OWL: Cooperative thread array aware scheduling techniques for improving GPGPU performance

Keyphrases

• Warp Scheduling 100%
• GPGPU 100%
• Scheduling Techniques 100%
• Cooperative Thread Arrays 100%
• Scheduling Policy 60%
• Performance Improvement 40%
• Long Memory 40%
• Memory Latency 40%
• Bank-level Parallelism 40%
• Low Energy 20%
• Robin 20%
• Two-level 20%
• Performance Enhancement 20%
• Lost Opportunity 20%
• Agnostic 20%
• Prefetching 20%
• Programming Model 20%
• Energy Budget 20%
• Thread-level Parallelism 20%
• Scheduling Decisions 20%
• Memory-intensive Applications 20%
• Thread Block 20%
• Cache Contention 20%
• CUD A 20%
• GPGPU Architecture 20%
• Locality-aware 20%
• Core Performance 20%
• Hardware Resources 20%
• Latency Hiding 20%
• OpenCL 20%

Computer Science

• Scheduling Policy 100%
• Memory Latency 66%
• Level Parallelism 66%
• Performance Improvement 33%
• prefetching 33%
• Average Performance 33%
• Programming Model 33%
• Intensive Application 33%
• Proposed Mechanism 33%
• Enhance Performance 33%
• Thread Level Parallelism 33%
• Hardware Resource 33%