[Update December 1, 2022. You can read our post-show wrap-up here.]

The Supercomputing Conference, SC22, is where the High-Performance Computing (HPC) and Artificial Intelligence (AI) worlds come together to show off the latest advances in supercomputing, machine learning, networking, storage and visualization. It’s always an amazing show. And it’s great to be back in person, from November 13th to the 18th in Dallas, TX.

Cerebras is definitely going to be contributing to the amazement factor this year. We’ll have a booth stuffed with our unique hardware and staffed by HPC and AI experts, a range of great sessions and some new performance results that will knock your socks off!

Hands-On With the World’s Fastest Accelerators at Booth 2833

Look for our hanging sign as you come though the main entrance. We’re in a good neighborhood, just to the right of Intel, Microsoft and AWS.

We’ll have a pair of real Cerebras CS-2 systems, the world’s most powerful AI and HPC accelerator, and a petting zoo where you can get a look at the inner workings of this amazing feat of semiconductor, electrical and mechanical engineering. And of course, you can grab a selfie with the biggest chip in the world, the Cerebras Wafer-Scale Engine.

You can also explore our easy-to-use software stacks for both AI and HPC. Our expert staffers can take you as deep as you care to go into AI customer use cases, and we’ll have a dedicated SDK demo, which opens up new possibilities to greatly accelerate a range of widely used HPC codes such as stencils and field equation modeling.

You can also take a seat in our mini theater, where we’ll host a series of short talks from our customers’ technologists and our own. Stay tuned for more details.

Announcing our Academic Partnership Program

We recognize that not everyone who wants one can put a CS-2 system under their desk, but fear not, we’re here to help. Through our new Academic Partnership Program, we can assess your workload and optimize it for our systems, connect you to our network of Cerebras-equipped institutions, help you submit grant proposals to secure system time, and then amplify the impact of your research. So join us, and co-design the future!

Technical Sessions

On Monday, November 14th at 9:30am CST, Machine Learning Solutions Engineer, Cindy Orozco Bohorquez joins a team from Argonne National Laboratory to present the results of “A Comprehensive Evaluation of Novel AI Accelerators for Deep Learning Workloads”. This is the first evaluation of its kind, exploring the challenges and opportunities in integrating technologies such as Cerebras CS-2 in supercomputing systems.

In the afternoon of the same day, from 1:30 – 5pm CST, Cindy is also presenting a tutorial on “Programming New AI Accelerators for Scientific Computing”. A highlight of the session will be a hands-on session using the Cerebras systems at Argonne National Laboratory’s ALCF AI Testbed.

On Tuesday, November 15th at 2:30pm CST, our Director of Product and all-around AI guru, Natalia Vassilieva, will be knocking off those socks at the Exhibitor Forum. The title of her presentation is something of a spoiler: “Near-Perfect Scaling on 13 Million Core Cluster Made Easy”

Later that same day, at 4:30pm CST, Mathias Jacquelin, Member of Technical Staff will present an exciting paper, “Scalable Distributed High-Order Stencil Computations”. This paper was co-authored with Mauricio Araya-Polo and Jie Meng from TotalEnergies and details how the unique properties of the Cerebras Wafer-Scale Engine enable this traditionally memory-bound algorithm to become compute-bound and achieve near-perfect weak scaling.

On Wednesday, November 16th, from 10:30 am, we’ll be at the awards ceremony for the 2022 Gordon Bell Special Prize for HPC-Based COVID-19 Research. Why? Because we might win! A team from Argonne National Laboratory, Cerebras and others has been nominated for their paper: “GenSLMs: Genome-scale language models reveal SARS-CoV-2 evolutionary dynamics“.

On Thursday, November 17th at 12:15 CST, VP of Product Management, Andy Hock, joins our colleagues from Oak Ridge and Argonne, and Lawrence Livermore National Laboratories at a Birds of a Feather session titled “Transformers at Scale for Science”. These sessions often feature passionate discussions between experts in the same field and this topic is a particularly rich one, so sparks may fly!

Lastly, on Friday, November 18th at 8:30am CST, Natalia is sitting on a panel exploring the question “Quantum Computing: A Future for HPC Acceleration?” With opinions from leading European and US supercomputing institutions and Intel corporation, it promises to offer a compelling reason to see SC22 through right to the end.

About those performance results…

Alas, I’m not going to spoil the surprise. If you’re lucky enough to be going to the show, whether your interest is AI, HPC, or a blend of the two, I guarantee you’re going to be impressed.

For everyone else, you’ll just have to wait a little longer.

Rebecca Lewington, Technology Evangelist | October 19, 2022

Full details of our sessions can be found at cerebras.net/events/sc22.

If you’d like to arrange a private meeting with either our technical or commercial staffers, please reach out to our esteemed Duncan Hoskinson here and we’ll set something up.