The CPU Comeback Is Upon Us

What happened

Earlier this year, leaders at Amazon Web Services delivered a new mandate to their engineers: they need to conserve CPU cycles at all costs. AWS has reportedly experienced an explosion in wait times for CPU server capacity as AI workloads strain the company’s cloud infrastructure. The issue seemingly took AWS off-guard, and for good reason. The AI boom led to a surge in demand for GPUs and, later, memory.

OpenAI’s inadvertent hack of Hugging Face saw its model fire off as many as 300 actions an hour, and a single agent can spawn sub-agents that make tool calls of their own. And there’s one more important complication that may increase the workload on a CPU as models become more complex: safety guardrails.

Safety and policy checks on an agent’s actions are often specific rules that inspect syntax and log files, Kundu says. Guardrails may also use small models (under a billion parameters) to analyze task complexity or intent. Though they could be executed on a GPU, they often aren’t, because their small size and the need to minimize latency keeps the work on the CPU. Source: Euijun Chung, Yuxiao Jia, et al.

Euijun Chung, a PhD student at the Georgia Institute of Technology, recently co-authored another paper, with findings that complement Kundu’s work. Chung and his co-authors found that when a server has too few CPU cores, it falls behind on dispatching work to the GPUs. That causes the GPUs to stall as they wait for instructions. In addition to that, the paper touches on another key element of LLM workloads: tokenization.

Tokenization is a key first step in LLM inference. It converts text into integer token IDs that can be processed by the model. Unlike the matrix math required for most LLM inference, tokenization is branchy, data-dependent sequential string manipulation. Though it can be parallelized by chunking text, it’s not massively parallel in the same way as the bulk of LLM inference is.

Why it matters

CPUs were mostly left out of the story, as their relative lack of parallelization made them a poor fit for AI model inference, the process of running and serving large language models (LLM) to users. But the rise of agentic AI systems, which allow AI models to operate autonomously and call on sub-agents, is changing the narrative.

Matt Kimball, vice president and principal datacenter analyst at Moor Insights & Strategy, says 2026 has brought a spike in CPU demand, much of it due to agentic AI. “It’s one thing to have this agentic workload, and let’s say, it spawns 100 agents. If I’m going to roll this out across my enterprise, those 100 become tens of thousands, hundreds of thousands, or millions of agents,” says Kimball.

” Kimball’s comments refer in part to “tool use,” which is shorthand for an LLM’s ability to access the internet, open files on a desktop, and generally use a variety of software to accomplish its task. LLMs trained for tool use learn how to call on other software.

While the LLM’s inference is still primarily executed on a GPU or similar AI accelerator, the tool calls the LLM makes are typically pushed to the CPU. “Many components of an agentic AI task are inherently CPU based jobs,” explains Souvik Kundu, senior staff research scientist at Intel.

” An LLM tasked with programming software, for example, will likely make tool calls to write code to files, move or replace files, download required packages, and build the software once the LLM believes it’s complete. Kundu co-authored a paper on agentic AI optimization alongside researchers from the Georgia Institute of Technology.

They found the CPU is often idle while LLM inference is executed on a GPU and that, conversely, the GPU is often idle when tool calls are executed on the CPU. 8x under sustained load. It’s a start, but the gains chase a moving target. Agentic systems generate work at machine speed and multiply it as they go.

What to watch

Tokenization of small prompts is a relatively trivial task and won’t tax even an entry-level CPU. However, an agentic model that makes tool calls must parse and tokenize the results of the call. “If you have an ongoing sequence of, say, 100,000 tokens, and you have a tool result of a 1,000 tokens, the tokenizer will have to tokenize the whole sequence again.

And you have to do tokenization at every agentic tool call,” says Chung. This both increases the frequency of tokenization and increases the number of tokens involved. It’s conceivable that future tokenizers will find ways to mitigate this, Chung says, but it remains a problem for modern LLM inference.