Partial Flattening: A Compilation Technique for Irregular Nested Parallelism on GPGPUs

Keyphrases

• GPGPU 100%
• Nested Parallelism 100%
• Compilation Techniques 100%
• Memory Consumption 66%
• Depth-first 66%
• Task Allocation 66%
• Thread Hierarchy 66%
• First-order 33%
• Two-level 33%
• Computing Power 33%
• Hardware Support 33%
• Additional Task 33%
• CUDA Kernel 33%
• Task Stealing 33%
• Parallel Loops 33%
• Thread Block 33%
• CUD A 33%
• Parallel Tasks 33%
• CUDA Dynamic Parallelism 33%
• Distribution Scheme 33%
• Code Transformation 33%
• C Program 33%
• Execution Order 33%
• Recursive Call 33%
• Quick Sort 33%
• Bron-Kerbosch Algorithm 33%

Computer Science

• Nested Parallelism 100%
• Graphics Processing Unit 66%
• Hardware Support 33%
• Computing Power 33%
• C Program 33%
• Execution Order 33%
• Quicksort 33%
• Dynamic Parallelism 33%
• Code Transformation 33%