It’s been almost a year since Microsoft and Qualcomm changed the game in the laptop market with the launch of the Snapdragon X Elite chip at Computex 2024, taking the leap into a bright new world of ‘AI PCs’, followed swiftly by rivals Intel and AMD with their own shiny new laptop processors equipped with dedicated Neural Processing Units (NPUs). And in that year… well, nothing’s really changed, has it?
Intel recently confirmed this, reporting in an earnings call that demand for laptops with new AI-capable chips has been low, with the company in fact now facing a shortage of older chips due to unexpectedly high demand for laptops with 12th- to 14th-gen Intel Core i-series processors.
There are a lot of factors here, of course. Older laptops are cheaper, and in a time of such economic uncertainty (thanks for that, Trump tariffs) it’s no wonder that people would opt for a more affordable device, especially if they just need something functional for work or school. You certainly won’t find any laptops sporting Intel Core Ultra or AMD Ryzen AI Max chips in our best cheap laptops list.
There’s also the branding issue, which is a problem specific to Intel – and, quite frankly, a problem of Intel’s own making. It was revealed way back in 2023 that Team Blue would be dropping its iconic ‘i’ branding in favor of the new ‘Core Ultra’ processor nomenclature, a marketing move so wildly idiotic that it still baffles me to this day. Companies dream of building a brand as rock-solid as ‘Intel Core i7’, and you ruined it! In the words of Charlton Heston at the end of Planet of the Apes: You maniacs! You blew it up!
The AI Mystery
Ahem. Anyway. There are a few other issues surrounding the slow adoption of Intel’s AI PCs (and, one would assume, similar NPU-boosted laptops powered by chips from AMD and Qualcomm), but there’s one particular area I’d like to focus on. It’s a simple problem, but one with no clear solution: what exactly do these fancy new AI PCs do?
For the uninitiated, the main draw behind the AI PC ‘revolution’ is the presence of an integrated NPU: a dedicated processing unit for handling AI workloads. These NPUs were supposed to be the game-changer that catapulted on-device artificial intelligence – chiefly Microsoft’s AI helper for Windows, Copilot – but it’s clear to see that consumer interest in such things is still low.
Now, I’ve sat through more keynotes and briefings about the merits of Copilot and on-device AI than I can count (seriously, they’re all the bloody same), so I get the benefits. Running AI workloads – especially ones you might be giving personal information to – on your own device brings with it the advantages of better security, faster performance, and offline functionality. These are all good things.
But there’s no killer app; Copilot still isn’t very popular (hell, Microsoft is still struggling to get people to upgrade to Windows 11) and other common on-device AI programs – like photo and video editing software and tools for producing your own AI models, like LM Studio – are better suited to a system with a dedicated Nvidia GPU. Serious AI users are aware of this, too; Nvidia’s recent focus on AI hardware for both consumer and business use is clear evidence of this.
Plus Ultra!
It’s a bit frustrating, because these are good laptops. The new Dell XPS 13 with Intel Core Ultra is quite simply one of the best laptops I’ve reviewed in years – but not because of the AI features. Yes, Copilot works. It’s fine. The XPS 13’s real selling points are the premium design, fast overall performance, and phenomenal battery life. The ‘Copilot+ PC’ nonsense is an added bonus most people will ignore, nothing more.
In fact, the most popular AI tools don’t benefit from having an NPU in your computer at all; stuff like ChatGPT and Google Gemini run entirely in the cloud, with zero need for local processing. I imagine there are some poor souls out there who rushed to buy an ‘AI PC’ only to find that their Character.ai ‘girlfriend’ doesn’t reply to their attempts at flirting any faster.
Sure, part of this whole debacle simply comes down to marketing: consumers just aren’t really clear on the benefits of having an NPU, or what Copilot is good for, or whether Copilot is meant for personal or corporate users, or… the list goes on. Marrying some of Copilot’s functionality to the ever-despised Bing was another bum note from Microsoft.
With all of this in mind, it’s hardly a surprise that shoppers are feeling more keen on slightly older Intel laptops without any special AI powers. Intel’s earnings call discussion seemed to suggest that Team Blue might be reconsidering its efforts to remain on the cutting edge of processor tech, and given the extremely rough time Intel has been having lately, perhaps a new approach – one that champions simpler, more affordable tech – is the way to go?
Leave a Comment
Your email address will not be published. Required fields are marked *