DROPS

Document

Dark Silicon: From Embedded to HPC Systems (Dagstuhl Seminar 16052)

Authors: Hans Michael Gerndt, Michael Glaß, Sri Parameswaran, and Barry L. Rountree

Published in: Dagstuhl Reports, Volume 6, Issue 1 (2016)

Abstract

Semiconductor industry is hitting the utilization wall and puts focus on parallel and heterogeneous many-core architectures. While continuous technological scaling enables the high integration of 100s-1000s of cores and, thus, enormous processing capabilities, the resulting power consumption per area (the power density) increases in an unsustainable way. With this density, the problem of Dark Silicon will become prevalent in future technology nodes: It will be infeasible to operate all on-chip components at full performance at the same time due to the thermal constraints (peak temperature, spatial and temporal thermal gradients etc.). However, this is not only an emerging threat for SoC and MPSoC designers, HPC faces a similar problem as well: The power supplied by the energy companies as well as the cooling capacity does not allow to run the entire machine at highest performance anymore. The goal of Dagstuhl Seminar 16052 "Dark Silicon: From Embedded to HPC Systems" was to increase the awareness of the research communities of those similarities and to work and explore common solutions based on more flexible thermal/power/resource management techniques both for runtime, design time as well as hybrid solutions.

Cite as

Hans Michael Gerndt, Michael Glaß, Sri Parameswaran, and Barry L. Rountree. Dark Silicon: From Embedded to HPC Systems (Dagstuhl Seminar 16052). In Dagstuhl Reports, Volume 6, Issue 1, pp. 224-244, Schloss Dagstuhl – Leibniz-Zentrum für Informatik (2016)

Copy BibTex To Clipboard

@Article{gerndt_et_al:DagRep.6.1.224,
  author =	{Gerndt, Hans Michael and Gla{\ss}, Michael and Parameswaran, Sri and Rountree, Barry L.},
  title =	{{Dark Silicon: From Embedded to HPC Systems (Dagstuhl Seminar 16052)}},
  pages =	{224--244},
  journal =	{Dagstuhl Reports},
  ISSN =	{2192-5283},
  year =	{2016},
  volume =	{6},
  number =	{1},
  editor =	{Gerndt, Hans Michael and Gla{\ss}, Michael and Parameswaran, Sri and Rountree, Barry L.},
  publisher =	{Schloss Dagstuhl -- Leibniz-Zentrum f{\"u}r Informatik},
  address =	{Dagstuhl, Germany},
  URL =		{https://drops.dagstuhl.de/entities/document/10.4230/DagRep.6.1.224},
  URN =		{urn:nbn:de:0030-drops-58198},
  doi =		{10.4230/DagRep.6.1.224},
  annote =	{Keywords: dark silicon, embedded, hpc, parallel computing, performance analysis and tuning, power density, power modelling, programming tools, resource manageme}
}

Document

DOI: 10.4230/DagRep.5.8.1

Power-Bounded HPC Performance Optimization (Dagstuhl Perspectives Workshop 15342)

Authors: Dieter Kranzlmüller and Barry L. Rountree

Published in: Dagstuhl Reports, Volume 5, Issue 8 (2016)

Abstract

This report documents the program and the outcomes of Dagstuhl Perspectives Workshop 15342 "Power-Bounded HPC Performance Optimization". The workshop consists of two parts. In part one, our international panel of experts in facilities, schedulers, runtime systems, operating systems, processor architectures and applications provided thought-provoking and details insights into open problems in each of their fields with respect to the workshop topic. These problems must be resolved in order to achieve a useful power-constrainted exascale system, which operates at the highest performance within a given power bound. In part two, the participants split up in three groups, trying to address certain specific subtopics as identified during the expert plenaries. These subtopics have been discussed in more detail, followed by plenary sessions to compare and synthesize the findings into an overall picture. As a result, the workshop identified three major problems, which need to be solved on the way to power-bounded HPC performance optimization.

Cite as

Dieter Kranzlmüller and Barry L. Rountree. Power-Bounded HPC Performance Optimization (Dagstuhl Perspectives Workshop 15342). In Dagstuhl Reports, Volume 5, Issue 8, pp. 1-8, Schloss Dagstuhl – Leibniz-Zentrum für Informatik (2016)

Copy BibTex To Clipboard

@Article{kranzlmuller_et_al:DagRep.5.8.1,
  author =	{Kranzlm\"{u}ller, Dieter and Rountree, Barry L.},
  title =	{{Power-Bounded HPC Performance Optimization (Dagstuhl Perspectives Workshop 15342)}},
  pages =	{1--8},
  journal =	{Dagstuhl Reports},
  ISSN =	{2192-5283},
  year =	{2016},
  volume =	{5},
  number =	{8},
  editor =	{Kranzlm\"{u}ller, Dieter and Rountree, Barry L.},
  publisher =	{Schloss Dagstuhl -- Leibniz-Zentrum f{\"u}r Informatik},
  address =	{Dagstuhl, Germany},
  URL =		{https://drops.dagstuhl.de/entities/document/10.4230/DagRep.5.8.1},
  URN =		{urn:nbn:de:0030-drops-56757},
  doi =		{10.4230/DagRep.5.8.1},
  annote =	{Keywords: Exascale computing, Energy efficiency, Power awareness, Scalability}
}

Search Results

Documents authored by Rountree, Barry L.

Dark Silicon: From Embedded to HPC Systems (Dagstuhl Seminar 16052)

Abstract

Cite as

Power-Bounded HPC Performance Optimization (Dagstuhl Perspectives Workshop 15342)

Abstract

Cite as

Thanks for your feedback!

Could not send message