Supercomputing has evolved from niche applications, such as weather forecastsand is now powered by artificial intelligence and machine learning, with the power of these systems increasing dramatically.
The present and future of High Performance Computing were about to see The international conference for high-performance computing, networking, storage and analysisalso known as SC22, which took place from November 15th to 17th. SiliconANGLE Media’s live streaming studio, TheCUBE, covered the entire event, including exclusive interviews with experts from leading computing companies, focusing on supercomputing and cloud computing. (*Disclosure below.)
In case you missed it, here are three key takeaways from the SC22 event:
1) Supercomputing seems poised to explode globally.
Before kickoff of SC22 theCUBE host and industry analyst John Furrier Let’s put it this way: “It’s a hardware nerd fest of the top minds.”
This was evident across the convention floor with a flurry of important announcements from the industry’s top players. Nvidia Corp. introduced new high-performance computing solutions presented to researchers, Cerebras Systems Inc an AI supercomputerand Hewlett Packard Enterprise Co. announced a new portfolio for its Cray line of supercomputers.
During the Day 1 keynote analysis segment, theCUBE co-host Savannah Peterson – who describes herself as a “very open-minded hardware nerd” – said SC22 got her excited to see what early AI trends are taking root in business use cases. 5-10 years ago there was a lot of hype around quantum and AI, but these advanced technologies have historically been difficult to scale because the hardware wasn’t ready to support these types of endeavors.
“Now we have air cooling, we have liquid cooling, we have multiple [graphics processing units]. Dell showed me all eight of them, which they built into their beautiful million-dollar machine, which is extremely impressive for people who do complex calculations,” recalled Peterson. “But what excites me is that I love it when we merge business and science. I don’t think that happens often.”
Here is theCUBE’s full video of the Day 1 keynote analysis:
2) Dell plans large-scale AI and high-performance computing initiatives.
Among the many notable announcements at SC22 was Dell Technologies Inc. revelation his PowerEdge XE9680 system. Announced as the first system with eight Nvidia GPUs connected to Nvidia’s NVLink, Dell designed it to meet growing customer demand for large-scale artificial intelligence and high-performance computing initiatives.
Travis Vigil, senior vice president of product management at Dell, said the company is hearing from its customers that AI workloads and traditional HPC workloads have become almost indistinguishable. The important thing is to understand what customers are trying to do by taking all the different streams of data – both structured and unstructured – and bringing them together to make business decisions.
“It’s a really exciting time because customers are saying the same things that researchers and universities have always been trying to do with HPC: ‘I want to do it on an industrial scale, but I want to do it in a more open way, more flexible,'” he said he.
What Vigil calls “AI for the rest of us”. the latest iteration of Dell’s PowerEdge lineup will offer additional support for AI initiatives the company announced earlier this year, including Dell validated design for AI.
Here is theCUBE’s full video interview with Travis Vigil:
3) Where is all this leading? To space.
High performance computing has compelling use cases across multiple industries, but none is as exciting as the last frontier. at NASA’s Global Modeling Assimilation Office, scientists maximize the use of observations from space and use them to build a coherent, consistent physical system of the Earth. The requirements are enormous because it involves modeling all of NASA’s Earth data.
Over the next decade, NASA made a long-term Obligation to build one Open Source Scientific Community. That’s why scientists are excited about the technology they expect to be able to use soon, according to two NASA scientists who spoke to theCUBE during the event.
“For me, it’s the accelerator technologies. The limiting factor for us as scientists is how quickly we can get an answer,” said dr Bill Putman (pictured left)Associate Chief of the Global Modeling and Assimilation Office at NASA Goddard Space Flight Center, theCUBE said during the event.
“If we can get our answer faster through accelerated technologies, with the support of the NCCS and the data centers, but also the software developers that enable us to do that, then we can do more and push the questions even further,” Putman said.
dr Dan Duffy (To the right)Head of the Computational and Information Sciences and Technology Office at NASA Goddard Space Flight Center, said what excites him is the software frameworks surrounding the simulations his colleagues want to run.
“The technology, they deal with how to use it in the most energy efficient and most efficient way; it’s extremely important,” he said. “Is that what I’m really here to try to better understand how to support these scientists not only with the hardware but also with the software frameworks that will allow them to be successful?”
Here is theCUBE’s full video interview with Dr. Bill Putman and Dr. Dan Duffy:
To see more of theCUBE’s coverage SC22here is our full event video playlist:
(* Disclosure: TheCUBE is a paid media partner for the SC22 event. Neither Dell Technologies, the primary event coverage sponsor of theCUBE, nor any other sponsors have editorial control over the content on theCUBE or SiliconANGLE.)
