摘要: Two novel algorithms for simultaneous computation of a large number partial sums are introduced, their performance assessed, and architectures implementation suggested. The direct D operations replaced by O(D/log D). new approach is based on concept optimal partitioning redundancy removal in arithmetic intensive, high throughput computing that expected to be the basis class which represent a, departure from brute force parallel where inherent not detected or removed.
computer.org
ieeecomputersociety.org
osti.gov参考文章(5)
H.J. Caulfield, J.A. Neff, W.T. Rhodes, Optical computing: the coming revolution in optical signal processing Laser focus. ,vol. 11, pp. 100- 110 ,(1983)
Adly Fam, A multi-signal bus architecture for FIR filters with single bit coefficients international conference on acoustics, speech, and signal processing. ,vol. 9, pp. 468- 470 ,(1984) , 10.1109/ICASSP.1984.1172335
Adly T. Fam, An algorithm for parallel computation of partial sums technical symposium on computer science education. pp. 131- 133 ,(1984) , 10.1145/800014.808152
Kung, Why systolic architectures IEEE Computer. ,vol. 15, pp. 300- 309 ,(1982) , 10.1109/MC.1982.1653825
Adly Fam, Space-time duality in digital filter structures IEEE Transactions on Acoustics, Speech, and Signal Processing. ,vol. 31, pp. 550- 556 ,(1983) , 10.1109/TASSP.1983.1164114