Intel and Nvidia are teaming up to make a new chipset. Should Apple be worried?

Macworld

On Thursday, Intel and Nvidia announced a major partnership that shook up the chip market and a number of stocks. But Apple—and Apple fans—shouldn’t start freaking out.

The deal is pretty simple. First, Nvidia will buy $5 billion worth of Intel common stock. That alone sent Intel’s stock soaring.

For the server market, Intel will build custom x86 CPUs that Nvidia will use in its AI datacenter products. Nvidia already rules the AI datacenter business and if it can eke out a little more performance or efficiency with a custom CPU, that’s good, but it hardly matters to those who use Apple products.

A report last year said Apple doesn’t use Nvidia’s products to train its AI models, preferring Google’s Tensor cloud products instead. There have been major upheavals in Apple’s AI development since then, and it wouldn’t surprise us if Nvidia was a big part of Apple’s cloud-based AI training now, but whether Apple uses training server products from one external company or another isn’t a huge deal.

Apple is developing its own “Private Cloud Compute” AI servers using Apple Silicon, but those appear to be aimed at inferencing—where the trained model runs on a user’s requests and data to generate a result for them. Apple may still rely on outside hardware for AI training, which often has different and more power-intensive requirements.

Mac vs PC

The more interesting part of the partnership is the development of new consumer chips. Intel will build and sell new x86 RTX system-on-chips that combine an Intel CPU and Nvidia RTX GPU chiplets.

Intel already sells CPUs with integrated GPUs for the thin-and-light notebook market, but the GPUs in those products fall far behind competing products from AMD and Apple. AMD’s superior performance in graphics (and in some cases CPUs as well) has made it win some market share and has dominated the new handheld gaming computer market.

With this deal, Intel/Nvidia will likely be able to sell much better products aimed at thin and light laptops and handheld gaming PCs—currently, bigger gaming laptops and of course desktop PCs are dominated by systems that use discrete Nvidia GPUs, but those are power hogs by comparison.

Where does that leave Apple? Are these chips going to out-compete Apple’s impressive M-series chips? Are people going to switch from MacBooks to Windows laptops?

Well, we don’t know how the chips will perform. There are no actual products announced, nor any of their technical specs. We don’t know when they will be released—will these first Intel/Nvidia SoCs compete with next year’s M5 series, or M6? Maybe even M7?

We’ve heard claims of superior performance to Apple’s chips before, and in the end, the products that use those chips can’t keep up, especially not without using a lot more power. Apple has some catching up to do on graphics performance, but the CPU, video, and audio encoding, as well as neural engine AI processing, remains top-tier.

You can already get a faster laptop than an equivalent-priced MacBook (especially in graphics performance). But it’s not a MacBook. It doesn’t have the same build quality, battery life, and attention to detail.

But most importantly, people buy Macs because they want to use a Mac. It’s macOS and the apps that run on it (Apple’s and others), together with easy operability with other Apple products, that make people want a Mac. Maybe Intel’s new chips will re-create the Pentium-Power PC arms race of the mid-’90s, but ultimately, while this Intel/Nvidia partnership is interesting, it doesn’t do anything to fundamentally change much.