AMD Developer

From smarter LLM pruning to training performance insights on AMD Instinct MI355X + ROCm 7.0, here are three must-read updates for anyone building with AMD. 👇 Dive into the details Týr-the-Pruner: Search-based Global Structural Pruning for LLMs: https://bit.ly/44bt0f8 The vLLM MoE Playbook: A Practical Guide to TP, DP, PP and Expert Parallelism: https://bit.ly/48CPKWD Optimizing LLM Workloads: AMD Instinct MI355X GPUs Drive Competitive Performance: https://bit.ly/4iL04kf

More seamless compatibility between PyTorch and ROCm means researchers can focus on modeling, and kernel engineers can keep pushing performance. Alex (Wei) C., CEO of Nexa AI, shares why this matters for both sides of the stack.

Results on whether a hot dog is a sandwich or not remain inconclusive. Comment to settle the debate. 🌭

🎉 TensorWave turns 2! 🎉 Two years ago, we set out to prove that an AMD-only cloud could power serious AI at scale. Today, we’re supporting some of the most demanding AI teams with high-performance GPU clusters, open ecosystems, and a focus on real-world results... not hype. Huge thank you to our customers, partners, and the TensorWave team for believing in a different way to build AI infrastructure. Here’s to the next chapter. Head to https://tensorwave.com/ to learn more.

No vendor lock-in FTW.

👏

Making GenAI faster, smarter, and more cost-effective. Fireworks AI is tackling all three with tuned inference, fine-tuning workflows, and a platform built for real-world scale—powered in part by AMD. Hear more from their CEO and co-founder Lin Qiao.

Excited to share our NeurIPS 2025 Tutorial on How to Build Agents to Generate Kernels for Faster LLMs (and Other Models!) A collaboration across institutions: AMD, Stanford, Google DeepMind, Arm, Nvidia, Meta, Modular AI, UC Irvine, and ML Commons. - If you're an AI researcher, check out the parts of AI compute (e.g. GPU) that affect your inference & training jobs -- from why you should do batch sizes in certain multiples of 2 to how your trillions of matrix multiplications make their way through different types of memory on a chip - Also chat with us about the trend of self-improving LLMs that can write code to make them even faster. Agents -> SFT data -> RL with profiling tools for strong reward signals - Come build with us! It's an exciting future where LLMs can make the next generation of AI research (RL, even more sparsity, hybrid models with SSMs) super fast on any new hardware generation -- on-the-fly in your IDE Sina Rafati, Hao Li, Azalia Mirhoseini, Anna Goldie, Laurence Moroney, Vartika Singh, Mark Saroufim, Chris Lattner, Sitao Huang, David Kanter, Simon Guo, Kesavan Ramakrishnan, Vincent Ouyang, Tim Gianitsos, Nithyashree Manohar, Sharon Zhou, with acknowledgements to the AMD GEAK team Zicheng Liu, Dong Li, Ziqiong Liu, Pratik Prabhanjan Brahma and the AMD ROCm Kernel team ZHAOYI LI and the AMD Omni team Muhammad Awad, Keith Lowery, Cole Ramos and kernel engineers John Tyler, Muhammad Osama and of course the one and only Emad Barsoum. When they're out, will drop... * arxiv paper link * Github repo link