A handful of laptops are coming with Intel Core Ultra “Meteor Lake” chips and AI-focused processors. These “AI” laptops promise to streamline your workflow and perform AI tasks more efficiently than notebooks not equipped with Intel’s hot new CPUs. If these machines live up to their promise, they could end up being some of the best laptops out there, based on our testing.
Meteor Lake chips are built on a 4-nanometer process, making them smaller and more powerful than previous chips like 7-nanometers. Raptor Lake 13th Generation CPU Intel Core Ultra chips feature a new component called the Neural Processing Unit (NPU) that is intended to handle AI workloads more efficiently than the CPU or GPU. This aspect is the main selling point for laptops equipped with Meteor Lake chips.
I’m still not 100% convinced about incorporating “AI” into our daily lives. It’s not just because ChatGPT thought I was dead. But I’m willing to give this technology a fair shot, especially if it actually makes my life easier. From what I’ve seen so far, these new AI laptops could be a big deal. He gives three reasons for this.
On-device AI
Intel Core Ultra chips can perform on-device generative AI tasks more efficiently than processors without NPUs. This not only speeds up text and image generation, but also improves security, as everything can be done locally without going through cloud-based programs like ChatGPT or Stable Diffusion.
During Intel’s AI Everywhere event, we saw a demo showing how quickly the Meteor Lake chips’ NPUs can generate images. When the presenter typed “spaceship leaving Earth’s orbit” into GIMP (via the AI plugin), the image was generated in less than a second. If this had been done on his cloud-based ChatGPT, the images would have taken even longer, based on my own time using OpenAI’s program.
I had a similar experience at Qualcomm’s Snapdragon Summit in October. So the company announced the AI-focused Snapdragon X Elite chip, which also has an NPU. During a demo using a laptop with an X Elite chip and one with an Intel Core i7 CPU, the machines running on Qualcomm silicon instantly produced an image of a cat with green eyes. In contrast, a notebook with an Intel processor took about 20 seconds. This is a great demonstration of how quickly NPUs can help devices generate images.
Another benefit of on-device AI is that you don’t have to pay a subscription. This isn’t a guarantee, but I’ve never heard of any manufacturers I’ve talked to that have introduced a subscription service. If this is true, you might be able to unsubscribe from services like ChatGPT and generate text and images on your machine for free.
Apps equipped with AI functions
Several applications, including GIMP, Audacity, Zoom, and OBS, are optimized to take advantage of Meteor Lake’s NPUs. Not only does this make the apps easier to use, but it can potentially save you time and money, especially if you’re a professional who uses these apps for work.
During Zoom calls, Meteor Lake chips can better handle things like dynamic noise suppression during calls. This is because the NPU can handle these tasks more efficiently than the CPU, while freeing up the CPU for other work. If you use Audacity, you can create a prompt like “Add EDM beat” to get the music track you want. GIMP also allows you to create prompts to generate images.
I use GIMP and Audacity a lot. So I’m curious to see how AI capabilities can make my work easier or provide new capabilities.
improved game
Trying to play the best PC games on a machine with an integrated GPU can be a risky proposition. However, considering how the new chip incorporates some of the same technology as the Intel Arc GPU, playing graphically demanding games on a Meteor Lake laptop might not be so scary. not. That doesn’t mean you’ll get the same gaming experience as a machine with a discrete GPU like the Nvidia GeForce RTX 40 series, but the difference in graphics and performance may not be that great for certain titles.
At Intel’s AI Everywhere event, I saw Ghostrunner 2 running on two laptops side by side. One was equipped with a 13th generation Core i7 CPU, and the other was equipped with an Intel Core Ultra 7 chip. The Intel Core Ultra laptop on the right in the photo above achieved higher FPS (frames per second) numbers. The large XeSS graphics tout Intel’s XeSS supersampling technology. The technology works like Nvidia’s DLSS to render in-game scenes at a lower resolution than the display (so you can work faster and better) and then performs machine learning. to “upscale” the scene. Match the resolution of your display.
Note that Ghostrunner 2 ran at over 60 fps with DLSS enabled. Without it, the game slowed down to about 40 fps. While this falls short of the 60 fps standard that most gamers (myself included) demand, it’s still twice his frame count on laptops with 13th generation Intel Core CPUs. Again, I don’t think these results are universal for all games running on a Meteor Lake laptop, but the fact that you can get results like this from an integrated GPU is impressive. is.