The Challenger Stack - AMD, frontier labs, and the new compute playbook

​The AI compute landscape is shifting. Leading frontier labs are no longer building on a single chip. They're architecting multi-vendor stacks, and AMD is increasingly central to how the most ambitious teams scale training.

​Join us for an evening of candid conversation with the operators and analysts who are living this transition firsthand. Our confirmed speakers include:

• ​Jeff Tatarchuk, Co-Founder, TensorWave

• ​Dylan Patel, Founder, SemiAnalysis

• ​Tim Davis, Co-Founder & President, Modular

• ​Mark Papermaster, CTO & Executive VP, AMD

• ​Eugene Cheah, CEO & Co-Founder, Featherless AI

​Together, they'll take the stage for a 45-minute panel covering the real-world case for AMD compute at scale, what it actually looks like to run a multi-chip lab across AMD and NVIDIA, how frontier teams are thinking about hardware strategy as training demands grow, and the emerging compute playbook for labs that want optionality without sacrificing performance.

​This is an intimate event designed for decision-makers, whether you're leading infrastructure at an enterprise deploying AI at scale or building the models yourself. This is a conversation about the choices that will define your stack for the next several years.

​After the panel, stick around for a networking happy hour to continue the conversation with speakers and fellow attendees.