Prof. Amnon Barak
  The High-Performance Computing Systems research group

The group performs groundbreaking research and development in management
and run-time optimizations of high-end computing systems.

Areas of current activities include:

  • Algorithms for adaptive resource management in clouds.
  • Randomized and deterministic algorithms for resilient computing in Exascale clusters.
  • Run-time optimizations for accelerators (e.g. GPUs) in clusters using memory access patterns.
  • A cluster management system for providing a single-system image.
  • Parallel and distributed systems, algorithms and applications.
      On-going and previous projects

  Research interests:
  • Virtualization and run-time optimizations in clusters with multiple GPUs:
      the VirtualCL (VCL) and the MAPS Framework projects.
  • Parallel and distributed algorithms for resilient computing in Exascale clusters:
      the FFMK project.
  • Management of clusters and multi-cluster clouds: the MOSIX project.
  • High performance parallel and distributed computing.
      Selected publications           Erdös number: 1

  Research students:

I am looking for bright and highly motivated students at all levels,
who are interested to perform innovative R&D in my research group.

  The distributed operating systems lab course (67845)

Individual projects for graduate and advanced undergraduate students.
Possible topics: distributed systems, high-performance computing,
randomized and deterministic resilient algorithms,
parallel and distributed computing and applications,
management of clusters and clouds, operating systems, scheduling algorithms.