Efficient correction of normalizer shift amount errors in fused multiply add operations
作者： Charles Tsen , Adam Dreyer
DOI:
关键词:
摘要: A method for correcting a shift error in fused multiply add operation. The comprises adjusting normalized floating-point number before performing correction to produce an adjusted number, and the number. shifting mantissa of one direction. module comprising normalizer module, compensation logic, round. is operable normalize based upon estimated quantity leading zeros. logic manage rounder correct with only
lens.org 参考文章(80)
Heonchul Park, Le Trong Nguyen, Processor that decodes a multi-cycle instruction into single-cycle micro-instructions and schedules execution of the micro-instructions ,(1997)
Stefan Demetrescu, High speed memory and processor system for raster display ,(1984)
Edward A. Hutchins, Arithmetic logic units in series in a graphics pipeline ,(2004)
David C. Wyland, Risc processor using register codes for expanded instruction set ,(2000)
Ethan W. Gannett, Fragment visibility pretest system and methodology for improved performance of a graphics system ,(1997)
Ching-Yu Hung, Wissam A. Rabadi, Leonardo W. Estevez, Reconfigurable SIMD coprocessor architecture for sum of absolute differences and symmetric filtering (scalable MAC engine for image processing) ,(1999)
Walter Robert Steiner, Geometry processor for graphics display system ,(1986)
Harm Peter Hofstee, Martin Edward Hopkins, Michael Karl Gschwind, SIMD datapath coupled to scalar/vector/address/conditional data register file with selective subpath scalar processing mode ,(2001)
Jason Seung-Min Kim, Robert Quan, Vector floating point unit ,(2002)