High Performance Computing

janelia7_blocks-janelia7_secondary_menu | block

janelia7_blocks-janelia7_fake_breadcrumb | block

janelia7_blocks-janelia7_support_team_header | block

What We Do

node:body | entity_field

World-class computational infrastructure for Janelia's science.

Janelia's scientists reconstruct whole-brain connectomes from electron microscopy volumes measured in hundreds of terabytes. They train deep learning models on high-dimensional neural recordings, segment subcellular structures from FIB-SEM datasets, and stream multi-photon imaging data at gigabytes per second. With the launch of AI@HHMI - a $500 million, ten-year investment to embed AI throughout the scientific process - these computational demands are accelerating rapidly, as researchers build automated experiment pipelines, train generative models on massive biological datasets, and run distributed GPU workloads at a scale Janelia has never seen before.

The High Performance Computing department builds and operates the infrastructure behind all of it. The team manages over 25 petabytes of scientific data across multiple storage tiers and runs a Linux-based HPC cluster with more than 10,000 cores and over 600 GPUs in Janelia’s data center, supplemented by cloud resources for additional scalability. As Janelia's research grows more computationally intensive, the department continues to expand its capacity - adding next-generation GPU systems, scaling storage to keep pace with data acquisition, and helping researchers move seamlessly between on-premise and cloud environments.

janelia7_blocks-janelia7_block_right_hand_rail | block

Main Menu (Mobile)- Block

Main Menu - Block

High Performance Computing

What We Do