Signaloid

Signaloid is a Cambridge, UK computing company spun out of the University of Cambridge that builds hardware and a cloud platform for probabilistic workloads. Its UxHw (distribution-extended compute hardware) technology lets existing, unmodified applications compute directly on probability distributions, automating work that normally needs costly code changes or long Monte Carlo runs: Monte Carlo methods in finance and physics, importance sampling in reinforcement learning, and particle filters in physical AI and robotics. The company claims 10x–1,000x speedups on key workloads, and up to 1,000x better performance-per-watt for its C0-ASIC; these are company figures, not independent benchmarks. The platform ships as the Signaloid Compute Engine (including an AWS Marketplace machine image from June 2026), as edge hardware modules, and as the C0-ASIC, which has taped out, with engineering samples due to the first customer in Q3 2026; the UK Advanced Research and Invention Agency (ARIA) commissioned the ASIC for evaluation. Target markets are finance (pricing and risk models) and manufacturing (explainable AI-driven predictive maintenance), with Boeing named as a manufacturing customer and partnerships with imec and TSMC for the high-end hardware. Founded in 2019 (Signaloid Limited, Companies House 12028829, incorporated 3 June 2019) by CEO Phillip Stanley-Marbell, previously at Bell Labs, IBM, Apple, MIT and Cambridge. It raised a £3M seed round in 2023 led by IQ Capital with Type One Ventures and Creator Fund, and has said it is raising a Series A; no valuation has been disclosed.