TrueLoop — push your devices past the limit
Stateful Wave Computing · adaptive control & optimization middleware<br>Push your devices<br>past the limit.<br>A model-free, retained-state runtime between your application and your hardware. It turns each measurement into the next configuration — so your control, optimization, calibration, and certified-randomness loops run tighter, at higher dimension, on fewer measurements, under more drift.<br>Get a 90-day key →Watch it converge ↓<br>Drop the client into your loop in one line. No fee, no card.
scroll<br>Where it sits<br>A thin adaptive layer between the objective and the physics.<br>The runtime is middleware, not a replacement for anything you already run. The application or model that sets the objective stays where it is; the wave substrate that realizes it stays where it is; the runtime is the control plane in between, closing the loop each round. Perception and learning keep their place, the physics keeps its place, and the adaptive work that used to need a model or a search now happens in one measured step.
What it lets you do<br>It is a lever on your operating point.<br>Most loops are not limited by the problem — they are limited by the cost of the loop around it. Classical control needs a model and many measurements; classical search needs time. So loops run conservatively. The runtime relaxes the binding constraint: one measurement per round, no model, converging from the first — so you can push a loop you already run into a harder regime than its current control allows:
measurements to spare<br>→<br>one measurement per round — when each read costs money, time, dose, or shots
slow / batch cadence<br>→<br>real-time — per-tick, per-frame, per-packet decisions on the same hardware
a few channels<br>→<br>thousands — tracking quality stays flat as dimension grows, at linear cost
a comfortable deadline<br>→<br>milliseconds — a usable answer where sequential search runs out of clock
a tuned model per device<br>→<br>model-free — the same loop across devices and drift, no re-tuning
a gentle operating point<br>→<br>harder, driftier, more nonlinear — where a fixed-gain loop would lose lock
If your task has a writable control and a measurable output, the runtime can run it harder — not a new class of problem, but more from the loops and devices you already have. It declines, by design, where a classical method already wins: when measurements and time are abundant, or when the objective lives in correlations the marginals cannot see.
The headline results<br>The advantage grows with dimension and complexity.<br>Two distinct mechanisms point the same way. The measurement-efficiency edge is structural and unbounded; the optimization edge appears under a starved budget, where a digital solver runs out of evaluations and the runtime keeps improving as the problem grows. Every figure below is established against fairly-tuned baselines.
Every figure is established in controlled benchmarks against fairly-tuned baselines, with hardware validation in progress. The smaller, equally-honest results — the retained-state memory signature, the substrate-as-optimizer mode, and the cases where it declines — are detailed across the docs and compatibility page.
Where it wins<br>Three regimes where restarting a digital search loses.<br>These are the conditions you push a loop into to get the advantage. The runtime replaces gradient estimation with a residual-feedback update and carries state across related problems — which changes the outcome in three places, and the client tells you honestly when you’re outside them.
01 / DRIFT<br>Hold a target as the device moves<br>One measurement per round, no model, no per-device tuning — tracking 3 to 10 times tighter than finite-difference and SPSA gradient methods, which cannot assemble a step before the target drifts away.
02 / DEADLINE<br>Decide before search finishes<br>When the clock runs out before a digital solver can search, the runtime hands back a usable answer first — the regime where the crossover in the results above starts to bite.
03 / RELATED TASKS<br>Reuse the last solution<br>Retained configuration warm-starts each related problem from the previous physical trajectory — a re-convergence advantage that grows with relatedness and vanishes, correctly, when tasks are unrelated.
Outside these — static problems at ample budget, a well-tuned PID with time to settle, or objectives that live in dense correlations — a conventional method does as well or better, and the client returns DECLINE rather than overstating the fit.
See the yield curves live →·Read the QRNG docs →
How it works<br>One loop. Run it on our endpoint, or offline in your own systems.<br>You hand the runtime a measured statistic each round; it returns the next configuration. During the free evaluation the residual update law runs server-side on our endpoint — you receive results and next-configs, never the method. With a paid license you can also run the runtime offline inside your...