ST2GPU: An Energy-Efficient GPU Design with Spatio-Temporal Shared-Thread Speculative Adders
TimeTuesday, December 7th1:52pm - 2:15pm PST
Event Type
Research Manuscript
Virtual Programs
Presented In-Person
Cross-Layer Power Analysis and Low-Power Design
DescriptionModern GPUs employ thousands of cores, yielding higher performance but also higher power consumption. To meet performance targets while staying within a reasonable power budget, designers have to make these execution cores increasingly more power efficient. One way to increase their power efficiency is to employ power-efficient adders. We observe that consecutive arithmetic computations from the same code location are highly correlated and propose ST2GPU, a GPU architecture that uses history-based speculative adders that produce guaranteed correct results while saving 70% of the nominal adder power. We estimate that ST2GPU saves 21% chip energy with negligible performance and area overheads.