Scott Tease, Vice President & General Manager of HPC & AI at Lenovo
As the HPC world gathers in Hamburg, Germany for the International Supercomputing (ISC) conference, a deep and heartfelt sigh of relief can be heard from every corner of the globe. After two years of virtual events, Zoom calls, webinars, working from home and travel restrictions we can finally gather together in person for an ISC show. And as we share a handshake and a beer in Hamburg, we are at a much different place in the HPC community than we were in 2020.
The pandemic impacted all of IT in profound and very visible ways. Offices shuttered, with only essential personnel allowed onsite (which I found out at Lenovo did not include the VP of HPC & AI). Companies and governments rushed to set up remote access servers and VPNs to accommodate their now housebound staffs. Device sales boomed as schools scrambled to equip every student with a means to access the virtual classroom. While supply chains in nearly every industry buckled under the strain of curtailed component supplies from China and other countries, IT was brought to its knees.
For the HPC community, the effect was equally dramatic and will be felt for years to come. Research facilities around the globe suddenly shifted as many cycles as possible into the pursuit of COVID therapeutics and vaccines. Waves of public and private funding were instantly available to everyone who could throw compute power at the problem and solve it faster. This shift to COVID research would, understandably, displace other work on the cluster. As the pandemic dragged on, that ‘other work’ would need to get done somehow.
All at once, we had a pandemic separating us, amped-up need for compute resources to fight the pandemic, and a global supply chain shortage causing logistical nightmares. HPC directors were forced to rethink how HPC would be consumed going forward. Public cloud solutions helped to ease some of the pressure, but public cloud limitations and costs add up and aren’t a long-term solution for everyone. Some as-a-Service (aaS) type offerings came out as well, but they weren’t built for HPC and the unique utilization and configuration flexibly HPC demands.
The challenge wasn’t simply a matter of fixing the supply problem or getting people immediate access to compute cycles, it was a fundamental shift in how customers approached the acquisition of HPC equipment. Our Lenovo HPC team maintained open lines of communication with the community throughout the pandemic. We knew this was an inflection point in the industry. A new paradigm was emerging of three tiers of computing: on-prem purchase, which will continue to be the most economical way to acquire hardware, as a Service, to respond quickly with readily available, partitioned resources that are on site, and access to the cloud for additional burst capability and access to unique technologies. Using these guidelines, we announced our HPC-as-a-Service vision back in January at our (hopefully) last virtual event. We call it Lenovo TruScale HPC, and we’ll have experts on hand at ISC to answer your questions.
Our team is looking forward to seeing everyone in person and getting back to shows the way they were meant to be: in-person. Stop by our booth at ISC (#B209) to learn more details about Lenovo TruScale HPC. You can also see an AI solution that’s out of this universe. We’ll also have some very interesting, super-secret hardware to show. I can’t go into details but suffice to say it’s like an old bridge going over a deep blue cascading river. (If you get that reference, you’re our kind of people!)
ISC is on! It’s live and in-person, and we can’t wait to see all of you there!
