At this week’s J.P. Morgan Healthcare Conference in San Francisco, Nvidia made a series of announcements focused on leveraging its AI capabilities by partnering with healthcare leaders to tackle key medical challenges.
AI has tremendous potential to change the way we live, with some of the biggest applications in healthcare, helping people live longer and healthier lives.
The Santa Clara AI powerhouse’s announcements included:
- Partnering with pharma giant Lilly to build a $1 billion joint research lab (including compute, talent, and infrastructure over five years) in the San Francisco area that will use AI to accelerate development and availability of new drugs and medical treatments.
- Collaborating with Thermo Fisher Scientific on an autonomous lab infrastructure to make scientific instruments intelligent and laboratories more autonomous to drive scalable scientific discovery.
- A major expansion of Nvidia’s BioNeMo open development platform for AI-driven biology and drug development.
Improving clinical development with AI models
The co-innovation lab with Lilly is a five-year commitment dedicated to accelerating closed-loop discovery and creating AI models to improve clinical development. The two companies will also be working together on opportunities to apply AI across Lilly’s businesses, from manufacturing to commercial operations.
By combining Lilly’s deep domain expertise in drug discovery with Nvidia’s expertise in AI and accelerated computing, the two companies aim to accelerate how medicines will be designed and developed.
During the event, Nvidia cited the growth of powerful open AI models as a key driver of healthcare discoveries and delivery. Open models are the backbone of modern innovation. Nvidia provided several supporting data points, including 80% of the world’s startups are built on open models.
Additionally, for the first time in history, healthcare — particularly the United States’ $4.9 trillion industry — is deploying AI at nearly three times the pace of the broader US economy.
Alliance with Thermo Fisher to advance scientific AI infrastructure
Nvidia and Thermo Fisher Scientific, a leader in providing a wide array of products, services, and solutions for healthcare and research markets, are collaborating to power AI-based solutions and laboratory automation. The work will use Nvidia’s AI platforms and Thermo Fisher’s solutions to drive growth in laboratory automation, accuracy, and speed.
Leveraging NVIDIA DGX Spark as a benchtop AI supercomputer, Thermo and Nvidia built and deployed quality control and analytics agents to help scientists interpret data as it’s generated, flag anomalies, validate results, and recommend adjustments before the next run, making each experiment smarter than the last.
The ultimate goal would be to build a fully autonomous lab.
Expanding the BioNeMo open-development platform
Nvidia’s BioNeMo, a development platform and software ecosystem designed to accelerate AI-driven drug discovery and other biological research, is now a full open-development platform that enables “lab-in-the-loop” scientific workflows for biology and drug discovery.
BioNeMo includes a suite of open models, including RNAPro for predicting the 3D structure of RNA molecules, a new “version two” reasoning model for molecular synthesis, and KERMT, a chemistry-focused AI model jointly developed by Nvidia and Merck for drug discovery.
Nvidia is also introducing new libraries for large-scale molecular AI workloads, providing 10 to sometimes 100 times, depending on the workflow, speedup of turning hours of chemistry processing into minutes, and an open-source recipe for easily accelerating and scaling biology and chemistry foundation model training.
BioNeMo enables the full lifecycle of AI development in life sciences by connecting experimental data directly to model training. This turns every experiment into digital intelligence and creates a closed loop that continuously accelerates scientific discovery.
Multiplying training and robot deployment
Nvidia also announced a partnership with Multiply Labs, a San Francisco-based robotics company that builds automated manufacturing systems to produce advanced, individualized cell and gene therapies at an industrial scale.
Multiply will use Nvidia’s “three-computer solution” — NVIDIA DGX AI supercomputers for AI training, NVIDIA Omniverse and Cosmos on NVIDIA RTX PRO Servers for simulation, and NVIDIA Jetson AGX Thor for on-robot inference — to accelerate the training and deployment of their robots.
Using NVIDIA Omniverse and NVIDIA Isaac, Multiply creates digital twins of entire labs to train robots on thousands of sterile, precision tasks before they ever touch the real world, combining simulation, imitation learning, and Nvidia’s GR00T foundation model.
Multiply Labs is replacing manual, contamination-prone work with automated systems, and Nvidia quantified the benefits. In cell therapy manufacturing labs, robot systems have dropped per-dose costs by 70%. Dose throughput in the labs is also being increased by 100 times per square foot.”
Final thoughts
With AI, Nvidia has been a thought leader and visionary for the technology.
The company understands that AI has the potential to create a “rising tide” that could lead to an economic boom larger than the one the Internet brought. Its primary message to the industry is around openness, as that allows more companies to benefit.
These healthcare announcements are good examples, as they allow several partners to forge ahead and bring us closer to a future with accelerated development of more life-saving drugs and other treatments for the benefit of people everywhere.
Also read: A $28 million seed round and Pfizer partnership highlights how biomolecular AI is moving deeper into drug discovery.
