ACE: a digital floating point CNN emulator engine

Bela Feher, Peter Szolgay, Tamas Roska, Andras G. Radvanyi, Tamas Sziranyi, Marton Csapodi, Karoly Laszlo, Laszlo Nemes, Istvan Szatmari, Geza Toth, Peter L. Venetianer

Research output: Paper

2 Citations (Scopus)

Abstract

The architecture of ACE, a multiprocessor analogic CNN emulator engine consisting of 2 to 16 TMS320C40 floating point DSPs is introduced. The engine containing up to 512 Mbyte RAM (enough to store a 512*512*512* sized CNN cube) can be controlled through its SCSI port. It can either accelerate the multilayer CNN simulator CNNM or be accessed directly from the high level, C-based analogic CNN language ACL to achieve the simulation speed of approx. 2.8 μsec/cell/iteration/DSP for 3*3 linear templates.

Original languageEnglish
Pages273-278
Number of pages6
Publication statusPublished - dec. 1 1996
EventProceedings of the 1996 4th IEEE International Workshop on Cellular Neural Networks, and Their Applications, CNNA-96 - Seville, Spain
Duration: jún. 24 1996jún. 26 1996

Other

OtherProceedings of the 1996 4th IEEE International Workshop on Cellular Neural Networks, and Their Applications, CNNA-96
CitySeville, Spain
Period6/24/966/26/96

    Fingerprint

ASJC Scopus subject areas

  • Software

Cite this

Feher, B., Szolgay, P., Roska, T., Radvanyi, A. G., Sziranyi, T., Csapodi, M., Laszlo, K., Nemes, L., Szatmari, I., Toth, G., & Venetianer, P. L. (1996). ACE: a digital floating point CNN emulator engine. 273-278. Paper presented at Proceedings of the 1996 4th IEEE International Workshop on Cellular Neural Networks, and Their Applications, CNNA-96, Seville, Spain, .