platoseedplatoseed
An AI physics simulator for AI chip design
DeepSim is building an AI physics simulator. We are currently developing the only thermal simulator to meet AI chip design needs and are validating our tool with Intel. We are a team of three electrical engineering PhDs from Stanford with backgrounds in semiconductor fabrication and design.
DeepSim provides an AI-driven physics simulator designed for chip design, offering scalable, GPU-accelerated multi-scale simulations that aim to deliver faster design insights. The platform emphasizes automated setup and high scalability to enable engineers to explore larger, more detailed designs quickly.
A GPU-accelerated physics simulation platform that automates complex setup procedures and enables multi-scale (nano-to-macro) simulations. It claims 1000x faster performance and the ability to handle simulations much larger than traditional tools, while maintaining accuracy. The workflow is designed to be simple, allowing engineers to focus on higher-level design decisions rather than intricate configuration.
Who it’s for: Engineers and design teams in AI chip design and related semiconductor development who require high-fidelity, large-scale physics simulations with automated workflows.
Team with PhD engineers; mentions academics and partners; company appears to be in growth/innovation phase with a product-focused platform
Stanford EE PhD grad now building an AI physics simulation platform at DeepSim to solve the frustrating parts of my PhD. I'm also really into espresso making and will be happy to make you your best cup of coffee.
Alex is a co-founder of DeepSim, Inc., where he develops AI physics simulators. He earned his PhD in Electrical Engineering from Stanford, specializing in nanoscale thermal simulations, and brings over a decade of experience in the semiconductor industry.
Stanford EE PhD with background in semiconductor engineering, and physics modeling of integrated circuits. Excited about transforming the landscape of engineering simulations with new technology that combines physics, experimental data and AI.
Run complex physics simulations signficantly faster than any commerical tool.
DeepSim introduces an AI-accelerated 3D physics simulation platform that automates setup and runs designs up to 1000x faster with maintained accuracy, aimed at engineers working on complex systems (e.g., semiconductors, batteries, satellites). The launch highlights GPU-accelerated solving, scalable AI models, and early traction with high-detail thermal simulations.
Software and hardware to run neural networks faster and cheaper
The copilot for chip designers
