Accelerating Time to Insights for Automotive — Live Demo and Presentation at SC23 in Denver, Colorado.
Hello, and thank you for visiting my blog this week. As we approach the weekend, I’m excited to continue our mini-series about byteLAKE’s participation in the upcoming SC23 conference in Denver, Colorado.
As I mentioned in my previous post (you can read it here), this year, we’re shining a spotlight on the automotive industry. Over the past few years, in collaboration with our exceptional partners, Lenovo and Intel®, we’ve been working diligently on some groundbreaking projects centered around Quality Inspection Automation and Accelerating Time to Insights. Today, I’m thrilled to unveil the latest updates for two of our flagship products: byteLAKE’s Cognitive Services and byteLAKE’s CFD Suite (AI-accelerated CFD). In this blog post, I’ll focus on the latter, reserving the discussion of Cognitive Services for an upcoming post.
In a nutshell, the CFD Suite is all about harnessing the power of AI (Artificial Intelligence) to expedite the results of Computational Fluid Dynamics (CFD) simulations. Our journey over the years has been marked by extensive research, benchmarking, case studies with academic partners, and close collaboration with numerous business partners and clients. I want to express my heartfelt gratitude for all the insights, ideas, and feedback we’ve received during this journey. While I can’t list all of our valued clients just yet, I must acknowledge two names that have been instrumental in our progress: Lenovo Infrastructure Solutions Group and the Intel® teams. In my view, successful partnerships are built on creating and delivering impressive solutions collaboratively, and our long-standing partnerships with Lenovo and Intel® exemplify just that.
CFD Suite leverages AI (Artificial Intelligence) to accelerate time to insights and expedite the results of CFD (Computational Fluid Dynamics) Simulations.
Our most recent update to the CFD Suite product introduces “Learning-on-the-Fly,” an umbrella term encompassing several modules that are transitioning from the research phase to commercial availability. The first version of this update is slated for release shortly after the SC23 conference (on November 20, 2023).
Learning-on-the-Fly is an umbrella term for CFD Suite’s modules focused on acceleration and accuracy of AI-accelerated CFD Simulations.
So, what’s included in the CFD Suite’s November 2023 update?
- Learning-on-the-Fly: A suite of modules focused on accelerating and enhancing accuracy.
- AI Accelerator and AI Supervisor, both of which we introduced at the 10th Polish Chemistry Congress earlier this year. Learn more here: AI-accelerated CFD (Computational Fluid Dynamics) | PPT (slideshare.net) and in Polish here: CFD Suite (AI-accelerated CFD) — Sztuczna Inteligencja Przyspiesza Symulacje CFD (prezentacja) | PPT (slideshare.net).
- Improvements to the redesigned Deep Learning and Machine Learning modules, delivering over 10x acceleration for select CFD simulations.
- Enhanced accuracy control mechanisms.
- Optimizations for the latest hardware, including the 4th Gen Intel® Xeon® Scalable processor family (formerly codenamed Sapphire Rapids), Intel® Data Center GPU Flex Series, Intel® Data Center GPU Max Series (previously codenamed Ponte Vecchio).
- Refinements for previously supported hardware and frameworks, including Intel’s® OpenVINO™ and Intel® DL Boost Vector Neural Network Instructions (VNNI), as well as NVIDIA V100 Tensor Core GPU.
- Scalability features, now integrated into Learning-on-the-Fly, supporting HPC configurations, including multi-node setups and nodes equipped with numerous accelerators.
You’ll have the opportunity to witness all of these enhancements in action through live demonstrations at the upcoming SC23 conference in Denver, Colorado. Join byteLAKE team at Lenovo’s booth #601. After the event, we’ll also release a whitepaper summarizing our benchmarks with some of the hardware options mentioned above. We’ll delve into the acceleration of OpenFOAM®’s (*) Motorbike CFD Simulation and explain the seamless integration of CFD Suite with OpenFOAM® (*), and much more.
Experience the potential of CFD Suite (AI-accelerated CFD) in action at SC23, available at Lenovo’s booth #601.
Additionally, I’ll be delivering a live presentation at Lenovo’s Theater (booth #601) titled “Supercharging CFD Simulations with AI for Accelerated Insights.” Mark your calendar for Tuesday, November 14th, at 3:45 PM.
Join my presentation at Lenovo’s Theater at SC23 in Denver, Colorado, on Tuesday, November 14th, at 3:45 PM.
Looking ahead to 2024, here are some highlights from our roadmap:
- Auto-learning module as part of the Learning-on-the-Fly umbrella.
- Support for large meshes.
- Further enhancements in acceleration and accuracy for an expanding array of CFD solvers.
- AI is everywhere, but what more can it bring to Manufacturing and Automotive in specifics? Explore these AI solutions during the SC23 conference in Denver, Colorado. | by Marcin Rojek | Oct, 2023 | Medium
- Automated Quality Inspection for Automotive — AI in Action | by Marcin Rojek | Nov, 2023 | Medium
See you in Denver!
This offering is not approved or endorsed by OpenCFD Limited, producer and distributor of the OpenFOAM software via www.openfoam.com, and owner of the OPENFOAM® and OpenCFD® trade marks.
OPENFOAM® is a registered trade mark of OpenCFD Limited, producer and distributor of the OpenFOAM software via www.openfoam.com.