25 citations · 30 across the 2 of their papers we have counts for
3 papers · 1 filter
Optimizing GPU Cache Policies for MI Workloads
Johnathan Alsop, Matthew D. Sinclair, Srikant Bharadwaj +8
In recent years, machine intelligence (MI) applications have emerged as a major driver for the computing industry. Optimizing these workloads is important but complicated. As memor…
Holistic Management of the GPGPU Memory Hierarchy to Manage Warp-level Latency Tolerance
Rachata Ausavarungnirun, Saugata Ghose, Onur Kayıran +4
In a modern GPU architecture, all threads within a warp execute the same instruction in lockstep. For a memory instruction, this can lead to memory divergence: the memory requests…
CODA: Enabling Co-location of Computation and Data for Near-Data Processing
Hyojong Kim, Ramyad Hadidi, Lifeng Nai +5
Recent studies have demonstrated that near-data processing (NDP) is an effective technique for improving performance and energy efficiency of data-intensive workloads. However, lev…