Sutra advances in stages, each de-risking the next. Three are already done and reproducible; the decisive experiment — a hardware demo — is one phase away. Here is the whole path, honestly placed.
Quantum computing's bottleneck is not the number of physical qubits; it is how many of them each logical qubit costs. That overhead is set by the error-correcting code, the decoder, and how well both fit the hardware. Sutra's bet is that this whole stack is searchable and co-designable — that better codes, better decoders, and hardware-tailored fault tolerance can be discovered systematically rather than hand-crafted one paper at a time. We start where we already have a verified, reproducible result — better codes — and extend outward into the adjacent million-dollar problems of fault tolerance.
A BB-code constructor, distance estimator, heavy-hex metric and decoder — validated to reproduce all five of IBM's published BB codes (parameters and distance) exactly.
60,000 candidates swept (random at scale, then evolutionary) under the heavy-hex L1 ≤ 4 constraint. 14 of the top 50 beat Gross on k·d²/n after rigorous re-verification.
Exact MIP distance on the lead code, Smith-Normal-Form novelty analysis, and a fully reproducible release (code, data, demos) published under CC-BY.
Deeper exact-distance certification, circuit-level threshold runs at scale with a fault-tolerant syndrome schedule, and a live-literature novelty dedupe.
A fault-tolerant decoder schedule and a small-distance memory experiment on real IBM hardware — the experimental anchor the result is missing.
Generalise the pipeline beyond BB codes and into the field's other open problems — decoders, biased-noise tailoring, hardware-native embeddings, logical-gate layout.
A validated engine, verified codes, an exact certification and an open release are in hand. Turning that into a hardware-anchored result is exactly where support compounds.