$AMD EPYC Venice is far superior to $NVDA CPU.
Way too much misinformation out there. You can read the full thread below.
When Hyperscalers and AI Labs try to scale number of AI Agents and train it through RL, the highest core per socket matters.
Only @AMD could bring down Inference cost meaningfully in this Gen. I expect to hear more about EPYC Verano at Advancing AI event.
Not Financial Advice! DYOR!