Chip, Machine Learning, Software, System, Cloud

Cerebras Goes Big at Supercomputing 2021

Author: Rebecca Lewington, Technology Evangelist | November 16, 2021

There’s a lot to do at the Cerebras Virtual Experience

The Supercomputing Conference is where the High-Performance Computing (HPC) world comes together to show off the latest advances in high performance computing, networking, storage and analysis. It’s always an amazing show. 

We couldn’t be there in person, so we built a virtual booth instead, where you can hear from our extraordinary customers, take apart one of our CS-2 systems and learn about our on-premises and cloud consumption options. We can also take you inside our developer experience and introduce you to our new SDK, which opens up new possibilities at the leading edge of AI and HPC. Read on for more details about all the fun things you can do at the Cerebras Virtual Experience! 

 

Technical Talks

Join us Tuesday, November 16 at 11:00am CST to hear our VP of product management, Dr. Andy Hock, give a talk on “Powering Extreme-Scale AI and HPC Computer with Wafer-Scale Accelerators,” where he explores how our architecture can achieve massive performance gains by exploiting the underlying mathematical similarity between deep learning and key HPC workloads. (Watch at SC21) 

We’re thrilled to showcase the work of Dr. Mino Woo, postdoctoral researcher at the National Energy Technology Laboratory, in his technical talk “First Engineering Problem on Cerebras Wafer-Scale Engine”, where he demonstrates how wafer-scale integration can help them meet their grand challenge of reaching net-zero carbon emissions by 2035. (Watch on YouTube) 

System Virtual Tour

One of the joys of attending the Supercomputing show is the chance to look under the hood of the latest hardware and talk to the inventors. We can’t do that this time, so we’ve done the next best thing and built a virtual product tour. Learn what a marvel of mechanical, thermal, electrical and semiconductor engineering co-design the CS-2 system is. (Take the tour. Firefox, Chrome or Edge recommended.)

Customer Stories

I was lucky enough to sit down with Dr. Dirk Van Essendelft, Principal Investigator in the Advanced Computing and Artificial Intelligence Team, also at the National Energy Technology Laboratory, and our co-founder and chief architect, Michael James for a lively panel discussion about the lab’s mission to deliver solutions for an environmentally sustainable and prosperous energy future. (Watch on YouTube) 

I was also fortunate enough to talk to Dr. Venkatram Vishwanath, who leads the data science team at the Argonne Leadership Computing Facility at Argonne National Laboratory. Cerebras’ director of product, Jessica Liu helped me grill Dr. Vishwanath about the lab’s work and the ALCF AI-Testbed, which aims to evaluate the usability and performance of machine-learning-based high-performance computing applications, running on next-generation AI accelerator machines. (Watch on YouTube) 

The Cerebras Developer Experience

If you’ve only got two minutes (literally), watch Andy Hock’s introduction to our new software development kit (SDK), which is designed to allow developers to program at a more flexible, lower level directly for our architecture. (Watch on YouTube) 

We persuaded three of our engineers to go a little deeper: 

  • Overview of the Cerebras Software Stack, with director of product Jessica Liu (Watch on YouTube) 
  • Programmability & Architecture Benefits for HPC, with performance engineer Adam Lavely (Watch on YouTube) 
  • Introducing the Cerebras SDK, with product manager Justin Selig (Watch on YouTube) 

Ways to Unlock Cerebras Power

Many of our customers choose to deploy Cerebras’ systems in their on-premises data centers. But the sad truth is that not everyone has a data center, which is why we created our Cerebras Cloud offering with Cirrascale to bring Cerebras power to more usersVisit this part of our booth to learn more. 

Looking to the future

You may also have seen in the news recently that we raised $250 million in a Series F financing round, which will enable Cerebras to further expand our business globally and deploy our industry-leading CS-2 system to new customers, while continuing to strengthen its leadership in AI and HPC compute. 

As Andrew Feldman, our CEO and co-founder puts it: “This new funding allows us to extend our global leadership to new regions, democratizing AI, and ushering in the next era of high-performance AI compute to help solve today’s most urgent societal challenges – across drug discovery, climate change, and much more.” 

We can’t wait.