Apple launches new M6 and M5 Ultra Apple silicon chips — debuting in new Mac Mini and Mac Studio

News
By
Last updated

A trio of Apple announcements

M6 M5 Ultra
(Image credit: Apple)
Recent updates

August 25, 11:35 a.m. ET: This story was updated with more details from Apple regarding the M6's Dual 16-core neural engine and M5 Ultra's 32-core neural engine.

Apple has today unveiled its next-generation M6 Apple silicon chip, as well as a powerful M5 Ultra. The new chips will debut in new versions of its Mac Mini and Mac Studio, respectively, available to pre-order from today.

The M6 is Apple's first chip on a 2-nanometer process, utilizing a 12-core CPU with two super cores, four performance cores, and six efficiency cores. The chip also features a 12-core GPU and a 12-core GPU featuring neural accelerators, alongside a "Dual 16-core Neural Engine." The top version of the chip has 170GB/s of memory bandwidth.

That dual 16-core neural engine means that there are indeed two neural engines. There's a connection between the engines, allowing them to run one model quickly across both, or two separate models independently.

The M5 Ultra is Apple's first chip using its UltraFusion technology to form a quad-die architecture (The M5 Max chips it is connecting both used two dies). The chip goes up to 36 cores on the CPU and up to 80 cores on the GPU, with 1.2TB/S of unified memory bandwidth, which Apple says is 50% higher than the M3 Ultra. Apple also claims the 36-core CPU with 12 super cores and 25 performance cores will offer "up to 1.25x higher single-threaded performance and up to 1.3x higher multithreaded performance than M3 Ultra."

The M5 Ultra has a 32-core neural engine, but this is different from the dual 16-core neural engine on the M6. Because the M5 Ultra uses a pair of M5 Max chips over UltraFusion, the 32 cores across the dies to run two models in parallel for increased performance.

Both M6 and M5 Ultra feature GPUs with neural accelerators and their fastest cores, marking a leap forward for Apple's graphics processors when it comes to AI.

Apple users who have been maxxing out previous versions of the Mac Mini will be capped at 32GB of memory on the Mac Mini with M6, like a choice the company made due to the component shortage that has been affecting the entire industry. If you want 64GB like the M4 Pro, you'll need the M5 Pro model. But the M5 Ultra in the Mac Studio won't have that problem, with Apple offering up to 512GB of unified memory on that chip.

Latest Videos FromTom's Hardware

M6 also features updates to hardware-accelerated ray tracing, dynamic caching, and the shader core, which the company claims will allow for faster rendering and higher frame rates in games.

Mac Mini and Mac Studio

The new Mac Mini and Mac Studio aren't seeing any design changes, though there are substantial internal changes. Both systems are available for pre-order today, but won't arrive to customers and in retail stores until September 22.

Mac Mini with M6 running XCode.

(Image credit: Apple)

Besides the M6 version of the Mac Mini, there will also be versions with M5 Pro. Both models will be bumped to Wi-FI 7 and Bluetooth 6, as well as 2.5Gb Ethernet (up from 1Gb), with a 10Gb option available as an upgrade. The Mac Mini also supports genlock via USB-C, synchronizing displays and cameras to prevent video drift during broadcasts.

For those working with AI, Thunderbolt 5 on the M5 Pro Mac Mini will allow customers to run clusters for on-device models.

The Mac Mini with M6 will start at $899, while the M5 Pro option will begin at $1,699 (each drops $100 for education pricing). Given that the Mac Mini with M4 started at $599, it's seeing substantial price hikes here.

Mac Studio running MATLAB and LM Studio

(Image credit: Apple)

The Mac Studio will come with M5 Max or the new M5 Ultra chip, and Apple is positioning it for on-device AI. It will now offer up to six Thunderbolt 5 ports with support for as many as eight displays. Like the Mac Mini, it's also getting Wi-Fi 7 and Bluetooth 6 with Apple's N1 chip, which debuted in the iPhone 17 lineup. Like the Mac Mini, the Mac Studio also supports genlock for perfectly synced video.

The Mac Studio with M5 Max will start at $2,499 ($2,299 with an education discount), and the M4 Ultra model will start at an eye-watering $5,499 ($5,099 with an education discount).

Both systems will be compatible with macOS 27 Golden Gate, including Siri AI and refinements to the Liquid Glass design introduced in macOS 26 Tahoe.

Google Preferred Source

Follow Tom's Hardware on Google News, or add us as a preferred source, to get our latest news, analysis, & reviews in your feeds.

TOPICS
Andrew E. Freedman
Andrew E. Freedman
Senior Editor

Andrew E. Freedman is a senior editor at Tom's Hardware focusing on laptops, desktops and gaming. He also keeps up with the latest news. A lover of all things gaming and tech, his previous work has shown up in Tom's Guide, Laptop Mag, Kotaku, PCMag and Complex, among others. Follow him on Threads @FreedmanAE and BlueSky @andrewfreedman.net. You can send him tips on Signal: andrewfreedman.01

11 Comments Comment from the forums
  • usertests
    12-core CPU with two super cores, four performance cores, and six efficiency cores
    Previous had 4 "super", 6 "efficiency". Are M6 "performance" cores individually faster than M5 "super" cores?

    They have a bad habit of comparing their latest chips to the chips from 2 generations ago. At least warranted for Mac Mini since they skipped M5.

    The top version of the chip has 170GB/s of memory bandwidth.
    All M5 models had 153.6 GB/s using LPDDR5X at 9600 MT/s. Are they using LPDDR5X-10625 for some M6 models and a little slower for others, or what?

    2027 Macs to Get AI-Focused M7 Chips as Apple Skips High-End M6

    M6 - Late 2026
    M7 - First half of 2027
    M7 Pro - End of 2027
    M7 Max - End of 2027
    M7 Ultra - 2028

    If Mac Mini is getting M7 soon (and it's not debuting in iPad or something instead), maybe the strategy is to have the M6 as the "cost-sensitive" version (starting at $899, or $799 EDU). Which can be sold concurrently with an M7 version using more expensive LPDDR6. That's rumored to raise the base version to 240 GB/s bandwidth (what is that, LPDDR6-10000 on 192-bit?).
    Reply
  • Doomsday7
    I am not sure about it. Even 64GB unified RAM do mean safety operating LLMs with 36B only. So as long as Apple does not find itself a way out of RAM crisis… Stay with my old Macbook Pro is driving me nuts. Where is that new RAM reduction as Purdue Univeristy is proposing to half LLM memory need and increase speed significantly? This is what Apple formerly was known for: Technology lead. Maybe M8?
    Reply
  • User of Computers
    They make no claims about M6's single-core performance compared to its predecessor, leading me to believe that there are only minor architectural improvements to be had. In that case, it strikes me as odd that Apple would reduce the number of S-cores from M5 to make room for their new middle P-cores.
    If I had to guess, it probably has something to do with multi-threaded effiency.
    Reply
  • alan.campbell99
    I was curious about the mentioning of genlock support. It's not something I personally dabbled with but I remember way back when it was a capability that you could get using an Amiga so surely this support isn't necessarily new?
    Reply
  • bit_user
    alan.campbell99 said:
    I was curious about the mentioning of genlock support. It's not something I personally dabbled with but I remember way back when it was a capability that you could get using an Amiga so surely this support isn't necessarily new?
    Back in the Amiga days, you'd use it to perform realtime overlay of graphics with video and not lose sync with audio or need to do full-frame buffering.

    These days, I guess the main point of genlock is probably for video walls.
    Reply
  • usertests
    Doomsday7 said:
    I am not sure about it. Even 64GB unified RAM do mean safety operating LLMs with 36B only. So as long as Apple does not find itself a way out of RAM crisis… Stay with my old Macbook Pro is driving me nuts. Where is that new RAM reduction as Purdue Univeristy is proposing to half LLM memory need and increase speed significantly? This is what Apple formerly was known for: Technology lead. Maybe M8?
    At least you can cluster them with Thunderbolt 5.

    Got a link for Purdue proposal?
    User of Computers said:
    They make no claims about M6's single-core performance compared to its predecessor, leading me to believe that there are only minor architectural improvements to be had. In that case, it strikes me as odd that Apple would reduce the number of S-cores from M5 to make room for their new middle P-cores.
    If I had to guess, it probably has something to do with multi-threaded effiency.
    If the middle cores are not significantly slower than the best/super cores, there may be no real loss to single-threading across 4 cores (amount of "super" cores M5 has) accounting for any small improvements they made. I don't know enough about Apple's core types to say.

    Either way, I'm guessing it would be rare for the M6 to underperform vs. the M5. The MT boost will be noticeable, while a small ST deficit on two cores would be hidden.

    My theory based on the rumored fast M6 and M7 release schedule, and that there will be no M6 Max/Ultra, is that this is a quick refinement of the M5 with LPDDR5X for "budget" products. M7-based products will be released shortly with substantial boosts to AI performance, but using more expensive LPDDR6 memory.
    Reply
  • bit_user
    usertests said:
    At least you can cluster them with Thunderbolt 5.
    Eh, multi-GPU has definite bottlenecking issues and limitations. You'd at least want to use PCIe 5.0 x16, which is way faster than TB5.

    Benchmarks of Intel's BattleMatrix solution showed it could deliver gains on batch processing workloads, but that would primarily interest service providers and not most end users.
    Reply
  • JamesJones44
    User of Computers said:
    They make no claims about M6's single-core performance compared to its predecessor, leading me to believe that there are only minor architectural improvements to be had. In that case, it strikes me as odd that Apple would reduce the number of S-cores from M5 to make room for their new middle P-cores.
    If I had to guess, it probably has something to do with multi-threaded effiency.
    I've seen 6% quoted by a lot of people and they reference https://nanoreview.net/en/cpu-compare/apple-m6-12-core-vs-apple-m5. However, I don't know if that is an official run or "estimated", I've so far taken it with a grain of salt.

    Like you, I'm guessing the IPC improvements are marginal to nonexistent. It could largely die shrink with small tweaks to accommodate two additional cores and the change of two LP cores to E/Middle cores (Apple calls them "performance cores")
    Reply
  • Doomsday7
    Before Xiaomi is implementing this! https://arxiv.org/html/2606.00365v1
    Reply
  • User of Computers
    usertests said:
    My theory based on the rumored fast M6 and M7 release schedule, and that there will be no M6 Max/Ultra, is that this is a quick refinement of the M5 with LPDDR5X for "budget" products. M7-based products will be released shortly with substantial boosts to AI performance, but using more expensive LPDDR6 memory.
    I definitely agree. There shouldn't be anything noticible from consumers' or reviewers' perspectives.
    Reply