We asked Intel to define 'AI PC'. Its reply: 'Anything with our latest CPUs' If you're confused about what makes a PC an "AI PC," you're not alone. But we finally have something of an answer: if it packs a GPU, a processor that boasts a neural processing unit and can handle VNNI and Dp4a instructions, it qualifies – at least according to Robert Hallock, Intel's senior director of technical marketing. …
"There are cases where a very large LLM might require 32GB of RAM," he noted. “Everything else will fit comfortably in a 16GB system.”
Seems a bit like:
I have to say that in 1981, making those decisions, I felt like I was providing enough freedom for 10 years. That is, a move from 64k to 640k felt like something that would last a great deal of time. (Gates)
32 bits, it is enough for an experiment, it is 4.3 billion terminations -- even the defense department doesn't need 4.3 billion of anything (Cerf)
"I think there is a world market for maybe five computers." (Watson)
(and so on)
Much like second hand cars did.
Who want's this shit sprayed everywhere ? I don't.
Same as the heresy I committed when I last chose a car *and* TV. Told the sales person I didn't give a shiny shit how smart anything was because I will never use it. Instead I will use what I bring to the table through the screen(s).
That's because Intel's definition of 'AI PC' is a laptop with GPU, NPU Intel's latest Meteor Lake CPU.
The reason desktops processors are not part of Intel's definition and no Intel AI PC Raptor Lake CPUs with NPUs is NPUs are not required when you can fit a GPU with dedicated AI accelerators into a desktop. Nvidia RTX range has Tensor Cores not sure what AMD RX 7000 series has. My 2 year old gaming PC with 2 Nvidia RTX 3090 Tis is an AI PC.
Anyone that needs to run AI models on a desktop PC has no need for a NPU they will get or already have a desktop with a high end GPU with AI accelerators,.
...Intel talks about them a lot. But for the past few years most GPU's on the market have had Tensors cores built in. They are specifically designed to run the neural networks inside of AI models. So I would argue any gamer with Tensor cores in their GPU already has an AI PC. Welcome to the future.
I'm looking forward to this AI. I can just tell it go through the DVD iso's I haven't converted to MP4 and run them through Handbrake, save only the episodes (no extras), make sure that any that need subtitles are ripped properly either as a track or burnt-in ones, give them meaningful series and episode names, sort them into directories then stick them on the NAS in the right hierarchy and update my tracking file. I'm looking forward to asking the AI to "sort out my iTunes Library", including rationalizing all the "sort album artist" fields, fixing the artwork, upping the volume on the tracks that are a bit quiet and shifting the song announcements on live albums from the end of the previous track to the start of the track that it's introducting - all without changing the play count and last played date metadata. That's just a couple of things I've spent a few hours doing this week, but there are tons of things I can think of for AI to do for me. Will it be out in time for Christmas? What colour will it be?
The one thing that humans have a hard limit to is time. This affects how much information they can consume, or in your case how much music and how many videos its physically possible to listen to or watch in one lifetime. Even a super fast computer will take a noticeable amount of time to write to 32GBytes of memory.
I can see a use for very high performance GPUs in gaming environments where you're creating detailed worlds in real time (and where the detail glitches for the most part will go unnoticed). LLMs are useful but when I played around with them I found that they generated material which either took a long time to consume (read) or took me a relatively long time to utilize. (I don't get involved in Marketing so generating reams of BS in real time isn't something I need to do -- everything I do has to be verified and checked.)
The history of performance oriented computing has been for the most part applications soaking up surplus computing power, often with meaningless, if pretty, visual effects on graphical desktops. (For those without state of the art hardware its just a mess -- nothing works properly so its a constant fight to get anything done.) Adding more power is useful to sell more computer but I doubt if it will increase the quality of work, it will be just adding yet more monkeys and typewriters.
> at least according to Robert Hallock, Intel's senior director of technical marketing
But why listen to what a marketer says, about anything, ever? Their job is literally to say whatever will make product sell. If this were five years ago he'd be explaining how the latest CPUs were "blockchain-ready" and how important that is.
"Intel used the "Centrino" brand to distinguish Wi-Fi-enabled PCs"
Oh, that is what that brand was all about! Could never quite figure it out, as all the other PCs came with WiFi as well.
Must have been the wrong color of WiFi or something, same way as the cores in your GPU are clearly not as good as the NPUs that Intel are providing, so you obvioysly don't have an AI PC.
A CRM with AI to create a summary of your "relationship"? Oh boy, yes, that's what we need, for corporations to be able to quickly summarize customer relationships into easily understood text for some phone bank to call back and then "helpfully" recommend whatever costly solution fits best.
(no, we don't really need that, but it would "save time" for the marketing department without saving anyone else's time much at all)
If your job is to market a product, you have to meet one of two objectives:
Convince potential buyers that your product fits their needs better than the competitors offerings or;
Create a market by explaining your offering does something people didn't realize they really need to do.
This chap works in marketing and it looks like he's behaving like a well adjusted marketer. Good for him. ;)
“When I use a word,’ Humpty Dumpty said in rather a scornful tone, ‘it means just what I choose it to mean — neither more nor less.’
’The question is,’ said Alice, ‘whether you can make words mean so many different things.’
’The question is,’ said Humpty Dumpty, ‘which is to be master — that’s all.”― Lewis Carroll, Through the Looking Glass
I've been really intrigued by the concept of AI PC (https://www.lenovo.com/us/en/lenovoauraedition/) lately. It seems like we're at a turning point where everyday computing can be enhanced significantly by integrating AI capabilities. I think the potential for personalized user experiences is huge—imagine your computer learning your preferences and optimizing tasks for you!
However, I do have some concerns about privacy and data security. With AI processing so much personal information, it’s crucial that manufacturers prioritize these aspects. Overall, I’m excited to see where this technology goes and how it will shape our daily interactions with computers.
What does everyone else think about the balance between innovation and privacy?
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
