What You Ought to Know:
– Verily, a precision well being AI firm, as we speak introduced a collaboration with NVIDIA to combine NVIDIA’s full AI know-how stack into its Pre platform.
– The strategic integration is designed to assist Verily’s well being system, life science, payer, and authorities prospects and companions extra effectively develop and deploy AI throughout healthcare.
Unparalleled Acceleration for Omics and AI Growth
The collaboration supplies researchers with quick entry to cutting-edge instruments and {hardware} inside Verily’s trusted analysis atmosphere (TRE), Workbench.
- GPU Integration: Verily has added NVIDIA’s GPU-accelerated libraries and frameworks, together with NVIDIA NeMo, NVIDIA Parabricks, and NVIDIA CUDA-X Knowledge Science. Researchers now have entry to among the newest GPUs, together with NVIDIA Blackwell and NVIDIA Hopper.
- Velocity and Effectivity: This integration has demonstrated the flexibility to enormously speed up analyses, from hours to minutes, when utilizing Parabricks and B200 to course of genomic knowledge in comparison with conventional, CPU-only strategies.
“With NVIDIA’s cutting-edge AI capabilities now out there in our Pre platform, we’re offering researchers with unparalleled instruments to reinforce the pace and effectivity of AI mannequin improvement and omics evaluation,” mentioned Stephen Gillett, Verily CEO.
Powering the NIH All of Us Analysis Program
A crucial factor of this effort is the acceleration of analyses inside the Nationwide Institutes of Well being’s (NIH) All of Us Researcher Workbench. This Workbench, which hosts one of many world’s largest genomics datasets and helps practically 20,000 registered researchers globally, will now be powered by the Pre platform via Verily’s partnership with Vanderbilt College Medical Middle.
As an illustration of the platform’s enhanced functionality, Verily’s researchers developed the primary multimodal basis mannequin utilizing the NIH All of Us Analysis Program dataset. This mannequin integrates digital well being information (EHR) and genomics via polygenic threat scores, enabling new insights into illness prediction and proactive well being administration. Researchers have been capable of practice the mannequin 10x sooner than earlier approaches utilizing NVIDIA NeMo Automodel and H100 GPUs.
Increasing AI Capabilities Throughout the Pre Platform
Past the Workbench integration, Verily plans to make NVIDIA GPU-accelerated instruments out there throughout different Pre options, enhancing knowledge administration and trade.
- Refinery: This highly effective curation engine ingests, harmonizes, and enriches multimodal knowledge, utilizing a clinically knowledgeable, FHIR-native knowledge mannequin to remodel unstructured knowledge into structured knowledge belongings.
- Trade: A safe discussion board for sharing and accessing AI-ready datasets, fashions, and brokers to advance precision analysis and care.
