What Is an AI PC, and Why Would Higher Education Need One?

What Qualifies Something as an AI Computer?

Hardware differentiates an AI PC from a run-of-the-mill computer.

“AI PCs have all the traditional elements of the PC that we know today, except there is an enhancement,” says Hernan Londono, chief technology and innovation strategist for the public sector at Lenovo.

“Beyond having just memory, a hard drive and CPU, now we begin having new elements — such as an NPU, a neural processing unit, and GPU, a graphics processing unit — for processing artificial intelligence workloads,” he says.

With this more robust processing capability, an AI PC is designed “to perform tasks that require some type of intelligence — to do problem-solving, decision-making, content creation,” says Allen Bourgoyne, director of product for enterprise platforms at NVIDIA.

“PCs traditionally have been more about data retrieval: I pull up a spreadsheet, I create the document, I do all the work,” he says. With an AI PC, “it’s moving to more of a creative-type tool, rather than simply a data retrieval tool.”

RELATED: How strategic planning and expert support enhance AI PC rollouts.

What Can an AI PC Do That a Traditional Computer Cannot?

Much of AI processing today takes place in the cloud or on traditional PCs, which can handle some basic AI tasks such as tackling generative prompts. But an AI PC can do things faster and do them locally. That has big implications for higher education.

“If you have an AI-enabled PC, it’s able to handle those AI tasks in real time. You’re capturing a meeting or a lecture, and it’s processing that on the device,” said Adrienne Garber, Dell’s lead technology and innovation strategist for higher education.

“It’s not only faster, it’s also more secure, because you’re not uploading and downloading and transferring data in and out of the cloud,” Garber says.

How Can AI Computers Benefit Higher Education Institutions?

On the research side, “we’re beginning to see the researcher managing the data set closer to his or her practice” with AI PCs, Londono says. Local, real-time analysis minimizes latency and accelerates insights, helping researchers experiment and iterate faster and more effectively.

Field researchers benefit as well. When investigators are out in the field, “you might not have the connectivity that you would have in a classroom or a lab space on campus,” Garber says. With an AI PC, field researchers can take advantage of AI capabilities even while working on that disconnected edge.

There are administrative uses cases too: “In the back office, you’ve got lots of information — student information, financial information, university information,” Bourgoyne says. In the past, “to prepare a report, prepare a budget or look at students’ enrollment and class schedules, all of that was just a lot of brute-force labor.”

With the power of an AI PC, he says, “we can use AI assistance to help mine through all that data, pull out what’s important, create reports, and generate more concise and effective documentation,” all without exposing sensitive data beyond the device itself.

Teachers meanwhile can take advantage of real-time transcription, adaptive learning and personalized learning. “You are enabling student success through the use of an AI PC,” Garber says. “And because it’s all on the device itself, you have less chance of a breach, of externalized threats.”

LEARN MORE: Why colleges and universities are offering faculty development for AI use.

Is Now the Right Time for AI PCs In Higher Education?

Many schools are facing a hardware inflection point. “Windows 10 support ends in 2025, and 30% of PCs are 4 years old-plus, so they can’t even upgrade to Windows 11,” Garber says. With a move to AI PCs, “you have an opportunity now to make an investment that will allow you these additional productivity and capacity opportunities.”

As schools replace aging machines, “they should look for longevity,” Bourgoyne says. “We call it future proofing. Today, if you’re not buying systems that can execute AI to the level you need, you might want to rethink that.”

What to look for in an AI PC? Speed matters, especially when supporting big research workloads and complex administrative tasks.

“People talk about AI PCs in terms of trillions of operations per second, with entry-level AI PCs being around the 40 to 45 TOPS. That’s very well suited for your traditional teaching and learning lab,” Londono says. “On the research side of things, now we begin to talk about things like 100 to 150 TOPS.”

A good rule of thumb? “It’s never a losing proposition to buy a little bit more than you think you need,” Bourgoyne says. “I always recommend figuring out what you need, and then giving yourself a little cushion.”