Signaloid massively accelerates computations involving probability distributions, enabling efficient uncertainty quantification for AI models and making data-driven computing systems more relevant and safe.
Signaloid is a spinout from Cambridge University that was founded by Phillip Stanley-Marbell (Assoc. Prof Cambridge and MIT). Signaloid provides a computing platform to quantify, transmit, and track data or modelling uncertainty – flexibly and dynamically – without the need to adjust computing workloads or code. Applications include materials modelling, autonomous systems, computational finance, machine learning, and noisy intermediate-scale quantum computing. The technology is capable of processing workloads orders of magnitude faster than incumbent methods via its proprietary computing platform.
Signaloid was the first recipient of ARIA's AI Hardware project funding, having been awarded £5m to develop a silicon implementation of their technology to accelerate matrix inversion calculations for AI/ML applications.
