AMD: The Next Trillion-Dollar Giant in the AI Revolution
While the market underestimates its potential, AMD is quietly building the foundation to dominate the next decade of artificial intelligence.
🔹 Introduction
Today we’re talking about a company that’s being massively underestimated by the market.
While everyone’s focused on Nvidia, few are paying attention to the real silent powerhouse behind the next AI revolution: AMD.
🔹 The growth no one is seeing
Since 2015, AMD has multiplied its revenue by over 6x, reaching $25.8 billion in 2024.
This growth has been driven by:
Strategic partnerships with giants like Meta, Oracle, and Microsoft
A modular chiplet design that enables fast evolution and adaptability to new workloads
Flawless operational execution and expanding margins across the board
But here’s the kicker: AMD’s best results haven’t even started showing up on the books yet.
🔹 The plan is in motion (and it’s working)
In 2024, AMD kicked off its most ambitious growth cycle yet — the AI era.
The focus isn’t on competing head-to-head with Nvidia in training, but on dominating the most valuable segment of the AI chain: inference.
Real-world example?
Meta is running all Llama models on AMD’s MI300X GPUs.
And more: the ROCm software integration is speeding up. AMD already has multiple large-scale deployments with hyperscalers — a huge milestone proving the tech maturity of its hardware and software stack.
🔹 Growth hidden by the wrong narrative
Despite the progress, the market is still bearish.
Why?
Because it’s focused on short-term guidance and missing the big picture:
The datacenter segment's quarterly revenue grew from $0.8B to $3.9B in just 3.5 years.
The new MI350 GPU generation is launching ahead of schedule
Margins and cash flow are bouncing back fast with more operating leverage
Lisa Su made it clear on the last earnings call:
“We are positioning AMD to grow strongly across all segments, led by datacenter and AI PCs.”
🔹 The power of the chiplet platform
AMD’s secret weapon? Its chiplet architecture.
This model allows for highly specialized products with lower marginal costs.
Real examples:
Adding FPGAs to CPUs → AI capabilities in portable PCs
Delivering more memory per GPU → dominance in inference
Flexibility to tailor products to multiple markets with unique synergies
🔹 AMD’s business portfolio
AMD isn’t just a GPU company. It’s a platform with access to several massive markets:
Datacenter AI: tens of billions in expected revenue over the coming years
AI at the Edge: a $200B market with Xilinx as a key asset
AI PCs: clear leader in CPUs
Gaming and Embedded: currently soft, but recovery expected
🔹 The vision gap
While many investors are selling AMD due to the lack of immediate explosive results, long-term investors understand this:
AMD’s real growth is still ahead
The market will reprice the company once the bets made between 2020–2024 start paying off
AMD is on track to become a multi-trillion-dollar company
🔹 Conclusion
I believe AMD could be the 10x investment of the decade.
The tech is mature.
Customers are betting on it.
The software is scaling.
The roadmap is clear.
And the market still hasn’t woken up.
The only question now is:
Are you in — or are you waiting for the re-rating we all know is coming?
I hope you enjoyed this article. If you did, feel free to share it and follow me here on Substack, where I’ll keep bringing as much value as I can—completely for free.
If you’d like to follow my portfolio more closely and support my work, you can become a paid subscriber. It’s 100% optional, but truly appreciated. See you in the next one!