A report by the research firm, eMarketer, released on Tuesday, revealed that despite AI-capable computers being introduced in the middle second quarter, 14% of worldwide shipments were made up of them. Canalys.
Canalys reports that 8,8 million AI-capable computers were shipped in the period. 40% of these PCs ran Microsoft Windows, and 60% macOS.
Canalys defines AI-capable PCs as desktops or notebooks that include a chipet or block dedicated to AI workloads such as a Neural Processing Unit (NPU).
The report stated that with all major processor vendors AI-capable PCs roadmaps well underway, it is now set for an increase in device availability, and adoption by end users, in the second half and beyond of 2024.
Ishan dutt, Canalys’ Principal Analyst, stated in a recent statement that the second quarter 2024 saw a significant increase in AI-capable PCs.
He continued, “June saw a launch of Copilot+ computers incorporating Qualcomm Snapdragon X chips based on Arm Architecture.” While shipment volumes were small in the first quarter due to limited weeks and geographic coverage, the wide commitment of Windows OEMs towards adopting these products into their product portfolios bodes very well for the category.
He added that “AI-capable PCs are now poised to gain more traction in 2024’s second half.”
AI-PC market growth driven by supply
Dutt explains that OEMs and processor vendors are targeting a wider customer base through the availability of new product categories at more price points. Meanwhile, channel partners are signaling their preference for AI features in PCs. In a poll conducted in May, Dutt noted that nearly 60% of respondents said they expected customers to prefer devices that had a Copilot button.
Canalys estimates that the market for AI-capable computers will reach 44 million units worldwide in 2024. This figure is expected to rise to 103 million in 2025.
Mikako Ktagawa is a director and analyst at GartnerAccording to, an advisory and research company located in Stamford (Conn.), OEMs have demonstrated a strong commitment towards AI-capable computers. She did admit to TechNewsWorld that “the AI PC trend is not driven by demand, but rather supply.”
The rise of AI-capable computers is the hardware industry’s attempt to solve the “chicken-and-egg” problem by offering a large installed-base for application developers. Dev Nag, founder and CEO of QueryPalChatbots are a new way to communicate with your business.
“There aren’t a ton of mainstream applications that need NPUs yet — as opposed to, say, GPUs — but there are a number of compelling possibilities that these vendors are hoping to kickstart,” he told TechNewsWorld.
Nag listed a variety of possibilities including real-time local video enhancements, advanced photo editing, language translation in real time, and productivity improvements like intelligence document summaries, and advanced searches for local files.
The Shortage of Applications Slows Adoption
Nag stated that “OEMs remain cautiously optimistic.” “They are interested in a wide upgrade wave. They also want to remain strategically aligned with Microsoft.”
He added that “they’re naturally focused to integrate NPUs in the high-end business and consumer markets before the broader market.”
The Canalys report revealed that AI-capable computer penetration in the Windows PC market above US$800 doubled between Q1 and Q2, from 7% to 14%.
Nag predicted that “we’ll see a gradual rise in AI PC market shares to the mid-to-high 20% level next, but the absence of applications will slow down adoption.”
“If the application market starts to move towards taking advantage of this hardware,” he continued, “we could see the first ‘killer apps’ driving faster adoption in 2025.”
The enterprise will probably be the largest contributor to early adoption. “Large enterprise will probably be the first segment to adopt AI computers going forward,” Kitagawa explained. “Large businesses prefer this way, making AI computers more useful.”
Consumers confused and dazed
Although the enthusiasm of vendors towards AI-capable computer systems is high, that has not been true for consumers. “Sofar, consumer reaction has been tepid,” according to Andrew Hewitt. He is a principal analyst at Forrester.
He cited Forrester’s recent study, conducted in the second quarter of this year, which found that over half of U.S. customers do not know what AI PCs are.
“There’s still considerable confusion” when it comes down to understanding the differences between the AI PC and the traditional PC, he said. This is, in my opinion, because consumers are not familiar with the differences between a GPU, CPU and the newly-available neural processing unit. [NPU], which is one — but not the only — hallmark of an AI PC.
“The average consumer is likely to also associate generative AI software like ChatGPT or Copilot with AI PCs, rather than an upgrade in hardware.”
Nag stated that many people buying Apple laptops don’t know their purchase is AI-capable, thanks to Apple Neural Engine. They also may not consider app compatibility when making a selection.
“Microsoft’s introduction of the Copilot Key earlier this is year was not something the markets clamor for yet, but a attempt to lead the marketplace towards a Vision,” he said.
He said that some early adopters are sceptical of potential gatekeeping around AI dedicated NPUs since many GPUs have the same or even greater AI compute performance. Nag added that outside of the early adopters’ group, “the reaction is less confusion than it is inattention.” Apps are more likely to drive user awareness than hardware.
More marketing hype than functionality
Hewitt agrees that AI PCs are most likely to be adopted by the enterprise. He said that the opportunity for mass adoption will not be available until 2025, when many companies are preparing to migrate from Windows 10 to 11, in order to meet Windows 10’s end-of life deadline.
He added that “according to our report we see many of these uses cases surrounding personas such as creatives, data scientists, and developers who will benefit the most from apps which can take advantage of NPU.”
HP Newquist Executive Director of The Relayer Group, a science and technology exhibition company in New York City and author of “The Brain Makers: Genius, Ego & Greed In The Quest For Machines That Think,” argues that, at this stage, the AI PC is a marketing tactic and little more.
He told TechNewsWorld that manufacturers are preloading PCs with AI apps, which are typically built using large language models such as OpenAI’s ChatGPT. In some cases, extra processing power was added, similar to what you would find on gaming PCs.
Copilot is a part of Windows and is therefore a Microsoft offering. However, it is available online to anyone using Microsoft, so this feature isn’t limited to AI PCs.
He continued: “At this stage, the inclusion of high-powered AI processing units on these machines – which Microsoft calls neural processors – is more of a speed feature than an actual AI improvement.” “Until more AI apps are able to be run locally — and not rely on massive amounts of cloud power and storage — the AI processors won’t be utilized to their fullest potential. Apps will still rely on the Internet for much of their computing.”
“At the current stage, building PCs with this kind of processing power would be like building eight-lane highways for mopeds,” said he. “The highway has been built, but mopeds cannot tear up the roads as they are not designed to travel at that speed.”
