Roche expanded its AI and compute ambitions by striking a deal with Nvidia to build an ‘AI factory’ intended to accelerate drug discovery and diagnostic development. The initiative follows a broader industry push—led by companies like Eli Lilly—to deploy large compute infrastructures and generative models for molecular design, clinical data integration and imaging analytics. Roche said the collaboration will focus on scalable model development and high‑performance compute tailored to pharmaceutical workflows, while Nvidia brings GPU infrastructure and AI software stacks. One‑sentence clarification: an ‘AI factory’ refers to integrated compute, data engineering, and model‑building pipelines that enable iterative, at‑scale AI model development for R&D use cases. Why it matters: the deal signals a continued arms race over compute and AI capabilities in pharma, with implications for target discovery speed, in‑silico candidate generation, and competitive advantage among top R&D spenders.