Parallel Computing & GPUs

UC Davis has various resources available for parallel computing, including GPU compute clusters.

HPC@UCD

The UC Davis High Performance Computing Core Facility houses and administers many of the research clusters on campus. Math has previously had a share in Peloton which may still be active, and some of these clusters may be more broadly available to the research community. The College of Letters and Sciences (CLAS) has sponsored two from the list: Demon and Peloton.

Click here to eee the list of HPC supported clusters, and read more on their website.

MSBC — GPU Cluster

The Departments of Mathematics and Statistics have a shared GPU cluster housed in the Math Sciences Building, using funding shared with Letters and Sciences, and administered by Statistics staff. This is intended for use in both courses and research. Department members should email help@math.ucdavis.edu for access, and provide their campus username. Access is via SSH, through campus IPs or VPN (UC Davis has VPN instructions on their support website).

Updates and details about the joint Math-Stats GPU cluster is available on the website.

Affiliated Organizations

Several Department members also work with the UC Davis TETRAPODS Institute of Data Science (UCD4IDS), an interdisciplinary group focusing on foundations of data science and machine learning.

Outside Resources

Some businesses offer compute services. Here are some that offer free services.

  • ACCESS by NSF - Previously known as XSEDE, this is an NSF program to help US researchers, educators, and students to get access to large-scale computing resources at no cost. Their 'About' page is very informative.
  • Google Cloud for Researchers - Faculty, postdocs or PhD graduate students can submit a proposal for $1000-5000 of credit for use with Google's cloud services. Their website has a tool that lets you estimate how much time/credit you might need for your project.
  • Microsoft Azure for Students - Students can have $100 free credit per year for use with Microsoft's Azure cloud computing.
  • Microsoft Azure for Academic Research - There are free credits to use Azure cloud computing for research and education purposes. Details are difficult to pinpoint.
  • AWS Cloud Credits for Research - Amazon's AWS has very specific support of up to $5000 for publicly available research or broadening of cloud research.