Secure Research Data & Computing (SRDC) Platform
Research IT offers the Secure Research Data and Compute (SRDC) Platform for researchers working with highly sensitive (P4) data. SRDC is a secure compute environment that includes high performance computing (HPC) with GPUs, computing on virtual machines (VMs) with Linux and Windows servers, and protected storage for both options.
The SRDC Platform has been designed for use by a broad range of researchers, without compromising usability or performance. Researchers working with protected health information (PHI), other highly sensitive human subjects data, controlled unclassified information (CUI), or other high risk data coming from third parties are strong candidates for SRDC.
A no-cost allocation on the Platform is available to all principal investigators. In addition, researchers may scale up their resources by purchasing additional compute nodes which may be added to the system platform. This is often referred to as a "condo" model. * Condo computing is a shared ownership model in which researchers use equipment funds from grants or other sources to purchase and contribute compute servers (i.e., nodes) to the system. A detail overview of the system with technical specifications may be found here.
To learn more about SRDC, please send email to: srdc-serviceteam@lists.berkeley.edu.
High Performance Computing with Moderately Sensitive Data
Savio, a high performance computing cluster, may be used for moderately sensitive (P2/P3) data(link is external)(link is external). Savio is suitable for a wide range of research applications, including tightly coupled applications that require a low latency, high bandwidth interconnect, or very fast I/O.
Researchers working with de-identified public health or human genetic data are strong candidates for Savio.
Computing on Virtual Machines with Moderately Sensitive Data
The Analytics Environment on Demand (AEoD) Service is offered to departments, organizations and schools on campus, and provides Windows and Linux virtual machines typically suitable for researchers who need more compute power than is offered by a standard laptop or desktop. AEoD is offered as a partnership model with Research IT as one partner, providing the compuational infrastructure, and a department or org as the other partner, offering the local IT expertise to manage the virtual machines. AEoD virtual machines may be secured to work with moderately sensitive (P2/P3) data.
AEoD is most suitable for researchers needing interactive computing, using common software packages (Stata, ArcGIS, RStudio, etc.) in a familiar desktop environment. AEoD virtual machines may be scaled to meet computational needs, and are available in different sizes, from 2 - 20 cores, 4 - 256 GB RAM, plus performant storage...
Researchers performing data analytics or geospatial analysis with moderately sensitive public health information, FERPA(link is external)(link is external) data, or data related to animal research are strong candidates for AEoD.
To learn more about AEoD, contact Research IT consulting: research-it-consulting@lists.berkeley.edu