Thomas J. Sheffler and Siddhartha Chatterjee. An object-oriented approach to nested data parallelism. In Proceedings of the Fifth Symposium on the Frontiers of Massively Parallel Computation. IEEE, February 1995.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....of Palmer et al. [122] described an approach to realise work ecient data parallel indexing based on bounding the amount of replication and randomly dispersing concurrent reads. Also, nested data parallelism has been integrated into an object oriented language in the work of Shef er and Chatterjee [149]. The strength of the method lies in the compilation strategy, which translates nested data parallel C into ordinary C . The language V [32] integrates nested data parallelism into imperative languages using the example of C. Because of the imperative context, new problems like pointer lifting ....

T. Sheer and S. Chatterjee. An Object-Oriented Approach to Nested Data Parallelism. In Proceedings of the 5th Symposium on the Frontiers of Massively Parallel Computation, pages 203-210, McLean, VA, USA, February 1995.

....besides NESL. Proteus is a high level architectureindependent programming language designed for rapid application prototyping [21] V extends nested data parallelism to the imperative programming model of C [13] and the language developed by Sheffler and Chatterjee adds nested constructs to C [25], using a flattening technique to transform nested data parallel code into ordinary C . However, all three of these languages use CVL as a base implementation layer, and are therefore limited by its performance problems on current distributed memory multiprocessors. Adl, a functional language ....

T. J. Sheffler and S. Chatterjee. An object-oriented approach to nested data parallelism. In Proceedings of Symposium on The Frontiers of Massively Parallel Computation, Feb. 1995.

.... in Fortran If we consider expressing nested data parallelism in standard imperative programming languages, we find that they either lack a data parallel control construct (C, C ) or else lack a nested collection data type (Fortran) A data parallel control construct can be added to C [11] or C [30], but the pervasive pointer semantics of these languages complicate its meaning. There is also incomplete agreement about the form of parallelism should take in these languages. The FORALL construct, originated in HPF [16] and later added into Fortran 95, specifies data parallel evaluation of ....

....evaluation of the inner products for all rows. Nested parallelism is a consequence of the use of parallel operations such as sum and elementwise multiplication, projection, and indexing. Discussion Earlier experiments with nested data parallelism in imperative languages include V [11] Amelia [30], and F90V [1] For the first two of these languages the issues of side effects in the underlying notation (C and C, respectively) were problematic in the potential introduction of interference between parallel iterations, and the efforts were abandoned. Fortran finesses this problem by ....

[Article contains additional citation context not shown here]

T. J. Sheffler and S. Chatterjee. An object-oriented approach to nested data parallelism. In Proceedings of the Fifth Symposium on the Frontiers of Massively Parallel Computation, pages 203--210, McLean, VA, Feb. 1995.

.... in Fortran If we consider expressing nested data parallelism in standard imperative programming languages, we find that they either lack a data parallel control construct (C, C ) or else lack a nested collection data type (Fortran) A data parallel control construct can be added to C [11]orC [30], but the pervasive pointer semantics of these languages complicate its meaning. There is also incomplete agreement about the form of parallelism should take in these languages. The ###### construct, originated in HPF [16] and later added into Fortran 95, specifies data parallel evaluation of ....

....evaluation of the inner products for all rows. Nested parallelism is a consequence of the use of parallel operations such as sum and elementwise multiplication, projection, and indexing. Discussion Earlier experiments with nested data parallelism in imperative languages include V [11] Amelia [30], and F90V [1] For the first two of these languages the issues of side effects in the underlying notation (C and C, respectively) were problematic in the potential introduction of interference between parallel iterations, and the efforts were abandoned. Fortran finesses this problem by ....

[Article contains additional citation context not shown here]

T. J. Sheffler and S. Chatterjee. An object-oriented approach to nested data parallelism. In Proceedings of the Fifth Symposium on the Frontiers of Massively Parallel Computation, pages 203--210, McLean, VA, Feb. 1995.

.... we consider the expression of nested data parallelism in standard imperative programming languages, we find that they either lack a data parallel control construct (C, C ) or else lack a nested collection data type (Fortran) A nested data parallel control construct can be added to C [13] or C [37], but the pervasive pointer semantics of these languages complicate its meaning. There is also incomplete agreement about the form parallelism should take in these languages. The FORALL construct, originated in HPF [21] and later added into Fortran 95 [1, 24] specifies data parallel evaluation ....

....3. Fortran 95 definition of a quad tree for hierarchical force calculation in 2D (top) and its use in a hierarchical force calculation (bottom) 8 3. 3 Nested data parallelism in an imperative setting Earlier experiments with nested data parallelism in imperative languages include V [13] Amelia [37], and F90V [2] For the first two of these languages the issues of side effects in the underlying notation (C and C, respectively) were problematic in the potential introduction of interference between parallel iterations, and the efforts were abandoned. Fortran 95 finesses this problem by ....

[Article contains additional citation context not shown here]

T. J. Sheffler and S. Chatterjee. An object-oriented approach to nested data parallelism. In Proceedings of the Fifth Symposium on the Frontiers of Massively Parallel Computation, pages 203--210, McLean, VA, Feb. 1995. 3, 3.3, 6

....functions that work on any type of sequence. Because Nesl is purely functional, it suffers the usual problems when updating subsets of aggregate variables, and it makes it impossible for users to manage memory use. Also, Nesl s runtime system is incompatible with other languages. The Amelia [28] project introduces nested data parallelism into C . We use many of the same techniques as the Amelia project, but do not suffer the problems with unrestricted pointer semantics of C . The pC [11] language defines a new programming model called the distributed collection model. This model is ....

T. J. Sheffler and S. Chatterjee. An object-oriented approach to nested data parallelism. In Proceedings of the Fifth Symposium on the Frontiers of Massively Parallel Computation, pages 203--210, McLean, VA, Feb. 1995.

No context found.

Thomas J. Sheffler and Siddhartha Chatterjee. An object-oriented approach to nested data parallelism. In Proceedings of the Fifth Symposium on the Frontiers of Massively Parallel Computation. IEEE, February 1995.

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