The HPE Data Science Institute, NVIDIA Deep Learning Institute, Texas A&M Institute of Data Science, Texas A&M High Performance Research Computing, and Texas Engineering Experiment Station collaborated to create an online workshop. The workshop was developed to help students increase their knowledge in deep learning systems and other tools.

One of the workshop speakers was Kristopher Keipert, senior solutions architect at NVIDIA. Keipert provided insight on the purpose of his presentation and what people could gain from it. For more information and for further workshops offered by NVIDIA, visit www.nvidia.com/DLI.

• At the NVIDIA workshop, what was your presentation about?
  KK: My presentation was on the capabilities of the new NVIDIA A100 Tensor Core GPU. Since the workshop was focused on data science, we focused on the third-generation tensor cores that speed up AI training operations by up to 20x.
   
• What was the inspiration for creating this presentation?
  KK: The HPE Data Science Institute is both a nexus for cutting edge data science research, and a training hub for the next generation of data scientist practitioners. Knowing the audience, we were inspired to create a presentation that showed the cutting edge of hardware and software available to researchers looking to accelerate their data science workloads.
   
• What did you want the audience to gain from your presentation?
  KK: I wanted the audience to get up to speed on the latest technology for data science acceleration. Naturally, researchers tend to size their problem to the resources available to them. I hope that our introduction to the A100 Tensor Core GPU has inspired the researchers to think about larger, more complex problems that they can now explore.
   
• What was your favorite part of the workshop?
  KK: The excellent questions asked by the audience. While most of us have quite a bit of practice with Zoom meetings at this point, it can still feel difficult to connect with each other. I really appreciated how highly engaged the audience was, along with all of the insightful questions.
   
• Why should other people attend this event?
  KK: In data science and machine learning, research and development is moving at breakneck speeds. The state of the art today probably won’t be the state of the art two weeks from now. If any of us can hope to keep up, communication is key. These events give everyone the opportunity to share learnings and ideas. We have a lot to learn from each other!