MatX: MatX is an AI chip startup founded by former Google engineers specializing in hardware for large language models. It aims to outperform Nvidia GPUs in AI inference efficiency through custom silicon designs. The startup attracted significant venture backing to accelerate its challenge to established players.
Olix: Olix is a U.K. startup developing photonics-based processors tailored for AI computing. It emphasizes energy-efficient alternatives to conventional chip architectures. Investors are backing it amid Europe’s booming AI chip scene.
Arago: Arago is a Paris-based startup developing photonic chips to reduce energy consumption in AI systems. It emerged from stealth with focus on light-based computing for inference. Reports indicate preparations for substantial new funding.
Etched: Etched is a San Jose-based AI chip startup founded by Google alumni developing specialized processors for AI workloads. It targets superior inference performance over traditional GPUs. The company raised substantial capital to enter the market dominated by Nvidia.
Euclyd: Euclyd is a European AI chip startup working on alternatives to Nvidia’s dominance in inference hardware. It is actively seeking major funding to scale production. The company represents growing momentum in non-U.S. AI silicon development.
Nvidia: Nvidia is the dominant semiconductor company providing GPUs repurposed for AI training and inference workloads. It leverages its vast resources to acquire inference technologies from startups like Groq and invest in photonics for next-generation chips. The news highlights Nvidia racing against a surge of well-funded rivals developing specialized inference hardware.
Axelera: Axelera is a Netherlands-based AI semiconductor company building in-memory computing chips for edge and data center inference. It ships hardware to global customers for physical AI applications. The startup recently expanded funding to fuel commercial growth in AI acceleration.
Fractile: Fractile is a U.K. AI chip startup creating novel architectures for efficient inference deployment. It has backing from strategic investors like the NATO Innovation Fund. The company is planning a large funding round to compete in the heated AI hardware race.
Seedcamp: Seedcamp is a prominent European seed-stage venture capital firm investing in deep tech and AI infrastructure. It supports hardware innovations like reversible computing chips for energy efficiency. The firm backed Vaire Computing amid surging interest in AI chip challengers.
Ayar Labs: Ayar Labs specializes in photonic interconnects to enhance data transfer in AI systems. It focuses on integrating optical engines into AI hardware for improved performance and energy savings. Recent funding supports ramping production for next-generation AI infrastructure.
Optalysys: Optalysys is a U.K.-based firm pioneering photonic computing solutions for AI acceleration. It focuses on hardware that leverages light for faster, more efficient processing. Plans for significant fundraising underscore its role in the inference shift.
Carlos Espinal: Carlos Espinal is the managing partner at Seedcamp, an early-stage European venture firm. He views AI chip investments as shifting from niche to core infrastructure plays. In the news, he highlighted Seedcamp’s backing of innovative chip startup Vaire Computing.
Cerebras Systems: Cerebras Systems develops large-scale AI processors optimized for intensive computing tasks. It positions itself as a direct challenger to Nvidia by focusing on architectures suited for AI inference. The company recently secured major late-stage funding to expand its role in the competitive AI chip market.
Nato Innovation Fund: The NATO Innovation Fund is a venture capital initiative by NATO allies investing in dual-use deep technologies including AI and defense hardware. It backs startups developing efficient AI chips to address geopolitical tech dependencies. The fund has invested in U.K.-based Fractile, amid rising European AI chip activity.
Patrick Schneider-Sikorsky: Patrick Schneider-Sikorsky is a director at the NATO Innovation Fund focusing on deep tech investments. He emphasized that existing GPU architectures fall short for large-scale AI inference, supporting bets on novel chip designs. In the news, he commented on the shift toward inference-optimized hardware via investments like Fractile.

European Growth: AI chip developers in Europe are pursuing large funding rounds to build independent inference hardware stacks.
Inference Shift: Attention in AI hardware has moved from training-focused GPUs to purpose-built chips for efficient model deployment in applications.
Nvidia Strategy: Nvidia counters rivals through licensing deals like Groq for inference tech and investments in photonics innovations.
Investor Confidence: Venture firms now regard AI chip startups as essential bets for future infrastructure rather than speculative niches.