Embedded Files
2024 Liquid Cooling API Telemetry and Controls Webinar
Two very important software API standards (Application Software Interface) are currently expanding their domain to include liquid cooling. These are the PowerAPI and Redfish. Ryan Grant and Jeff Autor, respectively, will introduce these standards and the current effort to expand them to liquid cooling. There will be ample time for discussion as well as Q&A.
2020 HPC Energy Efficiency in an Exascale Era Webinar
There are four facets of energy efficiency; systems, infrastructure, software and the environment; These are underscored with an operationally focused data measurement, collection and analysis system. This webinar touches on all four facets with insights from efforts at Lawrence Berkeley National Laboratory's supercomputing center.
2019 Water-Cooled Servers Common Designs, Components, and Processes Webinar
With more water-cooled IT products arriving in the marketplace, ASHRAE TC 9.9 felt the need to outline some of the common processes, parts, and materials for focus in use for future water-cooled designs. This white paper is an attempt to provide and make available those items that could be classified as common. The material published in this white paper compliments the materials published in Liquid Cooling Guidelines for Datacom Equipment Centers, second edition (ASHRAE 2014).
2019 Multi Source Cooling Control Algorithm Webinar
This webinar describes how the automation team at ENI Green Data Center (GDC) solved the problem of controlling the temperature inside a data room that can be air cooled both with direct free-cooling and with water-based chillers and heat exchangers using a sliding mode control design. They created an algorithm which provided: 1. a directional, incremental “walking ” function to seek out and test alternative solutions, biased toward a preferred solution, 2. stabilization on an optimized solution, 3. a destabilization function to retest the solution periodically without allowing jitter, and 4. a method of changing preference to autonomously re-optimize for another configuration.
2018 Analysis of Contracts between Supercomputing Centers and Electricity Service Providers Webinar
Increases in peak electricity demands and the growing use of renewable energy — with associated intermittency and variable output — present new challenges to electricity service providers (ESPs). ESPs employ demand charges, variable tariffs and demand response (DR) programs to influence the consumption behavior of consumers to partially mitigate these challenges. Due to their high load and potential flexibility, supercomputing centers (SCs) are increasingly gaining importance in the grid. This webinar presents a qualitative study of service contracts between ESPs and SCs in the United States and Europe. From this we extract a contract typology used to understand how, and to what extent, variable tariffs, DR programs, and demand charges are imposed on SCs. Further, we highlight the actions taken by SCs in response to these contractual elements. Finally, we present perspectives on grid integration of SCs to enhance their collaboration with their ESPs to benefit supply stability and resilience.
2017 Energy Efficiency Considerations and HPC Procurement Webinar
The predominant goal for procurement of HPC systems is to identify the optimal solution to both technical and financial targets that maximizes the contribution of that system to the organization’s mission. Beyond the acquisition cost of the system, it is also important to consider the total costs of ownership, including the improvements necessary to host the system, the infrastructure that supports its operation, and the significant operational costs associated with that new HPC system. In this panel, HPC leaders will discuss and debate key procurement requirements and lessons learned that can contribute to greater energy efficiency and reduced operational costs. The document by the Energy Efficient High Performance Computing Working Group (EE HPC WG) “Energy Efficiency Considerations for HPC Procurement Documents: 2014” serves as a backdrop to the discussions and its aim and scope will be restated at the outset of the webinar. Following this, representatives from Sandia, LLNL, and TU Dresden will give a brief presentation of their experiences with the integration of energy efficiency considerations in recent procurements. To conclude the presentations representatives from Cray and IBM provided their feedback and take on the EE HPC WG document.
Ladina Gilly from the Swiss Supercomputing Center will moderate the panel. Panelists include James H. Laros III from Sandia National Laboratory, Anna Maria Bailey from Lawrence Livermore National Laboratory, Daniel Hackenberg from TU Dresden, Todd Takken from IBM and Steve Martin from Cray Inc.
2016 Data centre design standards and best practices for HPC centres
To what extent are existing data centre design standards known and applied within public research HPC data centre? Which design issues, common to public research HPC data centres, do the standards not cover? How do public research HPC data centres define design criteria where there requirements are not covered by existing standards? What approaches to future-proofing do public research HPC data centres apply? Is there sufficient similarity in the applied approaches to allow the definition of a standard for the design of public research HPC data centres? These are some of the questions that will be addressed by Ladina Gilly in this webinar. Ladina has had personal experience of these challenges from when she was planning the new Swiss Supercomputing Center (CSCS) data centre. She will look at HPC requirements v. enterprise design standards; the life cycles of buildings v. HPC systems; fast changing technology developments and optimizing CAPEX and OPEX.
The Sandia Power API is a proposed “de-facto” standard power Application Programming Interface (API) that endeavors to cover the entire software space, from generic hardware interfaces to the input from the computer facility manager. This Power API provides multiple levels of abstractions to satisfy the requirements of multiple types of users. It creates a reference implementation comprising a subset of the overall API functionality. The ultimate goal, however, is that vendors of the hardware and software components provide their own implementations. It is likely that some portion of these functions have already been written by vendors, but with slightly different calling arguments. For portability sake, it is hoped that the specific implementations can be melded to this proposed API.
This webinar introduces GEO: a new scalable, extensible, open source runtime for global energy optimization in HPC systems. GEO is a production-grade performance and power management framework with development, testing, and validation supported by Intel and collaborations across the Open Source Community. GEO provides built-in features ranging from static management of power policy for each individual compute node, to dynamic coordination of power policy and performance across all of the compute nodes hosting one MPI job on a portion of a distributed computing system. Through a plug-in architecture, GEO can be extended to support new control algorithms and new hardware power management features. The dynamic coordination is implemented as a hierarchical control system for scalable communication and decentralized control. The hierarchical control system can optimize for various objective functions including maximizing global application performance within a power bound.
Redfish is part of a broader effort to create and publish an open industry standard specification and schema that meets the expectations of end users for simple, modern and secure management of scalable platform hardware. Redfish will specify retrieval of “IPMI class” data including temperature sensors, fans, power supplies, power consumption and power thresholds. It will allow for performing common actions like setting power thresholds. It will interface with the building management console with viewing and configuring networks settings as well as managing local user accounts.
2015 Liquid Cooling Controls Case Studies Webinar
Today’s practice for liquid cooling is to use CDUs with constant flow-rate and temperature. Tomorrow’s products could be designed for variable flow-rate and temperature based on actual heat removal requirements, but there would have to be more and finer grained telemetry and controls. How much savings can be gained in energy savings compared to the incremental capital and operational costs? This webinar will explore case studies from major HPC centers that have implemented and/or are considering dynamic controls for liquid cooling.
Marriann Silveira, Lawrence Livermore National Laboratory, will open the webinar with a statement of motivation and goals.
Tom Durbin, National Center for Supercomputing Applications, describes controls that utilize variable frequency drives to minimize energy use in chilled water pumps.
David Martinez, Sandia National Laboratory, describes components and controls working in unison to prevent wasted energy and improve reliability.
Greg Rottman, U.S. Army Engineer Research and Development Center (retired) and Charles Rush, Hermetic Rush Services, describe four modes of operation; normal, precooling, comprehensive cooling, and load balance.
Torsten Wilde, Leibniz Supercomputing Center, shows work in progress on using models and simulations of the LRZ hot water cooling infrastructure.
David Grant, Oak Ridge National Laboratory, will provide closure and describe next steps.
2015 Swiss Electricity Service Provider Perspectives Webinar
Wolfgang Korosec is the Chief Information Officer at St. Galler Stadtwerke, an electricity provider in Switzerland. He will present his perspectives on power provisioning from a supplier point of view. This presentation will give a broad overview of the inner workings of an electricity provider. What are their sources and markets of electricity supply today and tomorrow? What factors influence pricing of electricity? Are there unique relationships and arrangements for large customers, like hospitals and supercomputer centers? What recommendations are suggested and what options exist for supercomputing centers to develop closer partnerships with their electricity providers. These are some of the questions that Wolfgang will answer in this webinar.
Wolfgang is CIO of Sankt Galler Stadtwerke, a power utility serving electricity, gas, water, heat and telecommunications. Wolfgang is also President of the Swiss Silicon Valley Association, a non-profit Swiss organization promoting technical and managerial exchanges between the Silicon Valley and Switzerland. He started his professional career as a consultant focusing on computer integrated manufacturing. Wolfgang joined ETH, the Swiss Federal Institute of Technology, where he had the unique opportunity to work on many innovative ICT projects during the rise of the Internet and the World Wide Web. Wolfgang Korosec holds a Master's degree in Mechanical Engineering and Economics from Vienna University of Technology.
2015 Liquid Cooling Controls Vendor Webinar
You are invited to attend a webinar where system integrators - including HP, Cray, Bull, Lenovo and RSC – will talk about their vision and roadmap for HPC liquid cooling controls.
Many HPC liquid cooling systems are operated with fixed parameters, like temperature and flow rate. There may be energy savings opportunities for these parameters to vary, based on real-time load changes.
What are the new and upcoming technologies that would provide liquid cooling controls to promote energy efficiency? What controls systems should be built into the HPC center building infrastructure and what should be included in the HPC system. What opportunities may exist to manage the power – and cooling - loads with resource and job scheduling capabilities.
These are some of the questions that have arisen in the Energy Efficient HPC Working Group Controls Team. This webinar provides a venue for exploring answers to these kinds of questions.
2015 Designing for Liquid Cooling Webinar: LANL Case Study
Liquid cooling is key to dealing with the heat density, reducing energy consumption, and increasing the performance of this generation of supercomputers and becomes even more predominant for the foreseeable future. The transition to liquid cooling, however, is not without challenges.
This webinar will focus on some of the challenges with facilities design for liquid cooling and uses Los Alamos National Laboratory (LANL) as a case study. Liquid cooling design, approvals, and construction requires long term planning. LANL plans decades ahead, and initiated the process over 4 years prior to the arrival of Trinity, their next liquid cooled platform. Liquid cooling enables deployment of energy efficient advanced technology systems and promises to improve the overall energy and water use efficiency by about 25%. Attend and hear about the design considerations as well as an update on the construction and implementation progress.
2015 Dynamic Power Management Webinar
Power management is one of the most critical challenges on the path to exascale. Supercomputing centers today are designed to be worst-case power provisioned, leading to two main problems: limited application performance and under-utilization of procured power. This talk will introduce hardware overprovisioning: a power-efficient design approach for future supercomputing centers that addresses the aforementioned problems.
Page updated
Report abuse