ZAP Power joins Nvidia Inception Programme
Deep-tech start-up ZAP Power Ltd, which is developing cryogenic power electronics for next-generation AI data centres, has been selected to join the NVIDIA Inception Programme, NVIDIA’s global programme designed to support cutting-edge technology start-ups.
The programme supports early-stage companies working at the forefront of AI, data centre infrastructure and advanced computing, providing access to technical expertise, industry connections and go-to-market support. Zap Power’s selection recognises the relevance of its work in addressing one of the fastest-growing challenges in AI infrastructure: power efficiency at extreme scale.
Zap Power is developing high-voltage DC power conversion technologies designed to operate within cryogenic environments. As AI workloads drive rack-level power densities to unprecedented levels, conventional power architectures introduce significant losses and heat. Zap Power’s approach aims to reduce these inefficiencies by aligning power conversion more closely with how modern compute hardware actually operates.
Participation in the NVIDIA Inception Programme will enable Zap Power to engage more closely with the AI ecosystem, refine its technology roadmap, and accelerate collaboration with partners across data centre, semiconductor and infrastructure markets.
“Being selected for NVIDIA Inception is a strong validation of both the technical direction we are taking and the urgency of the problem we are addressing,” said the founders of Zap Power. “AI infrastructure is reaching a point where energy efficiency is no longer optional. This programme gives us access to insight and networks that can help us scale faster and more effectively.”
The NVIDIA Inception Programme supports thousands of start-ups worldwide and is widely regarded as a marker of technical credibility within the AI and advanced computing landscape.
Zap Power is currently progressing its technology towards pilot deployments, with a focus on AI data centres and other power-constrained, high-performance computing environments.
