2015 Program (text)

Talks:
1) Hash Map Linearization, Dibakar Gope
2) Spatiotemporal Adaptivity in Tiled Architectures, Gokul Ravi
3) Self-learning Map-Seeking Circuits, Rohit Shukla
4) Transparent Fuzzy Barriers for GPUs, Joel Hestness
5) Accelerometer: A Roofline Model for Hardware Accelerators, Mohammed Shoaib
6) Synchronization Using Remote-Scope Promotion, Marc Orr
7) Revisiting Virtual L1 Caches, Hongil Yoon
8) Border Control: Sandboxing Accelerators, Lena Olson
9) Redundant Memory Mappings for Fast Access to Large Memories, Jayneel Gandhi
10) A Framework and Analysis of Behavior Specialized Accelerators, Tony Nowatzi
11) Rapid, Human-Friendly, and Accurate Prediction of GPU Speedup
Using Single-Bit Values of Program Features, Newsha Ardalani

Posters:
ChronoBar: Energy-Efficient Temporal Operand Collection, Michael Mishkin
Checkpointing Methods for TAGE Predictors, David Schlais
GPU Computing Pipeline Inefficiencies and Optimization Opportunities
in Heterogeneous CPU-GPU Processors, Joel Hestness
Evolution of Crypto Accelerators in SPARC Architecture : A Case Study,
Mohammed Shoaib
Pruning Design Space for Cache Hierarchies, Nilay Vaish
Energy-Proportional Computing: A New Definition, Rathijit Sen
Revisiting Virtual L1 Caches, Hongil Yoon
Exploiting Ordered Parallelism in Fine-Grained Task-Parallel Programs,
Aditya Venkataraman
Border Control: Sandboxing Accelerators, Lena Olson
Redundant Memory Mappings for Fast Access to Large Memories, Jayneel Gandhi
Agile Paging for Efficient Virtualized Page Walks, Jayneel Gandhi
Toward GPUs being mainstream in analytic processing, Jason Power
Seeking to Program Accelerators Like Homogeneous Shared Memory, Chris
Feilbach & Adam Sperling
Memory Processing Units, Vijay Thiruvengadam
MIAOW: Open source GPGPU, Vinay Gangadhar
Self-Learning GPU Warp Scheduler, Newsha Ardalani