Five thoughts from Lisa Su’s keynote at AMD Advancing AI - SiliconANGLE

Five thoughts from Lisa Su’s keynote at AMD Advancing AI - SiliconANGLE

AMD positions Helios as delivering 30% more tokens-per-dollar versus Nvidia and repositions Venice CPUs as critical for agentic AI, claiming 1.8x performance gains over the prior generation. For CTOs, this signals a viable vendor alternative reducing Nvidia lock-in and enables CPU-forward architectures for cost-optimized agent inference workloads.