12 companies

Showing 1–12 of 12 companies

Mythic

Mythic builds Analog Processing Units that merge compute and memory for energy-efficient AI inference, targeting data centres, automotive, robotics and defence.

femtoAI

femtoAI builds ultra-efficient edge AI inference technology and custom sparse processing hardware, enabling AI to run on battery-powered devices without cloud connectivity.

Neurophos

Neurophos builds photonic AI inference hardware - silicon photonics and programmable metasurfaces that do matrix multiplications optically - targeting power-constrained AI workloads in data centres.

Positron AI

Positron AI builds purpose-built hardware accelerators for large language model inference, aiming to provide an efficient and affordable alternative to GPU-based solutions.

Normal Computing Corporation

Normal Computing builds AI infrastructure software and physics-based hardware, including thermodynamic computing ASICs and an EDA platform for semiconductor engineering teams.

Tensordyne

Tensordyne builds custom silicon, hardware and software for low-power, high-performance multimodal generative AI inference, including its flagship Napier inference system.

Nexthop AI

Nexthop AI builds customized networking hardware, hardened network operating systems and pre-tested optical and electrical interconnects for hyperscalers and neo-clouds running AI workloads.

Sygaldry Technologies

Sygaldry Technologies builds hybrid quantum-classical AI servers designed to deliver exponential speedup for AI training and inference.

d-Matrix

d-Matrix builds specialised AI inference hardware using digital in-memory compute to reduce data movement and improve performance for generative AI workloads.

P-1 AI

P-1 AI builds engineering AGI for the physical world through Archie, an AI agent that automates cognitive tasks in the design of complex physical systems.

Delos Data Inc

Delos Data Inc builds cluster management software and server designs for large-scale AI inference data centers, enabling GPU scale-up domains above 1,000 GPUs.

Cerebras Systems

Cerebras Systems builds wafer-scale AI supercomputer chips that replace clusters of GPUs with a single, much larger processor for large-scale artificial intelligence workloads.