Analysis: Apple Gets Loud About AI PCs With New M3-Based MacBook Air

Apple calls the new MacBook Air the ‘best consumer laptop for AI,’ a jab against fresh AI PCs from OEM rivals as well as the AMD and Intel CPUs inside them. It also says that ‘every Mac’ with an M-series chip ‘is a great platform for AI.’

Apple is getting louder and more assertive about the AI capabilities of its PCs, calling the newly revealed M3-powered MacBook Air the “best consumer laptop for AI” on top of saying that “every Mac” with an M-series chip “is a great platform for AI.”

Unveiled Monday, the new MacBook Air comes in 13-inch and 15-inch sizes, and they both use the M3, the third generation of Apple’s base-level, Arm-based M-series system-on-chips that debuted in the MacBook Pro and iMac last fall.

[Related: Analysis: 5 Takeaways On The State Of AI PCs And Why Big Channel Vendors Are Excited]

The tech giant is promoting the new MacBook Air, expected to start shipping this week, as a significant upgrade over 2020’s M1- and Intel-based models, with the M3 enabling performance gains up to 60 percent over the former and as many as 13 times over the latter.

But what was most notable about Apple’s announcement was how it brought the AI capabilities of its Mac computers to the forefront. It happened as vendors in the Windows PC ecosystem have started to hype up the emerging category of AI PCs and release new client devices with advanced AI capabilities in an attempt to rekindle the ailing market.

With the statement that the new MacBook Air—which starts at $1,099—is the “best consumer laptop for AI,” Apple is taking a jab at recently revealed AI-focused laptops from rivals, such as Dell Technologies’ refreshed XPS line, Lenovo’s IdeaPad Pro 5i and HP Inc.’s Spectre x360 2-in-1s.

What helped give Apple the confidence to apply this superlative is what it says is the M3’s “faster and more efficient” 16-core Neural Engine, which sits alongside the CPU and GPU on the processor’s die to accelerate machine learning workloads on the device.

The Neural Engine has been in Apple’s M-series chips since they debuted in Macs back in 2020. This has put the company years ahead of Intel and AMD, which only debuted client processors with a neural processing unit (NPU)—designed to target similar workloads—at the beginning of last year and end of 2023, respectively, to boost the AI PC concept.

With the Neural Engine, Apple said, macOS is given the ability to deliver “intelligent features that enhance productivity and creativity” such as “powerful camera features, real-time speech to text, translation, text predictions, visual understanding [and] accessibility features.”

The company also highlighted how apps by ISVs are enabling advanced AI features. It cited as examples AI Math Assistance in Goodnotes 6, automatic photo enhancement in Pixelmator Pro and background noise removal from video in CapCut.

But it’s not just the NPU that Apple sees as an important element for AI workloads. The tech giant also pointed to how the unified memory architecture of its M-series chips helps the new MacBook Air and other Macs “run optimized AI models, including large language models and diffusion models for image generation locally with great performance.”

This is an area that has been of critical importance to vendors in the Windows PC ecosystem since Apple’s competitors are hoping to enable generative AI capabilities that were born in the cloud but are now being optimized to run on computer’s processor.

Apple even made time to mention how the MacBook Air, like other Macs, can run cloud-based AI solutions such as Copilot for Microsoft 365, Canva and Adobe Firefly.

All these things, in Apple’s view, lend to Macs providing what it called a “great platform for AI.”

None of these features are new with the latest MacBook Air. It certainly isn’t news to Mac heads who have been testing large language models, trying out new features like the ones listed above and pushing the limits of their devices. Ironically, the “killer app” for AI PCs that Intel cited last year to promote its now-released Core Ultra chips—Rewind—debuted on Macs in 2022 and still isn’t available yet on Windows devices.

But despite Apple’s AI PC bona fides, Monday’s MacBook Air announcement represents the most the company has ever said about the broader range of AI capabilities possible with its family of Macs. For instance, when Apple launched new Macs last fall with its M3, M3 Pro and M3 Max chips, the company mainly focused on AI features for content creation software and the benefits of the M3 Max’s 128 GB of maximum unified memory for AI developers to run large transformer-based models, which includes the many large language models in use today.

With Apple CEO Tim Cook promising in February to reveal what he called “incredibly” exciting generative AI features later this year, he made clear that he views AI as a “huge opportunity,” even if the company has been more reserved than rivals in talking about the space.

After all, as Cook said when pressed on Apple’s generative AI strategy last month, it’s more important for the tech giant to show than tell: “Our M.O., if you will, has always been to do work and then talk about work and not to get out in front of ourselves.”