AI Agents Are Quietly Draining a Resource Nobody Saw Coming: the Humble CPU
Amazon Web Services is rationing CPU capacity, Intel has sold out of server chips, and AMD has doubled its forecast. The AI boom just got a new bottleneck.
Key points
- Amazon Web Services told its engineers to conserve CPU capacity in 2025 as AI workloads caused a surge in wait times for CPU server resources.
- AMD doubled its server CPU forecast, and Intel has sold out of server CPUs through at least the end of 2025.
- Researchers from Intel and the Georgia Institute of Technology found that up to seven of eight stages in a typical AI agent pipeline run on the CPU, not the GPU.
- More CPU cores cut the time to a model's first response by roughly 1.5 to 7 times in tests with longer inputs.
- Analyst Matt Kimball of Moor Insights & Strategy calls the CPU demand spike an "absolute tell" that the chip is now central to any serious AI system.
For the last few years, the AI hardware story was simple: GPUs (specialised chips that do the heavy number-crunching) were scarce and pricey. CPUs, the general-purpose processors inside every laptop and server, barely got a mention. That's changing fast.
Amazon Web Services, the cloud-computing arm of Amazon that rents server capacity to businesses worldwide, has reportedly told its engineers to conserve CPU cycles after surging demand stretched wait times for CPU-based servers. The story was first flagged by IEEE Spectrum AI. We first covered the CPU shortage angle on 16 August 2026, around the same time enterprises were already struggling to track what their AI computing bills were costing them.
Why do AI agents need so many CPUs?
AI agents are software programs that carry out multi-step tasks on their own: searching the web, writing files, or calling other agents for help. The GPU still handles running the AI model itself, but almost everything else falls to the CPU.
Souvik Kundu, a senior research scientist at Intel, puts it plainly. "The CPU does the job of parsing output, figuring out which tool to invoke, making the API call or running the code, collecting the result, and feeding it back." Madhu Rangarajan, AMD's vice president of compute and enterprise AI, says AMD's own testing found seven of eight stages in a realistic agent pipeline run entirely on the CPU.
Picture an AI agent told to build a piece of software. It asks the GPU to generate the code, then hands off to the CPU to write files, fetch packages, run a compiler. The CPU is busy precisely when the GPU is resting, and the other way round.
The maths gets uncomfortable at scale. Matt Kimball, vice president and principal analyst at Moor Insights & Strategy, notes that a single enterprise deployment might start with 100 agents. Those agents spawn sub-agents. "One hundred become tens of thousands, hundreds of thousands, or millions," he says, each one making tool calls that land on a CPU.
What is making the bottleneck worse?
Two research papers, one co-authored by Kundu and one by PhD student Euijun Chung at Georgia Tech, point to a second problem: tokenisation (the step that converts words and punctuation into numbered chunks a model can read).
For a short prompt it's trivial. When an agent has built up a conversation of 100,000 tokens and receives a tool result of 1,000 more, the tokeniser must process the entire sequence from scratch. Chung explains that this happens at every tool call. His paper found that in tests using models including Meta's Llama 3.1-70B and Alibaba's Qwen 3-30B, adding CPU cores cut the time to a model's first response by 1.5 to 7 times, depending on sequence length.
Safety checks pile on further CPU load. Rules that inspect what an agent is about to do, along with small models under a billion parameters that check an agent's intent, tend to run on the CPU because they are compact and latency-sensitive.
What does this mean for prices and availability?
Intel's server CPU stock is sold out through at least the end of 2025. AMD has doubled its server CPU forecast. Arm, Qualcomm, and Nvidia have all announced new CPU designs aimed at agentic AI workloads. Both AMD's data-centre momentum and Intel's recent $15 billion share sale to fund new chip factories, stories we've covered closely this August, now look like early chapters of this same supply squeeze.
| Company | Move | When |
|---|---|---|
| Amazon Web Services | Rationed CPU capacity for engineers | Early 2025 |
| Intel | Server CPUs sold out | Through end of 2025 |
| AMD | Doubled server CPU forecast | 2025 |
| Arm / Qualcomm | New CPUs targeting AI agents | Announced 2025 |
| Nvidia | Prioritised Vera, its own Arm-based CPU | Part of Vera Rubin platform |
Kimball warns the crunch may reach ordinary consumers. Intel has reportedly shifted production toward server chips at the expense of desktop and laptop lines. If demand keeps climbing, expect the same script that played out with GPUs: tighter supply, higher prices.
Chung is blunt. "The average sequence length will grow and grow, so I'm expecting this problem to get worse in future workloads."
The real story here isn't that CPUs are suddenly glamorous. It's that enterprises scaling agentic AI have a new capacity risk they probably haven't budgeted for, and the chip industry is only just catching up.
Common questions
Does this affect me if I just use ChatGPT or Claude?
Not directly today, but if CPU shortages push up cloud costs, AI services could get more expensive or slower over time. For now, nothing changes for everyday users.
Is a CPU the same as a GPU? Why do both matter?
No. A GPU handles thousands of simple maths operations at once, which is ideal for running AI models. A CPU handles fewer tasks but can juggle many different types of work, which is what AI agents need when they browse, write files, or run code.
Should businesses building on AI agents worry about this now?
Factor CPU capacity into any agentic AI project now. If you rely on a cloud provider, check whether your plan includes guaranteed CPU resources, especially for workflows where agents call many tools in sequence.
