Intel Executive Discusses Engagement with Energy Department, Argonne, and ECP

Exascale Computing Project · Intel Executive Discusses Engagement with Energy Department, Argonne, and ECP

 

By Scott Gibson

Trish Damkroger, Intel Corporation

Trish Damkroger, vice president and general manager, High-Performance Computing, Intel Corporation

Patricia (Trish) Damkroger is an expert in high-performance computing (HPC) and has more than 27 years of managerial experience in both the public and private sectors. Her current role is vice president and general manager of the HPC organization in the Data Platforms Group at Intel Corporation.

She joins us in this episode of the Let’s Talk Exascale podcast to discuss what excites her most about exascale computing, the scope of Intel’s engagement with the US Department of Energy and Argonne National Laboratory, changes she sees affecting the HPC community the next several years, Intel’s exascale hardware and software efforts, and Intel’s role in preparing the HPC community to take advantage of exascale computing capabilities.

Interview Transcript

Gibson: Hi, Trish, welcome to the program.

Damkroger: Thanks for having me.

Gibson: You’ve worked with supercomputers both on the national labs side as well as on the vendor side at Intel. What excites you most about exascale computing?

Damkroger: High-performance computing is a tool along with experiment and theory to really enable scientists and engineers to tackle the world’s biggest challenges. With exascale, we’re enabling the convergence of modeling and simulation, high-performance data analytics, and artificial intelligence. This is enabled through exponential increases in hardware—such as memory, storage, CPU, and GPU—and software. I believe HPC changes the world and exascale computing capabilities will help scientists and engineers solve problems that previously were out of reach, or impossible. Exascale computing impacts will be far reaching across national security, scientific discovery, economic security, energy security, and health care.

Gibson: Tell me about the scope of the Intel engagement with DOE and Argonne.

Damkroger: We are very fortunate to have such a wonderful collaboration between DOE and Argonne. Intel, Argonne, and Department of Energy remain committed to delivery of the technology for the Aurora supercomputer, which is designed specifically for the convergence of AI [artificial intelligence] and high-performance computing. The system will include a future generation of Intel’s Xeon (Sapphire Rapids) on enhanced SuperFin technology, Intel’s Xe HPC compute micro-architecture (Ponte Vecchio), a future generation of Intel Optane persistent memory, and Intel’s oneAPI software. So, as we move into the future, we’re excited to collaborate with Department of Energy and Argonne in support of the United States’ leadership in advanced computing systems, including exascale, neuromorphic, and quantum computing. We will focus on creating next-generation semiconductor technologies and manufacturing processes, advanced system design, and software enablement to enable advanced scientific discovery, strengthen national security, and improve industrial competitiveness that will lead the next phase of innovation.

Gibson: Intel plays a key role in the nation’s exascale computing strategy. What changes do you see for the HPC community over the next several years as a result of the technology development efforts at Intel that are directly connected to exascale?

Damkroger: We have entered this new era of significantly higher need for compute power, data analysis, and data storage. And in turn this is driving the need for advanced technology that enables unparalleled innovation, new discoveries and transforms the way we do business using data analytics and artificial intelligence. At Intel, we are committed to investing our R&D into our product roadmaps and really positioning that answer with the diverse mix of data center technologies. We talk about scalar, facial, matrix, and vector technologies that focus that architecture to really move us forward. And all of this will impact our exascale computing and the future of exascale.

Gibson: Does Intel have both hardware and software initiatives/activities in place related to achieving exascale, and, if so, how would you describe the balance of your efforts between hardware and software?

Damkroger: Intel hardware and software technologies have been enabling supercomputers for many years. And, today, Intel technologies power most of the world’s Top500 supercomputers. Looking toward the era of exascale computing, Intel continues to invest in leading-edge technologies across six pillars and continues to invest in our product roadmaps. As I shared earlier, we designed Aurora to take advantage of our technologies and are confident Intel’s future CPU, discrete Xe-based GPUs, hardware accelerators, and Optane memory, all linked together with the oneAPI unified programming model, will deliver exascale performance. We are working closely with Argonne and ECP to harness the capabilities of Aurora across existing and new application development. oneAPI frees accelerated computing from the economic and technical burdens of proprietary programming languages. Its unified programming model offers freedom of choice of hardware architectures to accelerate applications. Based on industry standards and open specifications, oneAPI provides a fast path to performance across XPUs and easily integrates with legacy code. Intel’s oneAPI Toolkits, which will go gold later this quarter, offer developers fast, efficient development with a complete set of cross-architecture libraries and tools.

Gibson: Trish, what role does Intel play in helping prepare the HPC community—today and in the future—to take advantage of the capabilities delivered with exascale computing?

Damkroger: Intel is a proud partner with the HPC community to invest in, develop, and deliver hardware and software technologies that enable the IT workloads and use cases for the world’s most advanced computing systems. We strive to work closely with the HPC community to implement our technologies and drive broad ecosystem adoption. We are driven by this passion, knowing that our technologies will be used to address the most critical challenges in national security, energy assurance, economic competitiveness, health care, and scientific discovery. The exascale computing capabilities will enable scientists to solve problems that previously were out of reach, and truly the effects on the lives of people and the world will be immense. I am a huge advocate that HPC changes the world.

Gibson: Trish Damkroger of Intel, thank you for being on the program. Thank you for listening to Let’s Talk Exascale.