ZML is betting on fast inference across AI chips — and it could reshape the market

I’d like to share a development from the AI and inference space.

French startup ZML has launched LLMD, a free server for running open-source LLMs across a wide range of chips: Nvidia, AMD, Google TPU, Apple Metal and Intel Arc.

The goal is straightforward: break down technological silos, reduce vendor lock-in, and give companies more flexibility in choosing their infrastructure.

Importantly, the product is not open source yet. ZML is offering it for free to study adoption and later monetize where it creates the most value.

Why this matters: inference optimization is becoming critical to the cost and scalability of AI.

Are you already looking at hybrid AI infrastructures?
#AI #inference #startups #HackerNews