English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Performance optimization can be a daunting task especially as the hardwarearchitecture becomes more and more complex. This paper takes a kernel from theMaterials Science code BerkeleyGW, and demonstrates a few performance analysisand optimization techniques. Despite challenges such as high register usage,low occupancy, complex data access patterns, and the existence of severallong-latency instructions, we have achieved 3.7 TFLOP/s of double-precisionperformance on an NVIDIA V100 GPU, with 8 optimization steps. This is 55% ofthe theoretical peak, 6.7 TFLOP/s, at nominal frequency 1312 MHz, and 70% ofthe more customized peak based on our 58% FMA ratio, 5.3 TFLOP/s. An array oftechniques used to analyze this OpenACC kernel and optimize its performance areshown, including the use of hierarchical Roofline performance model and theperformance tool Nsight Compute. This kernel exhibits computationalcharacteristics that are commonly seen in many high-performance computing (HPC)applications, and are expected to be very helpful to a general audience of HPCdevelopers and computational scientists, as they pursue more performance onNVIDIA GPUs.

8 Steps to 3.7 TFLOP/s on NVIDIA V100 GPU: Roofline Analysis and Other Tricks