How fast is PRISM at fleet scale?

On real US-equity data (Mar–Apr 2026), PRISM prices roughly 144 personalized accounts per second per core versus about 14 for a commercial CPU baseline — about 10.2× — and parallelizes linearly. It clears 100,000 personalized accounts in about 11.6 minutes on a single core versus about 118 minutes for the baseline.

What is the largest universe PRISM has solved?

PRISM runs a full 100,004-asset real backtest spanning 47+ exchanges at about 1.8 seconds per rebalance, with a walk-forward Sharpe of 0.737 and a maximum drawdown of −4.22%. An exact commercial solver times out near 5,000 names (~33 s); PRISM prices the full book deterministically with no timeout.

How good is PRISM's solution quality?

PRISM matches the exact reference optimum to about 1e-7. On after-tax outcome it is competitive with an exact commercial baseline — matching it in 58–63% of regimes and winning outright in about 33%. PRISM does not claim to beat the exact single-account optimum every time; its advantage is delivering near-exact results at a fleet scale the exact baseline cannot reach.

How much tax alpha does PRISM capture?

PRISM captures the full ~$238k harvestable tax budget on a $5M, 192-name book and is the only engine measured still producing six-figure tax alpha at 5,000+ names. It beats open-source solver baselines on after-tax wealth in about 88% of tested regimes and rules-based tax-loss harvesting in about 95%.

Benchmarks & Evidence | PRISM by Asymmetry Computing

00 · Verify it yourself

The one benchmark you can run without us.

Most numbers here are on real US-equity data we assembled. This one is on public Kenneth French (FF30) data anyone can download — so you can reproduce it on your own machine, with no licensed data and nothing to take on faith.

Public-data benchmark

23.28×

Faster than a commercial baseline at matched quality, on the public Fama–French 30-industry (FF30) dataset. Reproducible — no licensed data required.

Nothing to take on faith

public

The inputs are the standard Fama–French dataset the whole field already knows — nothing bespoke, nothing cherry-picked.

Then on your data

8-wk

An 8-week paid pilot runs the same comparison on your real book, constraints, and deadline — every losing case shown.

Latest · extreme scale · July 2026

The newest runs, at the frontier of scale.

The most recent recorded benchmarks push past prior ceilings — a half-million-account book and a million-device fleet, each returned inside its operational window, at matched quality. One engine, finance to grid.

Direct indexing · 500,000 accounts

~2 min

A 500,000-account book on a 1,000-name universe priced in ~134 s cold / ~63 s warm on a single GPU — roughly 25–30× a tuned 10-core commercial/open-source deployment (extrapolated from a measured 2,000-account subsample), every configuration gate-passing.

Real-time dispatch · 1,000,000 devices

~15 s

A million-device fleet allocated and disaggregated to feasible per-device setpoints in ~15 s — 25× end-to-end vs a tuned commercial solver, within ~0.2% of optimal, well inside the 5-minute market window.

Same engine, both

1 engine

The same optimization core spans finance and grid — deterministic, feasible by construction, and auditable end-to-end at every scale on this page.

Extreme-scale runs are chunked to fit a 20 GB card; PRISM times are fully measured wall-clock, baselines extrapolated where noted — so we say "roughly," not a spuriously precise figure. The batched whole-book/whole-fleet solve is the moat, not any single isolated problem.

01 · Scale incumbents can't reach

The workflow exact solvers physically can't run.

Nobody runs an exact solver across a 100,000-account book before the open — it times out long before. PRISM does, deterministically, and stays fast at every universe size.

Full-market backtest

100,004

Real assets across 47+ exchanges, solved at ~1.8 seconds per rebalance. Walk-forward Sharpe 0.737, max drawdown −4.22%.

Exact baseline ceiling

~5,000

Where an exact commercial solver times out (~33 s). PRISM prices the full book past that ceiling with no timeout.

Fastest measured

190–40k

PRISM is the fastest engine measured at every universe size across this range — not a single cherry-picked point.

02 · Fleet throughput

~10× more accounts per core — the margin lever.

Personalized accounts grow faster than AUM. Throughput per core decides how many accounts an ops team can run, and whether the nightly batch clears on time.

PRISMaccounts / sec / core

144

Commercial CPU Baseline 1accounts / sec / core

14

100,000 accounts · PRISM

11.6 min

Priced on a single core — and it parallelizes linearly across cores.

100,000 accounts · baseline

118 min

The same book on the commercial CPU baseline, single core.

Determinism

bit-exact

Repeated runs return content-hashed, identical outputs — no solver timeouts, no flaky reruns.

03 · Tax alpha

The only engine still harvesting six figures at scale.

As beta commoditizes, after-tax outcome is how direct-indexing and SMA desks compete. This is where PRISM is measured strongest — and where simpler approaches quietly leave money on the table.

Approach	Tax budget captured · $5M, 192 names	Six-figure alpha at 5,000+ names?	After-tax win rate vs PRISM
PRISM	~$238k (full budget)	Yes	—
Open-source solver baseline	~$3k at scale	No	PRISM wins ~88%
Rules-based tax-loss harvesting	partial	No	PRISM wins ~95%

Tax figures are measured on real-calibrated books; the robust real metric (tax captured) is measured directly. Your numbers are computed live on your data during the pilot.

04 · Quality, honestly

Near-exact — and we name where we don't win.

A sophisticated buyer trusts the vendor who states their non-fit first. So here it is, both sides.

Where PRISM wins

Matches the exact reference optimum to ~1e-7.
Competitive with an exact commercial baseline on after-tax outcome: matches it in 58–63% of regimes, wins outright in ~33%.
Runs the fleet-scale and full-universe workflows where exact solvers can't run at all.
Deterministic, reproducible, auditable — every trade traces to your constraints and tax logic.

Where PRISM does not

It does not beat the exact single-account optimum every time — no heuristic does.
It is not a bull-market beta play; diversification and tax discipline have a cost.
The quantum lane is a research track that adds zero tax alpha today — the production results are classical and measured.

05 · The measured curves

Every headline number, charted.

The same evidence behind the claims above, plotted straight from the recorded runs — factor-constrained QP on real US-equity universes, March–April 2026, p50 of five timed runs. Lower is better on solve time; speedup is versus the commercial baseline on CPU.

Figure 1 · Solve time vs universe size — factor-constrained QP

PRISM (GPU) Commercial baseline (CPU) Open-source solver 1 (CPU)

PRISM grows sub-linearly while the CPU baselines climb faster: at N=9,000 PRISM finishes in 303.6 ms versus 541.4 ms (the commercial baseline) and 994.8 ms (Open-source solver 1), all at sub-ppm quality gaps.

Figure 2 · Speedup by transition workflow — PRISM (GPU) vs Commercial baseline (CPU), N=5,000

Transition Rebalancegap < 0.004%

124.85×

Market Impactgap < 0.01%

25.86×

Tax-Aware Transitiongap < 0.01%

21.32×

Crisis Resiliencegap < 0.001%

20.92×

Every transition workflow clears a 20× floor against an exact commercial baseline, with every quality gap held under 0.01%. The rebalance lane — the most common intraday operation — reaches 124.85×.

Solver	Runtime p50 (ms)	Gap vs reference	Stability (CV)
PRISM (GPU)	138.7	< 0.00001%	11.7%
Commercial baseline (CPU)	208.1	reference	19.2%
Open-source solver 1 (CPU)	466.9	< 0.00001%	10.0%
Open-source solver 2 (CPU)	768.6	< 0.00001%	12.0%
GPU baseline	870.0	0.0014%	8.0%
Open-source solver 3 (CPU)	11,311	< 0.00001%	10.0%

Figure 3 · Structured QP at 5,000 real assets. PRISM is the fastest exact-quality solver in the field — 1.5× under the commercial baseline and 3.4× under Open-source solver 1 — while staying within a tight stability band. We keep the honest qualifier in: GPU baseline posts a lower run-to-run CV, and the commercial baseline remains the correctness reference every other engine is measured against.

Figure 4 · Fleet throughput — personalized accounts priced per second, per core

PRISM (GPU)~11.6 min for 100,000 accounts

144 acct/s

Commercial CPU baseline~118 min for 100,000 accounts

14 acct/s

~10.2× per-core throughput, and PRISM parallelizes linearly across cores. The whole personalized book is priced inside the overnight window, before the open — the workflow the fleet is actually billed on.

Figure 5 · The advantage holds at the ceiling — speedup vs the commercial baseline at N=75,257 real assets

Market Impact

12.28×

Tax-Aware Transition

5.60×

Rebalance

4.42×

Crisis Resilience

2.27×

Speedup peaks at 124.85× at N=5,000 and settles into a durable multiple as the universe grows to the full real-data ceiling. The institutional question at scale is persistence, not a monotonic climb: every transition workflow stays multiples faster — in a regime where exact solvers no longer finish at all.

Solver	QP Speed	Replay	Transition	Quality	Stability	Max Scale
PRISM (GPU)	10	10	10	10	9	10
Commercial baseline (CPU)	7	6	3	10	7	8
Open-source solver 1 (CPU)	6	5	0	9	7	5
GPU baseline	5	2	1	2	6	5
Open-source solver 3 (CPU)	2	2	1	3	5	4
Open-source solver 2 (CPU)	4	1	0	9	3	4

Figure 6 · Competitive overview — six solvers across six dimensions, scored 0–10 on real-data benchmarks (March 2026); higher is better. PRISM leads on speed, transition workflows, and scale; the commercial baseline and Open-source solver 2 match it on single-problem quality; GPU baseline edges run-to-run stability. No solver wins every axis — which is the point of showing all six.

06 · Trust primitives

Built to pass compliance, not just benchmarks.

In this buyer set, trust is the product and compliance is the gate. PRISM ships the de-riskers up front.

Deterministic & reproducible

Content-hashed, bit-reproducible outputs — re-derivable for any audit or exam date.

Tax-rule correctness

Wash-sale handling and lot-level tax accounting, validated by an internal test suite — 18 tests pass.

Validate against ground truth

An exact-reference comparator ships so you can check PRISM against the exact optimum on your own data.

Transparent, not a black box

Every trade traces to explicit constraints and tax logic — no opaque ML in the trade path.

07 · How we benchmark

Matched workloads. Honest qualifiers. Losses kept in.

A benchmark is useful only when it tells you whether an engine can clear your workload with acceptable quality, latency, and reliability — not when it collapses to one headline number.

Real data

Core results run on real US-equity data and a real 100,004-asset benchmark — not toy problems.

Matched comparison

Same universe, constraints, costs, and tax rules across every engine. Quality gap reported alongside runtime.

Reproducible

Fixed seeds, recorded runs, low run-to-run variance. Every figure ties to an artifact.

On comparators: baselines are labeled generically — Commercial CPU Baseline 1 / 2 (exact and fast-lane commercial solvers), open-source solver baselines, and rules-based tax-loss harvesting. We compare on outcomes (after-tax wealth, throughput, scale, feasibility), state the regime, and include the cases where a baseline wins. PRISM's results describe outcomes; its methods are proprietary. Read more: how to read optimization benchmarks without being misled →

Proof, not adjectives. Price the whole book before the open.