Concurrent Systems Laboratory

Computer Science Department

Research in the Concurrent Systems Laboratory (CSL) is focused on the design, implementation, and evaluation of computer systems that use state-of-the-art technology to achieve both high performance and high reliability. The focus of this research is often on parallel and distributed systems, exploiting parallelism for high performance and using redundancy in time and space for fault tolerance, leading to high reliability. Single processor, memory system, and input/output architectures are also of interest. The lab is directed by Yuval Tamir.

CSL is always engaged in multiple projects. These projects involve the development of system software, system architectures, chip micro-architecture, and distributed algorithms, as well as simulation studies and tool development.

Current Projects

  1. dependable system-level virtualization
  2. network design automation
  3. data race detection
  4. resource allocation in chip multiprocessors
  5. fault tolerance for chip multiprocessors


Past Projects

  1. dynamically-allocated multi-queue (DAMQ) buffers for switching elements of multiprocessor interconnection networks
  2. crossbar arbitration for switching elements of multiprocessor interconnection networks
  3. micro Rollback for low-overhead error detection in VLSI systems
  4. the UCLA Mirror Processor -- a RISC microprocessor capable of Micro Rollback, concurrent error detection, and rapid self-repair
  5. distributed error recovery schemes for message-passing and shared-memory multicomputers.
  6. fault-tolerant networks of real-time controllers.
  7. host-network interfaces
  8. distributed adaptive routing
  9. reliable network services
  10. fault-tolerant cluster management middleware