Warehouses in the IT channel are stocking up with AI-capable PCs - industry watcher Canalys claims these made up 20 percent of all shipments during Q3 2024, amounting to some 13.3 million units worldwide.

"Shipments," of course, simply means that these devices have left the makers' factory and been delivered to distributors, rather than 13.3 million AI-capable PCs being snapped up by buyers.

Nevertheless, Canalys says this figure represents a sequential growth of 49 percent over the previous quarter, meaning that vendors are ramping up output of AI-enabled systems in the hope these will prove popular.

Whether this will be the case at present is debatable. None of the industry experts we approached were able to give us any real insight on how well these AI boxes are selling compared with plain old PCs, or whether they are getting any traction with buyers.

In one sense, this doesn't really matter, since it is likely you won't be able to buy any computer that isn't AI-enabled before long. Recent forecasts estimate that such systems may account for 43 percent of PCs by 2025, and go on to make up the bulk of the market by 2026.

However, a look at the figures shows that Canalys is counting Macs in its AI-capable data, and these account for about half (47 percent) of all shipments, meaning that about 7 million Windows AI boxes were let loose into the channel during Q3.

What is an AI-capable PC anyway? Canalys defines it as a desktop or laptop that includes some dedicated hardware for accelerating AI workloads, such as the neural processing unit (NPU) embedded in many of the most recent processors.

But then Microsoft added to the confusion by introducing its own definition of "Copilot+ PC" to describe Windows systems with an NPU that performs at 40 TOPS or more, which is one measure of a computer's AI processing prowess.

As Canalys says in its report, both Intel and AMD are still awaiting Copilot+ PC support from Microsoft for their latest CPUs, adding another swirl of uncertainty to the mix.

Those Copilot+ PC boxes are likely to see lower sales than the plain AI-capable systems, analyst Kieren Jessop at Canalys told us.

"The Windows units will ramp in scale in line with a silicon transition. That will be split between Copilot+ PCs (which have an NPU > 40 TOPS) and under 40 TOPS PCs. So far the signals are that Copilot+ PCs will have a far slower adoption rate than AI-capable Windows PCs of under 40 TOPS," he said.

However, if it is performance you are concerned about, "it's important to note that GPUs still far outperform NPUs in terms of raw performance," Jessop said, while NPUs are more power-efficient and better suited for running perpetually.

On those shipment figures, Jessop explained that many of these will form part of normal buying cycles rather than be driven by specific AI use cases. In particular, the looming end of Windows 10 support next year will be pushing many organizations into a PC refresh.

"The 'AI capability' becomes a draw to businesses typically serving as a future-proofing measure rather than a primary purchase motivator at this stage," he said. Consumers, Jessop added, are "more likely to choose an AI PC for the extended battery life (via NPU advantages) rather than the AI features at this point."

Canalys says a key challenge for vendors will be to convince customers to future-proof for a potential wave of on-device AI use cases, which is almost entirely lacking at the moment.

In other words, AI PCs currently seem like a solution in search of a problem, but vendors will be keen to push them as they carry a 10 percent to 15 percent price premium over standard PCs, as Canalys has previously noted.

Bryan Ma, IDC VP for Devices Research, agreed that the industry appears to be desperately searching for a use case for AI PCs.

"The hope is that both Microsoft as well as third-party ISVs will emerge with more solid use cases in the upcoming years, but at this point such use cases are not obvious," he told us. "In the meantime, shipments of these AI PCs are a supply-side push, in large part on the basis of these systems featuring the latest-and-greatest processors rather than them being purchased specifically for on-device AI."

So the transition to the world of AI-capable PCs seems inevitable as vendors gradually fill out their portfolio with systems based on processors with embedded NPUs. But as recent figures from Gartner show, the lack of a rebound in PC sales figures appears to indicate that buyer interest in AI-capable computers is - so far - minimal. ®