We’ve been talking about for a while now the potential of artificial intelligence at the client, and Intel AI PC Acceleration marks the before and after in this model that can mark so many remarkable distances. However, in order to better understand its scope, it is necessary to first clarify what we mean when we talk about client-side AI and how it differs significantly from the client-server model on which the majority of AI platforms are based. that we know and use today.

When we approach Bard, ask Midjourney for a picture, or transcribe a text-to-speech recording using a technology-based service like Whisper, the only thing happening on our PC (or smartphone or whatever device we use) is this? we pass our request to the platform which is responsible for processing it, generating a response and sending it back to us. Thus, all computing tasks are performed remotely on the server where the said load arrives.

With a client-side AI computer, as you might have imagined, there is no external infrastructure available, all workloads are performed locally, on the client, that is, on the same system in which the requested function is requested. Its two main advantages are that it prevents us from depending on an external service and that it gives us much more privacy, since our data does not have to leave our system and be processed by the company’s infrastructure that can use it. for multiple purposes.

The problem is that, in general,AI computing is very resource intensive, but for some time now we’ve seen AI-based features that rely on specialized but consumer-market hardware to run on the client. The most obvious example is DLSS from NVIDIA and similar technologies that have subsequently spread. Thanks to the capabilities of NVIDIA hardware, enterprise AI-based software runs seamlessly locally.

NVIDIA has been a leader in this movement, and now Intel is joining in, providing both the hardware and the elements necessary to use it. As you may remember, a month ago Meteor Lake introduced the 14th generation Intel Core, also referred to as Core Ultra, which among its main novelties includes in its package several integrated devices dedicated, partially or completely, to computing with artificial intelligence. . And this is not just a movement in response to a trend, but part of a much larger plan, which Caitlin Anderson, Intel’s vice president and general manager of sales for the Client Computing Group, told us in an interview that must be lost on you.

Well, today we find significant progress in this sense, p Introducing Intel AI PC Accelerationa program launched by Intel for software development companies that will make it much easier to integrate AI-based features using the Intel Core Ultra platform, allowing them to get them directly to PCs in the optimal way.

Before announcing Intel AI PC Acceleration, the technology company had already collaborated with other companies and projects, among which it cites Adobe, Audacity, BlackMagic, BufferZone, CyberLink, DeepRender, MAGIX, Rewind AI, Skylum, Topaz, VideoCom, Webex, Wondershare Filmora, XSplit and Zoomwhich allowed them to add more than 300 AI-based features in total.

More information