Enabling High Performance Computing in Cloud Infrastructure using Virtualized GPUs

With the advent of virtualization and Infrastructure-as-a-Service (IaaS), the broader scientific computing community is considering the use of clouds for their technical computing needs. This is due to the relative scalability, ease of use, advanced user environment customization abilities clouds provide, as well as many novel computing paradigms available for data-intensive applications. However, there is still a notable gap that exists between the performance of IaaS when compared to typical high performance computing (HPC) resources, limiting the applicability of IaaS for many potential scientific users. Most recently, general-purpose graphics processing units(GPGPUs or GPUs) have become commonplace within high performance supercomputers. We propose to bridge the gap between supercomputing and Clouds by providing GPU-enabled virtual machines. Specifically, the Xen hypervisor is utilized to leverage specialized hardware-assisted I/O virtualization tools in order to provide advanced HPC-centric Nvidia GPUs directly in guest VMs. We evaluate this work by measuring the performance of two Nvidia Tesla GPUs and comparing to bare-metal hardware. Results show this method of leveraging GPUs within virtual machines is a viable use case for many scientific computing workflows, and could help support high performance cloud infrastructure in the near future.