NUMAchine is a cache-coherent shared-memory multiprocessor designed to have high-performance, be cost-effective, modular, and easy to program. Processors, caches, and memory are distributed across a number of stations interconnected by a hierarchy of unidirectional bit-parallel rings.

A 64-processor prototype of NUMAchine is currently being built at the University of Toronto. It will have 4 stations connected by a local ring and 4 local rings connected by a central ring. A station will have 4 R4400 MIPS processors, each with a 1-Mbyte second-level cache, 256 Mbytes of station memory, I/O devices and an 8-Mbyte Network Cache to cache off-station data.

  • NUMAchine Technical Report
  • Photographs of NUMAchine Hardware