Alessandro Morari and Roberto Gioiosa from Pacific Northwest National Laboratory won the best paper award at the recent International Parallel & Distributed Process Symposium in China. Their paper “Evaluating the Impact of TLB Misses on Future HPC Systems” presents a quantitative analysis of the effect of translation lookaside buffer misses on current and future parallel applications at scale.
TLB misses have been considered an important source of system overhead and one of the causes that limit scalability on large supercomputers. This assumption lead to HPC lightweight kernel designs that usually statically map page table entries to TLB entries and do not take TLB misses. While this approach worked for petascale clusters, programming and debugging exascale applications composed of billions of threads is not a trivial task and users have started to explore novel programming models and tools, which require a richer system software support.
Morari and Gioiosa are staff members in PNNL’s Computational Sciences & Mathematics division where they focus on high performance computing research. Download the paper (PDF).