J. R. Larus, "Parallelism in numeric and symbolic programs," Proceedings of the 3rd Annual Workshop on Programming Languages and Compilers for Parallel Computing, 1990. 158  Home/Search   Document Not in Database   Summary   Related Articles   Check

....semantics of the original sequential program. Dependences determines the available inherent parallelism of the loop. Some results from empirical studies of program characteristics have been recently published in the literature. Some of them cover aspects in the field of data dependence analysis [19, 15] analyzing subscript functions, distances associated to data dependences, kind of dependences, Other works report some measurements of the parallelism obtained by annotating a sequential program for monitoring and tracing its dynamic execution [11] It is worth knowing a priory the maximum ....

Larus J.R., "Parallelism in Numeric and Symbolic Programs", Proc. of the 3rd Workshop on Programming Languages and Compilers for Parallel Computing, Irvine-California USA, Pitman/MIT, August 1990.

.... researchers have shown that for numeric programs the most parallelism can be found at either the instruction level or the loop level [1] However for scalar programs, Larus suggests that there is not much loop level parallelism available because the loops tend to be small and have few iterations [5]. Therefore, it may be better to exploit fine grain parallelism for scalar applications. In this paper we analyze the effect of classical and superscalar optimizations on fine grain parallelism and speedup for scalar programs. Using code generated by the IMPACTI C compiler, we study the ....

J. R. Larus, "Parallelism in Numeric and Symbolic Programs", Proceedings of the 1990 Irvine Workshop, July 1990.

....and tracing systems have been described in the literature. MaxPar [6] is a CPS tool that measures the potential performance, parallelism, and behavior of optimistically parallelized codes given various architectural parameters. Other CPS tools have been used to measure program parallelism [7] and to perform dynamic dependence analysis [8] ETG tools instrument serial codes for trace generation on uniprocessor hosts [1] or parallel codes for trace generation on parallel hosts [9, 10] Examples of EDS systems include Tango [2] Proteus [11] and PEET [12] These systems use serial or ....

Larus, J., "Parallelism in Numeric and Symbolic Programs", Third Workshop on Programming Languages and Compilers for Parallel Computing, August 1990.

....assuming all data dependences are preserved. Kumar [15] used similar techniques and measured several larger programs with added assumptions that control dependences [12] are also preserved. Both of their works focused on fine grain parallelism at the operation or the statement level. Larus [16] measured the impact of cross iteration dependences on loops in numerical and non numerical applications using execution traces from C programs. Those earlier works addressed a lot of interesting issues and presented many interesting results. However, none of them truly attempted to answer all of ....

....causes I j and I i to differ should be marked as a DOACROSS loop. However, it is impractical to keep all of the iteration vectors for each data accessed because the length of the iteration vectors is not fixed and they use a large amount of storage. To detect cross iteration dependences, [16] creates an iteration descriptor for each loop iteration (see Figure 3) Each data accessed in the iteration also maintains some pointer information to keep track of the loop iterations. These descriptors form a tree like structure, with the descriptors of an inner loop pointing to the descriptor ....

[Article contains additional citation context not shown here]

J. Larus. Parallelism in numeric and symbolic programs. In 3rd Workshop on Programming Languages and Compilers for Parallel Computing, August 1990.

No context found.

J. R. Larus, "Parallelism in numeric and symbolic programs," Proceedings of the 3rd Annual Workshop on Programming Languages and Compilers for Parallel Computing, 1990. 158

No context found.

Larus, J., "Parallelism in Numeric and Symbolic Programs", Third Workshop on Programming Languages and Compilers for Parallel Computing, August 1990.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC