The next big wave of artificial intelligence (AI) is upon us with agentic AI. This is where AI goes from simply generating a response — whether it be text, a photo, or a video — to independently performing assigned tasks based on instructions, and, hopefully, appropriate guardrails.
With agentic AI comes the promise of agentic commerce, where AI agents can put items in your cart based on your requests, letting you just click a button to check out. It also offers the possibility of a virtual workforce that works side by side with human employees.
Will AI create the world’s first trillionaire? Our team just released a report on the one little-known company, called an “Indispensable Monopoly” providing the critical technology Nvidia and Intel both need. Continue »
Not surprisingly, many companies are pursuing the use of AI agents, and there are certainly numerous software-as-a-service (SaaS) companies going down this path. However, perhaps the best way to play agentic AI is on the hardware side. And when it comes to agentic AI hardware, one of the best-positioned companies is Advanced Micro Devices (NASDAQ: AMD).
An agentic AI winner
While best known as the distant No. 2 player in the graphics processing unit (GPU) market behind Nvidia, AMD is actually the market leader in data center central processing units (CPUs). This is a great position to be in, because with the rise of AI agents, there will be a huge need for high-performance CPUs.
GPUs are great at providing the muscle to train AI models and help run inference. What they are not good at is working with other tools and sequential reasoning. That is where CPUs step in, acting more like the brains of the operation. With agentic AI, data center servers will need a whole lot more CPUs to orchestrate the growing use of AI agents.
Image source: Getty Images.
Both AMD and rival Intel have discussed the GPU-to-CPU ratio in data centers shifting from 8-to-1 to 1-to-1 with agentic AI. That is leading to a huge surge in demand, with CPU demand now starting to outstrip supply. At…
..
