SNN Studio

What this studio is

Stochastic computing and neuromorphic hardware co-design toolkit. SC-NeuroCore is a research-to-hardware software stack for designing spiking and stochastic neural systems, validating their numerical behaviour, and moving selected models toward FPGA, ASIC, and embedded neuromorphic deployment.

Universal Stochastic Computing Framework for Neuromorphic Hardware — Rust SIMD engine, Python simulation, Verilog RTL, HDC/VSA, SCPN integration

Declared capabilities

The studio's capability manifest declares 123 capabilities and emits 7 evidence types through the federation gate.

Verbs

Compute backends

Evidence types

• studio.backend-benchmark.v1
• studio.bitstream-encoding.v1
• studio.cosim-parity.v1
• studio.fpga-deployment.v1
• studio.rtl-compilation.v1
• studio.sc-inference.v1
• studio.spike-analysis.v1

Evidence

Evidence	Schema	Kind	Boundary	Exactness
sc-neurocore-sc-inference-parity	studio.sc-inference.v1	measured	reference-validated	bit-exact
sc-neurocore-fpga-shd-synthesis	studio.fpga-deployment.v1	hardware-validated	reference-validated	bit-exact

Download the evidence bundle (RO-Crate) — measured, boundary reference-validated.

Measured performance

This studio publishes its multi-language benchmark databank — 65 of 70 cells measured, sealed and recorded in the transparency log, with the dispatch-order contradictions shown as findings.

Open the benchmark databank →

Published work

• SC-NeuroCore: Universal Stochastic Computing Framework for Neuromorphic Hardware

• Bit-exact benchmark suite for neuromorphic inference — An 8-capability, 70-cell benchmark databank that compares stochastic-computing and neuromorphic kernels across Rust, Mojo, Julia, Go, and Python — every cell emitted as a studio.sc-inference.v1 evidence bundle with claim boundary and exactness class.