Salk Institute Powers AI Research Infrastructure

Overview

The Salk Institute advances life-changing biomedical research across neuroscience, cancer, and aging. To quantify movement patterns via deep learning and computer vision, researchers needed to modernize legacy systems to handle data-heavy AI workloads without processing bottlenecks.

Solution

Salk implemented the VAST AI OS alongside NVIDIA and Run.ai as part of its BioComputing Initiative. VAST provided a unified scale-out architecture and global namespace that consolidates data lakes and streams massive image and video datasets directly to GPU compute nodes.

Results

VAST reduced data ingest latency and eliminated IT management overhead, accelerating AI model iteration cycles. Advanced data reduction shrank research datasets by up to 90%, empowering Salk researchers to process massive biomedical workloads and speed scientific breakthroughs.