The ADSP aims to accelerate discovery across a broad range of scientific domains which require data-intensive and machine learning algorithms to address challenging research problems. The program connects researchers with ALCF scientists to push the state-of-the-art in machine learning, workflows, data analysis, and algorithmic development.
The ADSP can be used to achieve computational readiness for ALCF programs such as INCITE and run on future systems such as Aurora, a new Intel-HPE system system which will be capable of over 1 exaflops.
The ADSP issues a regular call for proposals each year (in the spring). The program employs a competitive proposal process that awards allocations of time on ALCF supercomputers. All proposals are peer reviewed by a panel of experts for both the scientific impact and computational readiness. ADSP allocations are for two years and are awarded to researchers from academia, government research facilities, and industry. PIs will be required to submit a renewal application for the second year of the award.
For more details, please see the ADSP call information page.