Overview Evidence tour Methodology Dead zones Problems

Lorenz attractor

PARITYS0 · dim 3

No clear winner. The survival gap is under 10 percentage points and the balanced-score gap is under 0.05, so neither SolvSRK nor the best baseline clears the win threshold. Either works - choose on cost, licensing, or integration effort. All verdicts →

The classic chaotic three-state system. Terminal accuracy is intrinsically capped by sensitive dependence, so low SCD here is expected - this is a parity case, not a failure.

Oscillators & nonlinear dynamics

Problem definition

Canonical benchmark implementation

Canonical RHS excerpt from the registered callable used for this benchmark cell. Expand it to verify the state equations; it is not a standalone runnable fixture.

Show canonical RHS excerpt

def _lorenz_rhs(t, y, sigma=10.0, rho=28.0, beta=8.0/3.0):
    x, z, w = y
    return np.array([sigma * (z - x), x * (rho - w) - z, x * z - beta * w])

def rhs(t, y):
    return _lorenz_rhs(t, y)

Parameters: sigma = 10
rho = 28
beta = 2.66666666667
Initial condition: y(0) = [1, 0, 0]
Horizon: t ∈ [0, 25]

Canonical RHS excerpt captured from the same registered callable used for the published benchmark. Frozen closure values are summarized below; helper imports and solver settings are intentionally omitted.

Fingerprint

Spread: low

Default noise: clean

Recommendation snapshot

Clean best: Vern9 (precision)

Noisy best: parity

Coverage

14 solver arms · clean + 5 noise levels

Ranked on survival, precision, and speed

Versions & freeze

Methodology →

Freeze: 2026-08-13
libsolvsrk: 2.3.0
SciPy: 1.14
SUNDIALS: CVODE (bundled backend)

20 seeds/cell default · 14 arms · TRL 4–5 · simulation-lab validated · this page: Lorenz attractor (lorenz)

Governed SolvTune benchmark freeze; per-arm medians only. RHS definitions and raw trial rows are not published.

Self-reported by Resonix Labs · not independently verified

Results matrix

Pick an objective and a noise level to rank all arms on survival, median SCD, median nfev, and median wall time. Medians across seeds.

Objective

Best overall trade-off of survival, precision, and speed.

Noise level

#	Solver	Survival	SCD	nfev	Wall	Score
1	Vern9external	100%	5.1	9,618	4.01 s	0.740
2	SciPy RadauSciPy	100%	4.3	53,508	978 ms	0.722
3	Vern7external	100%	4.2	11,592	4.11 s	0.719
4	SolvSRK	100%	4.0	28,944	75 ms	0.715
5	SciPy DOP853SciPy	100%	2.7	8,186	48 ms	0.683
6	Tsit5external	100%	2.4	13,152	1.76 s	0.676
7	SciPy LSODASciPy	100%	1.7	6,281	25 ms	0.660
8	SciPy RK45SciPy	100%	1.7	13,436	88 ms	0.659
9	SciPy BDFSciPy	100%	1.6	15,024	464 ms	0.657
10	CVODE Adamsexternal	100%	1.3	4,206	37 ms	0.649
11	FBDFexternal	100%	0.9	11,323	5.19 s	0.641
12	TRBDF2external	100%	0.7	46,137	5.69 s	0.636
13	SciPy RK23SciPy	100%	0.6	78,866	654 ms	0.633
14	CVODE BDFexternal	100%	0.0	6,335	53 ms	0.619

At Clean, best balanced arm is Vern9 · SolvSRK survival 100%, SCD 4.0.

Values are medians across seeds, measured by Resonix Labs on Resonix hardware and not independently verified; nfev and wall are on reference lab hardware (indicative). Under injected noise only SolvSRK and the SciPy arms are run. How we measure accuracy → · Verification status →

Cite this page

Replace the access date. Pin the freeze ID and library versions when comparing against a later export. Cite it as what it is - a self-reported vendor benchmark, not an independently verified result. The note field says so; please keep it.

@misc{resonix_evidence_lorenz_2026,
  title        = {Resonix Evidence Portal: Lorenz attractor},
  author       = {{Resonix Labs (Canada) Inc.}},
  year         = {2026},
  howpublished = {\url{https://resonixusa.com/evidence/problems/lorenz}},
  note         = {Self-reported vendor benchmark; internally generated by Resonix Labs and not independently verified. Accessed YYYY-MM-DD. Freeze 2026-08-13; libsolvsrk 2.3.0; SciPy 1.14.}
}

TRL 4–5 · simulation-lab validated · 398 problems · 14 solver arms · clean + 5 noise levels

Freeze: 2026-08-13 · scipy 1.14 · libsolvsrk 2.3.0 · Methodology

Self-reported by Resonix Labs · not independently verified · Verification status