKEX-128`	exchange	57265	1896 KiB	1960 KiB
`DKEX-128`	init_a	57265	1884 KiB	1948 KiB
`DKEX-128`	init_b	57265	1888 KiB	1952 KiB
`DKEX-128`	pass1	57265	1900 KiB	1980 KiB
`DKEX-128`	pass2	57265	1876 KiB	1984 KiB
`DKEX-128`	pass3	57265	1896 KiB	1960 KiB
`DKEX-128`	derive_a	57265	1892 KiB	1956 KiB
`DKEX-128`	derive_b	57265	1876 KiB	1964 KiB
`DKEX-256`	exchange	58249	2068 KiB	2132 KiB
`DKEX-256`	init_a	58249	2056 KiB	2120 KiB
`DKEX-256`	init_b	58249	2072 KiB	2144 KiB
`DKEX-256`	pass1	58249	2064 KiB	2128 KiB
`DKEX-256`	pass2	58249	2048 KiB	2112 KiB
`DKEX-256`	pass3	58249	2044 KiB	2140 KiB
`DKEX-256`	derive_a	58249	2048 KiB	2112 KiB
`DKEX-256`	derive_b	58249	1980 KiB	2108 KiB
`DKEX-512`	exchange	59281	2124 KiB	2200 KiB
`DKEX-512`	init_a	59281	2108 KiB	2212 KiB
`DKEX-512`	init_b	59281	2132 KiB	2212 KiB
`DKEX-512`	pass1	59281	2104 KiB	2232 KiB
`DKEX-512`	pass2	59281	2104 KiB	2212 KiB
`DKEX-512`	pass3	59281	2124 KiB	2200 KiB
`DKEX-512`	derive_a	59281	2124 KiB	2212 KiB
`DKEX-512`	derive_b	59281	2124 KiB	2200 KiB

6. Transmission and storage overhead

Bandwidth counts all specified protocol messages and each required public key once. Public keys are transmitted bytes too. Certificates and transport framing are excluded. The published raw timing records are unchanged.

instance	passes	messages (bytes; raw API)	protocol-message bytes	public key A / B	bandwidth (bytes)	long-term sk (API cap)	shared secret
`DKEX-128`	3	800 / 3188 / 2420	6408	1312 / 1312	9032	2560	32
`DKEX-256`	3	1568 / 6195 / 4627	12390	2592 / 2592	17574	4896	32
`DKEX-512`	3	3392 / 7699 / 4627	15718	2592 / 2592	20902	4896	64

Symmetric primitives

Survey verdict: mixed — KEX core via ICCS; embedded ML-DSA uses its own SHAKE

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
`DKEX-128`	exchange	12%	0.5%	drng 4, pseudoXOF 1, sm3hash 186
`DKEX-256`	exchange	14%	0.2%	drng 4, pseudoXOF 1, sm3hash 585
`DKEX-512`	exchange	36%	0.2%	drng 4, pseudoXOF 5, sm3hash 1.23e+03

7. Raw evidence index

Paths are relative to performance/data/x86_1/ in the harness; each JSON file records its commands, environment and trials.

instance	item	file
`DKEX-128`	KAT log (sha256 `7159b0d28cca8ce6…`)	`kat/kex-04/DKEX-128.log`
`DKEX-128`	timing derive_a	`records/kex-04/DKEX-128__derive_a.json`
`DKEX-128`	timing derive_b	`records/kex-04/DKEX-128__derive_b.json`
`DKEX-128`	timing exchange	`records/kex-04/DKEX-128__exchange.json`
`DKEX-128`	timing init_a	`records/kex-04/DKEX-128__init_a.json`
`DKEX-128`	timing init_b	`records/kex-04/DKEX-128__init_b.json`
`DKEX-128`	timing pass1	`records/kex-04/DKEX-128__pass1.json`
`DKEX-128`	timing pass2	`records/kex-04/DKEX-128__pass2.json`
`DKEX-128`	timing pass3	`records/kex-04/DKEX-128__pass3.json`
`DKEX-128`	hash profile exchange	`profile/kex-04/DKEX-128__exchange.json`
`DKEX-256`	KAT log (sha256 `0edc6b7375877b75…`)	`kat/kex-04/DKEX-256.log`
`DKEX-256`	timing derive_a	`records/kex-04/DKEX-256__derive_a.json`
`DKEX-256`	timing derive_b	`records/kex-04/DKEX-256__derive_b.json`
`DKEX-256`	timing exchange	`records/kex-04/DKEX-256__exchange.json`
`DKEX-256`	timing init_a	`records/kex-04/DKEX-256__init_a.json`
`DKEX-256`	timing init_b	`records/kex-04/DKEX-256__init_b.json`
`DKEX-256`	timing pass1	`records/kex-04/DKEX-256__pass1.json`
`DKEX-256`	timing pass2	`records/kex-04/DKEX-256__pass2.json`
`DKEX-256`	timing pass3	`records/kex-04/DKEX-256__pass3.json`
`DKEX-256`	hash profile exchange	`profile/kex-04/DKEX-256__exchange.json`
`DKEX-512`	KAT log (sha256 `1fe8edad8396601a…`)	`kat/kex-04/DKEX-512.log`
`DKEX-512`	timing derive_a	`records/kex-04/DKEX-512__derive_a.json`
`DKEX-512`	timing derive_b	`records/kex-04/DKEX-512__derive_b.json`
`DKEX-512`	timing exchange	`records/kex-04/DKEX-512__exchange.json`
`DKEX-512`	timing init_a	`records/kex-04/DKEX-512__init_a.json`
`DKEX-512`	timing init_b	`records/kex-04/DKEX-512__init_b.json`
`DKEX-512`	timing pass1	`records/kex-04/DKEX-512__pass1.json`
`DKEX-512`	timing pass2	`records/kex-04/DKEX-512__pass2.json`
`DKEX-512`	timing pass3	`records/kex-04/DKEX-512__pass3.json`
`DKEX-512`	hash profile exchange	`profile/kex-04/DKEX-512__exchange.json`

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