The M5 Ultra Mac Studio tears through our benchmark tests

The Mac Studio review unit that Apple sent us to test this year is, put simply, kind of outrageous. It has an M5 Ultra chip with a 36-core CPU and 80-core GPU, 256GB of RAM, and 4TB of storage and costs $12,299. This thing is not for your typical content creation workloads. It’s for AI developers and some of the most demanding 3D visual effects houses out there.

Frankly, our usual benchmarks aren’t cutting it.

The latest version of Apple’s most powerful computer begins shipping today with new chips and prices that extend even further into the stratosphere. The new Mac Studio uses the same design that’s been with us since 2022, replete with ultra-fast ports both front and back, now kitted out with either an M5 Max or M5 Ultra chip, compared to the M4 Max or M3 Ultra of last year’s model.

The M5 Max model starts at a relatively modest $2,499 with an 18-core CPU / 32-core GPU, 36GB of RAM, and 512GB of storage. We’ve tested an M5 Max already, in the MacBook Pro, where it starts at $4,099 with the same amount of memory and 2TB of storage (the equivalent Mac Studio costs $3,299). The M5 Ultra is another beast altogether. It starts at $5,499 with 30 CPU cores, 64 GPU cores, 96GB of memory and 1TB of storage.

I’ve only had my hands on the new Mac Studio as well as a pair of the latest Mac Minis for a handful of days. That’s enough time to adequately test the M6 and M5 Pro chips in the Mac Mini, and to run our standard benchmarks on the M5 Ultra — but not enough time to seriously challenge it.

Conveniently, we still have the similarly specced M3 Ultra Mac Studio from last year — though it was only $8,099 at the time. Aside from CPU core count and architecture, they’re nearly the same, which makes them easy to compare.

The M5 Ultra Mac Studio is around 30 percent faster than our similarly configured M3 Ultra in both single-core and multicore synthetic CPU tests. It varies more in graphics tests, but some, like the new Geekbench 7 OpenCL test, also saw around a 30 percent uplift. And a heftier GPU rendering test like Cinebench scored 63 percent higher.

I’ve never tested anything this fast before. The storage speeds are off the proverbial charts. It renders the classroom test in Blender faster than I can tie my shoes. Our Premiere 4K video export test, which yields an MP4 file of about 1.7GB, is done in about 40 seconds. I can’t get up from my desk and get a cup of coffee that fast without risking a serious burn.

Apple is talking a big game about the M5 Ultra being “a monumental step for AI.” I don’t normally run local AI stuff (I just don’t find generative AI and LLM stuff very useful for my work — and I’ve tried). But I want to give it a serious go, so I’m open to suggestions. I’ve already reached out to our great Verge audience for ideas, and I appreciate all of you who commented. (Yes, even those of you with the “Can it run DOOM” jokes. I see you, and it never gets old. I mean, it does. But it doesn’t. I love DOOM.)

How long is this going to take me? Hopefully not very long, but it’s going to be a busy couple of months. Google is announcing its whole new line of Googlebook laptops at exactly the same time this article publishes. And there’s a Microsoft event coming up next week that sounds a whole lot like it’ll jumpstart Nvidia’s RTX Spark initiative. There’s never a drought of laptops to test, but this Techtember and Techtober seems like a deluge that’s about to engulf Your Friendly Neighborhood Laptop Reviewer.

I know, I know; “So many cool computers to test, what a burden.” Truly, it is a very exciting time for computers. An expensive one, for sure, but I don’t think the space has ever been this dynamic. So I’ll do my best to get back to you when I’ve got something cool to share about the M5 Ultra Mac Studio. Feel free to drop more ideas and tips in the comments here. Or email me or DM me on Bluesky if you’ve got suggestions you’d like to share. I also accept words of encouragement and compliments. 🙂

The M5 Ultra Mac Studio is a powerful machine. Let’s see if we can find its limits. Stay tuned.