<p>04/06/2019<br />
Dear users,</p>
<p>we encourage you to partecipate to an upcoming Workshop about<br />
Energy-efficient Computing on Parallel</p>
<p>Architectures, scheduled for 10-13September 2019 in Prague.</p>
<p> </p>
<p>The power requirements of large HPC facilities are becoming unsustainable for<br />
both were requirements</p>
<p>of large HPC facilities are becoming unsustainable for both technical and<br />
economic reasons. A significant</p>
<p>fraction of the total cost of ownership of HPC installations available<br />
nowadays is already driven by the</p>
<p>electricity bill, and the idea of charging users for the energy consumed by<br />
their applications is spreading.</p>
<p>This requires to find disruptive, smart and effective solutions, both at<br />
level of hardware and software to</p>
<p>maximize energy-efficiency and computing throughput of HPC systems within a<br />
given power envelop.</p>
<p>Several recently developed processors, such multi- and many-cores CPUs,<br />
GPUs, FPGAs and SoCs</p>
<p>(System on Chip) are engineered including features to improve computing<br />
energy-efficiency, e.g. higher</p>
<p>level of parallelism; dynamic clock frequency adaptation; capability of<br />
switching off idle cores, etc.</p>
<p>However, programming complexity and code portability can be affected by<br />
these choices, and computing</p>
<p>efficiency (as well as energy-efficiency) can be disrupted if<br />
applications are not properly coded to exploit</p>
<p>(or take into account) all of the hardware features, in the worst scenario,<br />
possibly increasing the</p>
<p>time-to-solution and the energy-to-solution. In addition, the heterogeneity<br />
of modern compute nodes and</p>
<p>systems, combined with the dynamic behavior of modern applications, require<br />
educated resource allocation</p>
<p>and resource management techniques to achieve energy efficiency.</p>
<p> </p>
<p>This workshop aims to strongly encourage the exchange of experiences and<br />
knowledge in novel strategies</p>
<p>to exploit, monitor, analyze, and optimize the energy-efficiency of recent<br />
computing systems. We focus on</p>
<p>new trends including, hardware and software tools, scheduling and resource<br />
management techniques but</p>
<p>also algorithm-design and techniques in general, able to minimize the<br />
energy-to-solution of workloads, and</p>
<p>to reduce the energy required to operate computing systems.</p>
<p> </p>
<p>Contribution should be submitted through the submission server ParCo2019<br />
ConfTool</p>
<p>(<a href="https://www.conftool.org/parco2019">https://www.conftool.org/parco2019</a>) by July 15th 2019.</p>
<p> </p>
<p> IMPORTANT DATES</p>
<p>July 15th 2019: deadline for extended abstract submission</p>
<p>July 30th 2019: notification of acceptance</p>
<p>September 13th: mini-symposium</p>
<p>October 31st : full paper submission (ten (10) pages length)</p>
<p> </p>
<p>HPC User Support @ Cineca</p>