Introducing the AI ML and Genomics Workloads (SDC 2019)

preview_player
Показать описание
"Both the Artificial Intelligence/Machine Learning and Genomics fields are growing quickly - consuming large amounts of compute and storage resources across labs and data centers. The SPEC Storage subcommittee, creator of the SPEC SFS 2014 benchmark, has been working with subject matter experts in the AI/ML and Genomics fields to capture their unique storage workload profiles. We have distilled the network and system call traces down to realistic industry-standard workloads.

In this session, we will review these workload definitions and show what results from SPEC Storage benchmark runs look like. These workloads will be available in a future release from the SPEC Storage subcommittee, expected in the first half of 2020."
Рекомендации по теме