• Publications
  • Influence
Exploiting locality for irregular scientific codes
  • Hwansoo Han, C. Tseng
  • Computer Science
  • IEEE Transactions on Parallel and Distributed…
  • 1 July 2006
TLDR
In this paper, we present two new locality improving techniques for irregular scientific codes. Expand
  • 88
  • 8
  • PDF
Efficient compiler and run-time support for parallel irregular reductions
TLDR
We develop L ocal W rite , a new technique which partitions irregular reductions so that each processor computes values only for locally assigned data, eliminating the need for buffers or synchronized writes. Expand
  • 31
  • 4
Improving Locality for Adaptive Irregular Scientific Codes
TLDR
We present a new graph partitioning algorithm based on hierarchical clustering, and show how to tune it to improve locality with low overhead. Expand
  • 41
  • 3
  • PDF
Efficient SIMD code generation for irregular kernels
TLDR
In this work, we propose a method to generate efficient SIMD code for loops containing indirected memory references in loops with array indirection. Expand
  • 47
  • 3
Improving compiler and run-time support for adaptive irregular codes
  • Hwansoo Han, C. Tseng
  • Computer Science
  • Proceedings. International Conference on…
  • 12 October 1998
TLDR
We introduce LOCALWRITE, a new compiler and runtime technique for parallelizing irregular reductions based on the owner-computes rule which eliminates the need for buffers or synchronized writes but may replicate computation. Expand
  • 38
  • 3
  • PDF
A Comparison of Locality Transformations for Irregular Codes
TLDR
We develop GPART,a new partitioning algorithm based on hierarchical graph partitioning, which produces partitions which almost match the performance of the best partitioning strategies, but with much lower overhead. Expand
  • 58
  • 2
  • PDF
Improving Compiler and Run-Time Support for Irregular Reductions Using Local Writes
TLDR
We introduce LOCALWRITE, a new compiler and run-time parallelization technique for parallelizing irregular reductions on distributed-memory multiprocessors based on the owner-computes rule. Expand
  • 35
  • 2
  • PDF
A comparison of parallelization techniques for irregular reductions
  • Hwansoo Han, C. Tseng
  • Computer Science
  • Proceedings 15th International Parallel and…
  • 23 April 2001
TLDR
A large class of scientific applications are comprised of irregular reductions on large data sets. Expand
  • 22
  • 2
  • PDF
Transparent Method Offloading for Slim Execution
TLDR
In this paper, we propose a transparent method offloading for slim execution, a novel approach to transparently relieve mobile devices of resource constraints. Expand
  • 22
  • 2
Access pattern based stream buffer management scheme for portable media players
TLDR
We propose a memory management scheme based on the notion of stream, a set of heap-allocated buffers with the same backtrace of function calls. Expand
  • 6
  • 2
  • PDF