Project Description

CADES is an ORNL facility to support R&D staff’s scalable computing and data analytics needs—making a research computing toolkit available to every research scientist. CADES today supports three protection zones (Open, Moderate, PHI), HPC compute, cloud, AI appliances, graph/analytics appliance, shared memory appliances, high-performance storage, high-speed networking and science DMZ, and object storage.

For our users and collaborators CADES helps accelerate the path to leadership computing and provides an on and off-ramp for data and scalable computing across facilities.

Scientific computing and data analytics needs are changing dramatically and CADES addresses these needs by offering:
• Multi-zone access (Moderate, PHI, Open); science DMZs for cross-facility high-bandwidth network transfers.
• Common data backplane: supporting simulation as well as analytics.
• Infrastructure heterogeneity: cloud, storage, cross-facility data workflows, GPU appliances, and graph appliances.
• Support for new scientific discovery modes: AI/machine learning and workflow middleware.