$AMD| People that think $META going Hyperscaler is bearish for @AMD are the same people that said "Do we even need AMD?" when the stock was around sub $100 or lower. Remember those idiots? Not Financial Advice! DYOR! These fund managers are just clueless. $META planned this for a long time, way before Chatgpt OpenAI got the attention or before market bought the GPU Hype from $NVDA or the world only run on GPU. If u dont understand it, just say u dont get it. I will spell it out right here for u: $META been working with $AMD long before $NVDA came to the picture, and they specifically want AMD for Inference. Inference is how Hyperscalers gonna make money long term, Dr. Su called this years ago. This is not the we overbuilt and sell all Semi false thinking. It is a simple decision that was planned for more than 5 years. Inference has much higher margin than Training, and AMD has the best chips with lowest TCO & TDP. That was why @Meta went heavy on $AMD chips early on and contributing to ROCm(Software) and hardware feedbacks. That is how META deserved the warrant. And yes Meta will buy much more than just 6GW from AMD, they will buy lots of CPUs/Agentic AI racks from AMD. $META will have 35-40% revenue growth from here for minimum of 8-12 quarters. $AMD will have crazy 3 digits growth from here because it is the main supplier of best CPU per socket for Agentic AI Era, where CPU:GPU Ratio is 1-5:1. Even with an optimistic estimate from Morgan Stanley, 6.75m EPYC Venice sold in 2027 will just be enough to rebalance about 10-15% of the imbalanced Ratio that was mostly GPUs from 2022-2025. CPU shortage is going to last for years, not months even with the most aggressive 2nm capacity Ramp from TSMC. If u dont understand this matter, u only have to read @MikeLongTerm thousands of thread on $AMD $TSM $META $MSFT $AMZN. Alright, that is it. Not Financial Advice! DYOR!