Electrical computers: arithmetic processing and calculating – Electrical digital calculating computer – Particular function performed
Reexamination Certificate
2006-06-16
2010-11-16
Malzahn, David H (Department: 2193)
Electrical computers: arithmetic processing and calculating
Electrical digital calculating computer
Particular function performed
Reexamination Certificate
active
07836118
ABSTRACT:
The present invention enables efficient matrix multiplication operations on parallel processing devices. One embodiment is a method for mapping CTAs to result matrix tiles for matrix multiplication operations. Another embodiment is a second method for mapping CTAs to result tiles. Yet other embodiments are methods for mapping the individual threads of a CTA to the elements of a tile for result tile computations, source tile copy operations, and source tile copy and transpose operations. The present invention advantageously enables result matrix elements to be computed on a tile-by-tile basis using multiple CTAs executing concurrently on different streaming multiprocessors, enables source tiles to be copied to local memory to reduce the number accesses from the global memory when computing a result tile, and enables coalesced read operations from the global memory as well as write operations to the local memory without bank conflicts.
REFERENCES:
patent: 5659778 (1997-08-01), Gingold et al.
patent: 5708836 (1998-01-01), Wilkinson et al.
patent: 5892962 (1999-04-01), Cloutier
patent: 5963745 (1999-10-01), Collins et al.
patent: 6122274 (2000-09-01), Kumar
patent: 6282583 (2001-08-01), Pincus et al.
patent: 6532484 (2003-03-01), Kechriotis
patent: 6681388 (2004-01-01), Sato et al.
patent: 7483937 (2009-01-01), Nakanishi
patent: 7506134 (2009-03-01), Juffa et al.
patent: 7565391 (2009-07-01), Hu
patent: 7584342 (2009-09-01), Nordquist et al.
patent: 2003/0195938 (2003-10-01), Howard et al.
patent: 2004/0010533 (2004-01-01), Castrapel et al.
patent: 2006/0064452 (2006-03-01), Nakanishi
patent: 2007/0198986 (2007-08-01), Panziera
patent: 2007/0271325 (2007-11-01), Juffa et al.
Eggers, et al. “Simultaneous Multithreading: A Platform for Next-Generation Processors,”IEEE Micro, vol. 17, No. 5, pp. 12-19, Sep./Oct. 1997.
Gunnels, et al. “Analysis of a Class of Parallel Matrix Multiplication Algorithms,” IPPPS 98, pp. 1-23, http://www.cs.utexas.edu/users/plapack/papers/ipps98/ipps98.html.
Office Action. U.S. Appl. No. 11/454,411, dtd. Jan. 13, 2010.
Danilak Radoslav
Juffa Norbert
Malzahn David H
NVIDIA Corporation
Patterson & Sheridan LLP
LandOfFree
Hardware/software-based mapping of CTAs to matrix tiles for... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Hardware/software-based mapping of CTAs to matrix tiles for..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Hardware/software-based mapping of CTAs to matrix tiles for... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-4250617