Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Jump to content

Tile processor

From Wikipedia, the free encyclopedia

Tile processors[1] for computer hardware, are multicore or manycore chips that contain one-dimensional, or more commonly, two-dimensional arrays of identical tiles. Each tile comprises a compute unit (or a processing engine or CPU), caches and a switch. Tiles can be viewed as adding a switch to each core, where a core comprises a compute unit and caches.

In a typical Tile Processor configuration, the switches in each of the tiles are connected to each other using one or more mesh networks.[2] The Tilera TILEPro64, for example, contains 64 tiles. Each of the tiles comprises a CPU, L1 and L2 caches, and switches for several mesh networks.

Other processors in a tile configuration include SEAforth24, Kilocore KC256, XMOS xCORE microcontrollers, and some massively parallel processor arrays.

References

[edit]
  1. ^ "The Tile Processor™ architecture: Embedded multicore for networking and digital multimedia - IEEE Conference Publication". August 2007: 1–12. doi:10.1109/HOTCHIPS.2007.7482495. S2CID 44858928. {{cite journal}}: Cite journal requires |journal= (help)
  2. ^ Wentzlaff, David (September 15, 2007). "On-Chip Interconnection Architecture of the Tile Processor" (PDF). IEEE Micro. doi:10.1109/MM.2007.89