Sherlock High Performance Computing Cluster
Do you need access to computational resources above and beyond what your laptop/desktop can handle to support your sponsored or departmental research? You may want to try out the Stanford Sherlock cluster. By using Sherlock researchers can focus on their research and not on system, networking and hardware administration. Purchased and supported with seed funding from the Provost, Sherlock comprises more than 1,400 compute servers and associated petabyte scale storage. More than 100 of those servers are available to any Stanford PI to run their computational codes and programs, with resources managed through a fair-share algorithm using the SLURM scheduler.
Faculty can also purchase additional dedicated resources to augment Sherlock by becoming an owner. Choosing from a standard set of server configurations supported by the SRCC staff, owners' servers are "joined" to the base Sherlock cluster. Owners have access to that base University-funded set of servers, through fair-share. But they also have priority access to the resources they purchased, whenever they want. When an owner's servers aren't in use, other owners can use them ... but non-owners cannot. The base Sherlock configuration is 170 servers. Since June 2014, an additional 1,312 servers have been added by owners. Whether an owner or not, researchers using Sherlock have access to more than 400 different computational tools and codes prebuilt by the Research Computing team.
As of July 2019 Sherlock comprises 1,483 compute nodes, 26,748 CPU cores, 728 GPUs and 2,053 TFlops of computing power used by 684 Principal Investigators and their approximately 4,100 research team members.