Performance, power consumption and thermal behavioral evaluation of the DGX-2 platform

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F61989100%3A27740%2F20%3A10244990" target="_blank" >RIV/61989100:27740/20:10244990 - isvavai.cz</a>

Výsledek na webu

<a href="http://ebooks.iospress.nl/volumearticle/53970" target="_blank" >http://ebooks.iospress.nl/volumearticle/53970</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.3233/APC200091" target="_blank" >10.3233/APC200091</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Performance, power consumption and thermal behavioral evaluation of the DGX-2 platform

Popis výsledku v původním jazyce

In this paper, we evaluate the performance, power consumption and its variation and also thermal behavior of the DGX-2 server from Nvidia. We present a development of specialized synthetic benchmarks to measure raw performance of GPUs for single, double, half precision and also Tensor Core units. With these benchmarks, we were able to reach peak performance and verify the specification provided by Nvidia. We achieved 130.79 TFLOPS peak performance in half-precision on Tensor Cores. We also measured the thermal stability of the DGX-2 system. It can hold its peak performance when all 16 GPUs are fully loaded except Tensor Core workload, when thermal throttling occurred with with up to 1% performance penalty. During single-precision workload we observed 23% variation of the power consummation of individual GPUs installed in the system. Finally, we have evaluated the behavior of the Tesla V100-SXM3 chip under the DVFS tuning. Running at optimal frequency, the compute bound workload can save up to 39% energy while the run-time increases by 51%. More importantly, memory bound workload can save up to 31% with 2% throughput penalty and during the communication over NVLink one can save up to 26% energy with no penalty. (C) 2020 The authors and IOS Press.

Název v anglickém jazyce

Performance, power consumption and thermal behavioral evaluation of the DGX-2 platform

Popis výsledku anglicky

In this paper, we evaluate the performance, power consumption and its variation and also thermal behavior of the DGX-2 server from Nvidia. We present a development of specialized synthetic benchmarks to measure raw performance of GPUs for single, double, half precision and also Tensor Core units. With these benchmarks, we were able to reach peak performance and verify the specification provided by Nvidia. We achieved 130.79 TFLOPS peak performance in half-precision on Tensor Cores. We also measured the thermal stability of the DGX-2 system. It can hold its peak performance when all 16 GPUs are fully loaded except Tensor Core workload, when thermal throttling occurred with with up to 1% performance penalty. During single-precision workload we observed 23% variation of the power consummation of individual GPUs installed in the system. Finally, we have evaluated the behavior of the Tesla V100-SXM3 chip under the DVFS tuning. Running at optimal frequency, the compute bound workload can save up to 39% energy while the run-time increases by 51%. More importantly, memory bound workload can save up to 31% with 2% throughput penalty and during the communication over NVLink one can save up to 26% energy with no penalty. (C) 2020 The authors and IOS Press.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

<a href="/cs/project/LM2018140" target="_blank" >LM2018140: e-Infrastruktura CZ</a><br>

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)<br>S - Specificky vyzkum na vysokych skolach

Rok uplatnění

2020

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

Advances in Parallel Computing. Volume 36

ISBN

978-1-64368-070-5

ISSN

0927-5452

e-ISSN

1879-808X

Počet stran výsledku

10

Strana od-do

614-623

Název nakladatele

IOS Press

Místo vydání

Amsterdam

Místo konání akce

Praha

Datum konání akce

10. 9. 2019

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

—