Anyone who has been around the HPC world for a while can appreciate how much easier system rollout and maintenance has become over the years, despite the ongoing challenges. Intel has long seen improving scalability and ease of deployment in HPC as critical. And getting from yesterday’s reliance on one-off systems to today’s solutions for workload optimized configurations has required a tremendous amount of behind-the-scenes collaboration with ISVs, standards organizations and OEM partners. It’s important work because the easier and more efficient HPC systems are to use, the faster we can advance critical science advancements and business applications.
Bringing order to chaos
Roughly 15 years ago, HPC system rollouts looked far different than they do today, and the challenges were significant. In the early days, every OEM created their own configurations and there were no defined standards. ISVs had to validate their software on various configurations. For example, it wasn’t enough to test on Red Hat Linux; ISVs had to test on specific clusters with the specific interconnects that customers planned to use. Intel realized it needed to make life easier for ISVs and customers alike.
Around 2010, Intel began working on ways to address the headaches associated with HPC system rollout and maintenance. As part of its early efforts, it developed the Intel HPC Platform Specification to help ensure more commonality and interoperability among systems for greater compatibility with ISV applications. Intel also introduced the Intel Cluster Ready program to create common “ready-made” configurations for key user groups. The program has matured to become today’s Intel Select Solutions program, building on the past decade’s lessons and technology advancements, such as the Intel Cluster Checker.
A focus on collaboration and problem solving
The ongoing progress Intel makes with solutions is only possible with the work of armies of engineers and close cooperation with a host of industry players. Intel is focused on meeting rapidly evolving ISV and customer needs. This requires close collaboration with ISVs—listening to their feedback, helping them with their solutions and providing assistance in partner engagements. Intel believes ISVs know what’s best for their software and works with them to optimize base Intel solutions for their verticals and target markets.
Intel also contributes heavily to key open-source codes, such as Linux, LAMMPS and GROMACS. Intel participates in organizations such as the OpenFabrics Alliance that are key to progress in HPC. And it works with cloud service providers to address gaps between on-prem and cloud solutions so users can consume compute cycles in the cloud efficiently and easily.
All of this ongoing cooperation and innovation has helped spark growth in the HPC ecosystem.
Moving ever closer to “out-of-the-box” optimization
HPC systems have come a long way thanks to the work of Intel and its partners. Today’s Intel Select Solutions simplify deployment and make it much easier for users to get closer to the maximum compute value from their systems. The Intel Select Solutions portfolio includes an array of configurations that are customized for critical workloads, including:
- Genomics analytics
- HPC/AI Convergence
- Simulation & modeling
Because Intel engineering groups test different configurations for each workload type and put them through the paces with key applications, customers can be confident that Intel Select Solutions will be ready for them to focus on the work at hand. And the reference configurations give OEMs a running start on their work.
“Boston was approached to join the Intel Select Solution program using our expertise and Intel’s products to develop a solution for Simulation and Modeling. The process was guided by Intel who were on hand to support us as needed to get the Solution over the finish line. Since its launch we’ve received ongoing Sales and Marketing support to help us get this product in front of the right customers”
—Konstantinos Mouzakitis, Head of HPC, Boston
It’s all about faster, more efficient problem solving
At the end of the day, organizations and businesses are interested in getting their jobs done as easily and efficiently as possible. Intel has long been proactive in working on standards and industry acceptance and trying to find the greatest amount of flexibility for HPC solutions so end users can make progress on their biggest challenges. Intel’s commitment to HPC solutions is the key.
Learn more about how Intel accelerates HPC progress through its long-term dedication to solutions in this blog post from Trish Damkroger, Intel vice president and general manager of HPC.
Intel technologies may require enabled hardware, software or service activation.
© Intel Corporation. Intel, the Intel logo, and other Intel marks are trademarks of Intel Corporation or its subsidiaries. Other names and brands may be claimed as the property of others.
"term" - Google News
May 24, 2021 at 02:25PM
https://ift.tt/3ukIw3w
Why Intel's Long-term Solution Mindset Is Critical to HPC Advancements - HPCwire
"term" - Google News
https://ift.tt/35lXs52
https://ift.tt/2L1ho5r
Bagikan Berita Ini
0 Response to "Why Intel's Long-term Solution Mindset Is Critical to HPC Advancements - HPCwire"
Post a Comment