Friday, May 1, 2026
Search

NVIDIA Locks In Eli Lilly and Thermo Fisher as BioNeMo Becomes Global Pharma AI Backbone

NVIDIA signed simultaneous AI platform deals with Eli Lilly and Thermo Fisher in January 2026, establishing BioNeMo as the foundational infrastructure layer for pharmaceutical research worldwide. The agreements triggered foundation model launches from at least five specialized biotech firms across Europe and North America. The move mirrors how AWS and Azure captured cloud infrastructure — before application markets matured.

Salvado
Salvado

May 1, 2026

NVIDIA Locks In Eli Lilly and Thermo Fisher as BioNeMo Becomes Global Pharma AI Backbone
Image generated by AI for illustrative purposes. Not actual footage or photography from the reported events.
Loading stream...

NVIDIA signed AI platform agreements with Eli Lilly and Thermo Fisher simultaneously in January 2026, positioning BioNeMo as the default infrastructure layer for global pharmaceutical research.1

The deals immediately triggered foundation model launches from five specialized biotech firms: Basecamp Research, Boltz, Owkin, Edison Scientific, and Natera — spanning biodiversity genomics, biomolecular structure prediction, and federated hospital data.1

BioNeMo provides pre-trained biological AI models covering protein structure, molecular generation, and genomics. Drug developers integrate these directly into pipelines rather than building from scratch. The model lowers the barrier for mid-sized biotech firms globally that lack NVIDIA's compute scale or the research headcount of competitors like Google DeepMind in London.

The infrastructure-layer strategy is well-tested. AWS and Azure became the substrate for entire software ecosystems. NVIDIA is replicating that playbook in life sciences — providing the platform, capturing margin, and staying out of direct competition with the applications built on top.

Thermo Fisher's partnership extends BioNeMo's reach into physical laboratory instrumentation across global scientific facilities. That hardware-software integration — connecting AI models to bench-level workflows — is significantly harder to replicate than software-only agreements.

European and North American biotech firms are building on shared infrastructure rather than maintaining proprietary compute. Owkin, based in Paris and New York, specifically targets federated learning across hospital datasets in multiple jurisdictions. Each firm specializes; all increasingly depend on NVIDIA's substrate.

Regulatory frameworks for AI-assisted drug approval remain incomplete across most jurisdictions — the EU AI Act does not yet fully address pharmaceutical applications, and the US FDA is still developing its AI/ML guidance for drug development. The infrastructure layer is consolidating faster than the regulatory layer can follow.

NVIDIA's position has shifted structurally. It is no longer primarily a chip supplier to pharma — it is a platform owner with recurring relationships at the top of the global drug development stack. That is a different and more durable business.1


Sources:
1 "NVIDIA BioNeMo Platform Adopted by Life Sciences Leaders to Accelerate AI-Driven Drug Discovery" — Finance.Yahoo

Salvado
Salvado

Tracking how AI changes money.