Chipmaker AMD said the personal computer is entering a new phase as artificial intelligence (AI) shifts from simple assistants to autonomous “agents” that can perform tasks on behalf of users, creating demand for a new category of machines called “agent computers.”
“For 40 years, the personal computer has been the most important tool in human hands,” Alexey Navolokin, general manager, APAC, AMD, said in a blog post. “It gave billions of people access to information and redefined how we work and create. Now it is evolving again.”
The shift is being driven by rapid advances in AI agents, which go beyond chatbots by executing multi-step tasks such as research, writing, planning, and analysis. These systems can move across apps and datasets with minimal human input, operating continuously rather than responding to one prompt at a time.
“Today’s AI agents can research, write, summarize, analyze, plan, and execute,” Navolokin said. “They don’t just answer questions. They take action.”
AMD said this evolution requires significantly more computing power than traditional PCs can provide, particularly for users who want AI workloads to run locally instead of relying on cloud services. This has led to the concept of an “agent computer,” a device designed to run AI agents full-time.
Unlike conventional PCs where users manually open applications and perform tasks step by step, agent computers are built for delegation. Users assign tasks through messaging platforms or work tools, and AI agents carry them out independently, often in parallel.
“You do not operate it like a PC. You delegate to it,” Navolokin said, describing a setup where a user can send instructions via messaging apps and receive completed outputs without direct interaction.
The implications could be significant across industries. For professionals, agent computers may increase productivity by automating repetitive tasks and preparing reports or summaries in advance. For creators, the technology could reduce time spent on logistics, allowing more focus on original work. Developers can use these systems as local environments to build and test AI-driven applications.
“It is not science fiction. It is happening now in real time,” Navolokin said.
AMD is making its latest processors, including the AMD Ryzen AI Max+ 395, as key enablers of this shift. These chips are designed to handle demanding AI workloads with high-bandwidth memory and efficient parallel processing, making them suitable for running multiple AI agents simultaneously.
The company also highlighted platforms such as Ryzen AI Max+ systems and desktop configurations like Framework Desktop as capable of delivering strong performance for agent-based computing, particularly in terms of cost efficiency.
A key part of AMD’s strategy is local AI processing, which it says addresses growing concerns around data privacy, cost, and reliability. Running AI models on-device allows users and businesses to maintain control over sensitive information while avoiding ongoing cloud fees.
“People want control over their data and AI they can use every day without limits,” Navolokin said.
As AI adoption accelerates, AMD expects demand to grow among consumers, startups, and small and medium enterprises seeking affordable, always-on AI capabilities.
“The PC remains the machine you use,” Navolokin said. “The agent computer becomes the machine that works for you.”