Two robots can complete the same task but behave in very different ways. So how do we measure success? Join the next AMD AI Research Club session with Pier Luigi Dovesi, Robotics & Autonomous Systems Lead, AMD Silo AI, and Yi Ru (Helen) Wang, Ph.D. student at the University of Washington, for a discussion on rethinking benchmarks for robotic manipulation. We’ll explore why success rates alone fall short, how structured evaluation can reveal deeper robot behavior, and the shift toward flexible robotic benchmarks. 📅 June 17 ⏰ 9–10 AM PT | 6–7 PM CET Register here: https://lnkd.in/ec_aRXun
AMD Developer
Semiconductor Manufacturing
Advancing AI innovation together. Built with devs, for devs. Supported through an open ecosystem. Powered by AMD.
About us
Advancing AI innovation together. Built with devs, for devs. Supported through an open ecosystem. Powered by AMD.
- Website
-
https://www.amd.com/aidevprogram
External link for AMD Developer
- Industry
- Semiconductor Manufacturing
- Company size
- 10,001+ employees
Updates
-
AMD Developer reposted this
Thank you Andrew Zigler for a fun conversation
Span of control used to be a hierarchy problem. You stacked a handful of direct reports under a manager, stacked managers under directors, and repeated until the org chart could reach all the work. But Anush E., VP of AI Software at AMD, is watching that math break. One manager now runs 10 engineers and 90 agents, with sub-agents crunching through the night, making that org chart feel much heavier than it ever did before. More layers = more removal from the work being done, which gets messy when agents churn out code as fast as they do. So directors and above get heavily abstracted from the design process, which can hurt alignment. This resonated a lot with me because it's something we care about a lot too at LinearB, where we equip engineering leaders with metrics and context to wrangle their AI rollouts and actually understand their org's adoption and impact. Meanwhile on the IC front, fellow teammates around that 1000x engineer get left behind. It's as if their coworker has become a context black hole where everything crosses an event horizon, and nothing can escape. That's how the workplace is shifting around the K-shaped curve he describes in this clip, and it's a reminder that it's everyone's responsibility (yes, even the 1000x engineer -- ESPECIALLY the 1000x engineer) to reflect on how to spread their capabilities, reduce the bus factor, and find the new natural shape of productivity for your org.
-
What's on your must-see list for Advancing AI? Now's your chance to explore the sessions, discover featured speakers, and find the topics most relevant to your work. Whether you're focused on performance, deployment, tooling, or the future of AI systems, you'll find plenty to add to your agenda. Start planning your experience for July 22-23 in San Francisco: https://lnkd.in/ezrBx-KN #AdvancingAI #AMDevs
-
-
What could you build with the ultimate local AI developer platform? AMD Ryzen AI Halo is built for #AMDevs working on agentic AI, generative AI, LLM applications, and beyond. With up to 128GB of unified memory and support for popular AI tools and frameworks, it's designed to help bring AI development closer to where innovation happens. Presale is now live. 🔥 Reserve yours now: https://lnkd.in/ekTrJAjz
Build the AI you want. Your Stack. Your Rules.
-
Catch up on the latest from the #AMDevs community! Links below ⤵️ ROCm 7.13: https://lnkd.in/edVftX9D Deploying Hermes Agent for Free on AMD Developer Cloud: https://lnkd.in/e3fiQiAS Accelerating LLM Startup on AMD Ryzen AI: https://lnkd.in/eUHBsNBC AMD Schola v2.1: https://lnkd.in/eJz5YsHn Girls in the Game: https://lnkd.in/eKGYaFbV Picard BAM Processing: https://lnkd.in/etynBfxW
-
Ready to see what we've been building? 🔥 The Advancing AI - Developer Track session catalog arrives June 8! Explore a packed lineup of technical sessions, expert speakers, ROCm certification opportunities, and 30+ hands-on workshops built for developers pushing AI forward. Check out this sneak peek and register now so you're ready to build your schedule when the catalog launches: https://lnkd.in/edueMWQp #AdvancingAI #AMDevs
-
Join the “Ask the experts” session to explore how a fully local multi-agent AI workflow runs on AMD GPUs. We’ll discuss how multiple agents and models work together to analyze unstructured visual content, extract information from data, and generate meaningful insights from complex, information-rich sources. Hosted by Mahdi G. and Shailen Sobhee ➡️ https://lnkd.in/euvA65Kw
-
-
SGLang is helping redefine what's possible in AI inference. Join Ying Sheng, Co-founder & CEO of RadixArk and Co-creator of SGLang, at Advancing AI. As one of the fastest-growing open-source projects in AI infrastructure, SGLang is helping developers push performance, efficiency, and scalability for modern LLM deployments. Hear directly from the people building the technology powering the next generation of AI applications. 📅 July 22-23, 2026 📍 San Francisco Register now: https://lnkd.in/erwWdkWf? #AMDevs #AdvancingAI
-