What Is An AI PC And Why Is Everyone Talking About Them?
AI PCs have become one of the hottest buzzwords in the tech world. In simple terms, when companies talk about AI PCs they usually mean computers that include a dedicated NPU, or neural processing unit, designed to handle AI workloads locally on your machine rather than entirely in the cloud.
For PC gamers and everyday users, that means new chips from Intel, AMD and others are shipping with extra hardware blocks specifically tuned for tasks like AI assisted video effects, background noise removal, image generation, and features like Microsoft Copilot that run partly on your own system.
At a recent Intel briefing, Intel fellow Tom Petersen shared just how quickly these AI ready systems have spread. According to Intel, the number of AI PCs in the world has already climbed to slightly more than 100 million, powered by recent generations of Intel processors that include NPUs.
This rapid growth has happened over only a couple of years, kicking off with Intel Meteor Lake chips and continuing with the newer Lunar Lake generation. Thanks to Intel’s huge presence in the laptop and desktop market, a big chunk of new PCs you see on store shelves now qualify as AI PCs whether buyers specifically asked for AI features or not.
From TOPS To ZOPS: Measuring AI Performance At Scale
When chip makers talk about AI performance they love to use the term TOPS which stands for trillions of operations per second. It is a rough way of describing how many AI related math operations a chip can crunch in one second.
Tom Petersen went a step further and tried to express the total AI horsepower that is now installed across all those 100 million plus AI PCs. He joked that the scale has got so large that we should not talk about TOPS any more but instead talk about ZOPS.
A ZOP in his words is a Zetta OP, an almost absurdly huge number of operations equal to a billion TOPS. Petersen claims that the installed base of Intel powered AI PCs now adds up to around four ZOPS of AI capability out in the wild.
It is a fun bit of marketing maths, but it does highlight a real trend. There is now a massive amount of AI specific hardware sitting inside consumer PCs. The big question is whether that hardware is actually being used for anything meaningful yet.
Are People Actually Using AI Features On Their PCs?
Even Intel is not sure how many of those AI PCs are truly being used for AI workflows. Petersen was clear that Intel does not track exactly what people run on their machines. So while we know the hardware is out there, we do not know how often the NPU is really firing up.
Petersen compared the current situation to the early days of floating point units in PCs. For a long time the FPUs that handled more complex math instructions were mostly sitting idle. Then over time software caught up, games and applications started to rely on them, and eventually they became an essential part of every CPU.
He suggested that a similar turning point for AI PCs might come from Microsoft. Specifically, he pointed to Microsoft Copilot and the more advanced Copilot Plus experiences as the potential trigger that could make all this NPU and GPU AI horsepower feel genuinely useful for mainstream users.
These Copilot features are designed to lean on local NPUs and GPUs for things like on device language models, smarter search, and context aware assistance that does not always need to ping a cloud server. As Microsoft rolls deeper integration into Windows 11 and beyond, the argument is that the AI hardware you already have in your PC will start to matter more.
Right now though, many people are still not buying new computers because of AI. Dell recently noted that consumers are not choosing systems based on AI branding. For most shoppers the decision still comes down to the usual things like price, CPU and GPU performance, battery life, storage, and build quality.
The 100 million plus figure for AI PCs probably says more about Intel’s commanding market share than a huge wave of AI focused demand from gamers or everyday users. Intel processors show up in a vast number of laptops and desktops, and PC sales generally are up compared to last year. So AI features are often just coming along for the ride when someone upgrades their machine.
What This Means For PC Gamers And Tech Fans
For the PC gaming community, AI PCs are an interesting side story to the usual focus on GPUs, frame rates, and visual quality. In the short term, the most obvious AI benefits will likely be things like enhanced background tasks and smarter tools that run in parallel to your games.
- Improved streaming experiences such as better background removal and noise suppression.
- AI assisted content creation for video editing, thumbnails and game clips.
- Potentially smarter upscaling, image enhancement, or AI driven in game assistants as developers experiment.
The core gaming performance still leans heavily on the CPU and especially the GPU, not the NPU. But as game engines and tools evolve, there is room for NPUs to pick up more of the AI side work so that GPUs can focus on rendering.
For now, Intel’s ZOPS bragging is a sign that the hardware stage is set. NPUs are here, they are shipping in huge numbers, and the installed base is already massive. The real test will be over the next few years as Windows, creative apps, and eventually games figure out how to turn all that theoretical AI power into features that PC gamers and everyday users actually care about.
Until then, many of us will keep buying PCs for the same old reasons while quietly inheriting AI capabilities with each new CPU generation. Whether you are excited about AI PCs or just mildly amused by new acronyms like ZOPS, it is clear that AI oriented hardware is going to be part of the PC landscape from here on out.
Original article and image: https://www.pcgamer.com/hardware/processors/intel-says-its-shipped-slightly-north-of-100-million-ai-pcs-which-equates-to-roughly-four-zops-or-zetta-ops-of-ai-processing-power/
