Over the past few years, the term “AI PC” has gone from sounding like a desperate attempt to revive the computer industry to something that could truly change the way we live with our personal computers. Please note that an AI PC is any system running a CPU equipped with a neural processing unit (NPU) designed specifically for artificial intelligence workloads. NPUs have been available in mobile hardware for years, but AMD is the first to introduce them to x86 PCs. Ryzen Pro 7040 chips.
Now with him Ryzen AI 300 chipsAMD is making its biggest push yet for AI-powered PCs – which could pay off in the future as we see more AI-driven features. Microsoft’s recall. (It should be noted that this also happened surrounded by privacy concerns and later slowed down.) To better understand how AMD is approaching the AI PC era, I spoke with Ryzen AI lead Rakesh Anigundi, Ryzen AI product lead and Jason Banta, CVP and GM of Customer OEM. You can listen to the full interview on the Engadget Podcast.
My most pressing question: How does AMD plan to engage developers with AI-powered features? NPUs aren’t exactly a selling point if no one is developing programs that use them. Anigundi said he understands that developers broadly “just want things to work,” so the company has built a strategy around three pillars: A strong software stack; performance hardware; and bring open source solutions.
“We don’t want to invent standards, but we have a philosophy of following standards,” Anigundi said. “That’s why we’re double-clicking ONNX, a cross-platform framework to really get the most out of our system. It ties in very closely with how we work with Microsoft, enabling their next-generation experiences and also OEMs. On the other hand, smaller ISVs Many innovations are taking place in [independent software vendors]this strategy also works very well.”
indicates Amuse 2.0 beta that AMD recently released as a way for the company to showcase the AI capabilities of its hardware. It’s a simple software for creating AI images and runs entirely on your NPU-powered device in the cloud, without the need to resort to OpenAI’s DallE or Google’s Gemini.
AMD’s Banta reiterated the need for a large toolset and software stack, but he noted that the company is also working closely with partners like Microsoft on prototype hardware to ensure the quality of the customer experience. “[Consumers] they may have all the hardware, they may have all the tools, they may have all the foundational models, but making that end customer experience great takes a lot of one-on-one time directly between us and those ISV partners.”
In this case, Banta also cites AMD’s relationship with Microsoft when it comes to building Copilot+ experiences for its systems. We saw a handful of AI features in the first batch of Qualcomm Snapdragon-powered Copilot+ machines. the new Surface Pro and Surface laptopthey are not yet available on Copilot+ systems running x86 chips from AMD and Intel.
“We make this experience perfect,” Banta said. At this point, you can consider Ryzen AI 300 machines “Copilot+ ready”, but not yet fully Copilot+ capable. (As I mentioned in my Surface Pro review, Microsoft’s current AI features are pretty basic, and that likely won’t change until Recall is officially released.)
That’s all Rumors that AMD is developing an Arm-based CPUthe company’s leaders, of course, did not reveal much. “Arm is a close partner of AMD,” said Banta. “We are working together on a number of solutions on our roadmaps… As long as [the] The general CPU roadmap, I can’t really talk about what’s around the corner.” But given the fact that the same rumor also points to NVIDIA developing its own Arm chip, and the amazing performance we’ve seen from the latest mobile chips from Apple and Qualcomm. , It wouldn’t be too surprising to see AMD follow the same Arm-paved path.
