S. Chatterjee, J. F. Prins, and M. Simons. Expressing irregular computations in modern Fortran dialects. In Fourth Workshop on Languages, Compilers, and Run-Time Systems for Scalable Computers, Lecture Notes in Computer Science. Springer Verlag, 1998.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....and, in case of lazy languages, opportunities for unboxing are increased; 2) it improves load balancing and data distribution in a parallel implementation. The transformation was mainly studied for the second reason. There is plenty of literature discussing the merits of attening for parallelism [7, 33, 11, 34, 21, 22]. We recently extended attening to be suitable for typed functional languages such as Haskell and Standard ML [10] In the following, we merely want to summarise the main properties of the code resulting from attening as this will be the input to the transformations discussed in this paper. For ....

S. Chatterjee, J. F. Prins, and M. Simons. Expressing irregular computations in modern Fortran dialects. In Fourth Workshop on Languages, Compilers, and Run-Time Systems for Scalable Computers, Lecture Notes in Computer Science. Springer Verlag, 1998.

....library have on portability. Section 5 summarises and evaluates the benchmarks. Finally, Section 6 brie y reviews related work and concludes. 2 The Architecture Space Previous work addressed the implementation of nested data parallelism on a range of di erent machines, such as vector processors [3, 10], shared memory multiprocessors [9, 10] and distributed memory machines [6, 14] However, the implementations, while being based on attening, used di erent optimisation techniques and enjoyed various levels of success. We are investigating to which extend we can target the full range of ....

....summarises and evaluates the benchmarks. Finally, Section 6 brie y reviews related work and concludes. 2 The Architecture Space Previous work addressed the implementation of nested data parallelism on a range of di erent machines, such as vector processors [3, 10] shared memory multiprocessors [9, 10], and distributed memory machines [6, 14] However, the implementations, while being based on attening, used di erent optimisation techniques and enjoyed various levels of success. We are investigating to which extend we can target the full range of architectures with a uniform compilation system ....

[Article contains additional citation context not shown here]

S. Chatterjee, Jan F. Prins, and M. Simons. Expressing irregular computations in modern Fortran dialects. In Fourth Workshop on Languages, Compilers, and Run-Time Systems for Scalable Computers, Lecture Notes in Computer Science. Springer Verlag, 1998.

.... transformation for both program development and compilation, and simplifies the definition of cost models [23, 1] 1 Fortran s FORALL loops are index based, which complicates the implementation in comparison to, e.g. Nesl s apply to each, but still, many of the basic principles are the same [8]. The aggregate data structures of collection oriented languages are often implemented by a library whose routines are in functionality close to the basic operations of the source language. Consequently, the library is both high level and complex, but carefully optimized for the targeted ....

....most machine dependencies of the implementation. We call this the library approach for Nesl s implementation, see [4] Fortran 90 95 s collection oriented operations are usually directly realized by a library, and even, FORALL loops can benefit from a collection oriented implementation [8]. The library approach, while simplifying the compiler, blocks optimizations by the rigid interface imposed by the library most importantly, it inhibits the use of standard optimizations for processor local code and hinders the efficient use of the memory hierarchy, i.e. optimizations for the ....

S. Chatterjee, Jan F. Prins, and M. Simons. Expressing irregular computations in modern Fortran dialects. In Fourth Workshop on Languages, Compilers, and Run-Time Systems for Scalable Computers, Lecture Notes in Computer Science. Springer Verlag, 1998.

.... is the attening transformation, a program transformation that maps nested to at parallelism and that was originally introduced by Blelloch and Sabot [5] It leads to ecient implementations on a range of high performance architectures in both the purely functional [7] and the imperative case [9]. Nevertheless, imperative languages face a serious problem here: Flattening is hard to automate in the presence of pointers and side e ects [9] Even special purpose extensions of imperative languages require serious restrictions [22] This may ultimately con ne imperative code to thread based ....

.... and Sabot [5] It leads to ecient implementations on a range of high performance architectures in both the purely functional [7] and the imperative case [9] Nevertheless, imperative languages face a serious problem here: Flattening is hard to automate in the presence of pointers and side e ects [9]. Even special purpose extensions of imperative languages require serious restrictions [22] This may ultimately con ne imperative code to thread based implementations of nested parallelism [2, 20] which currently seem to require sharedmemory architectures and certainly are not suitable for ....

S. Chatterjee, J. F. Prins, and M. Simons. Expressing irregular computations in modern Fortran dialects. In Fourth Workshop on Languages, Compilers, and Run-Time Systems for Scalable Computers, Lecture Notes in Computer Science. Springer Verlag, 1998.

....array. Parallelism in HPF programs comes through the use of a forall construct and special parallel operations on arrays, such as array permutation and reduction. The language is not oriented towards nested data parallelism, although pointer types can be used to support nested collections [26], and certain routines in the standard HPF library can be used to support nested data parallel operations. Titanium [30] is a dialect of Java designed for SPMD parallel programming. While Titanium uses Java as a base, it adds many extensions to the language, such as immutable classes, parallel ....

J. Prins, S. Chatterjee, and M. Simons. Expressing irregular computations in modern Fortran dialects. In D. O'Hallaron, editor, Languages, Compilers, and Run-Time Systems for Scalable Computers, pages 1--16. springer, 1998.

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