GPU, CUDA, and PyTorch Performance Optimizations

GPU, CUDA, and PyTorch Performance Optimizations Join the AI Performance Engineering Meetup (Amsterdam) for a hands‑on online session focused on squeezing every drop of performance from your deep learning workloads. This meetup is designed for AI engineers, researchers, and practitioners who work with GPU‑accelerated models and want to go beyond basic implementation. ### Event Details - Date: Monday, 17 August 2026 - Time: 6:00 PM (CEST) - Location: Online (Zoom link provided to registered attendees) - Recurrence: Every 3rd Monday of the month until December 2026 ### What to Expect The evening kicks off with Talk #0: Introductions and Meetup Updates hosted by Chris Fregly and Antje Barth. You’ll then dive into two technical talks – both titled GPU, PyTorch, and CUDA Performance Optimizations – that cover real‑world optimization patterns you can apply immediately. ### Who Should Attend - AI/ML engineers looking to reduce training and inference latency - Developers using PyTorch and CUDA who want to better understand GPU hardware utilisation - Anyone interested in systems performance engineering for AI ### Additional Resources - 📁 GitHub repo with code and slides: ai‑performance‑engineering - 📘 O’Reilly book on systems performance: Systems Performance Engineering - 🎥 Catch previous talks on the AI Performance Engineering YouTube channel - 🆓 Generative AI free course on DeepLearning.ai ### Ready to Optimise? Register for free on Meetup – the Zoom link will be visible after you RSVP.