First, I don't like making blind tradeoffs. If what I need (for whatever reason) is a really beefy ARM CPU, I'd like to know what the "Apple-less tax" costs me (if anything!)
Second, the status quo is that Apple Silicon is the undisputed king of ARM CPU performance, so it's the obvious benchmark to compare this thing against. Providing that context is just basic journalistic practice, even if just to say "but it's irrelevant because we can't use the hardware without the software".
We know how Apple's hardware performs on native workloads. We know how it performs emulating x86 workloads (and why). Surely "... and this is how this hardware measures up against the other guys trying to achieve the exact same thing" is a relevant comparison? I can't be the only person who reads "reaching desktop performance" and wonders "you mean comparable to the M1, or to the M3 Ultra?"
You're not. IMHO it's a fairly obvious, narrow and uncontroversial observation (and hence why its the top comment). That said, I personally still enjoyed the back and forth as many others one could imagine. There can be value in the counterarguments from multiple other usernames, as this facilitates sharpening reasoning for the conclusion from readers. (even when the original premise stays in tact)
The lack of others agreeing could be the result of many reasons. IMHO, a not insignificant one could be the incentive structure skews heavily towards lurking as HN rightfully disincentives "me too" type replies and not everyone always has something interesting to add
2c not an epistemologist ymmv
The cores, yes, but you can get an AmpereOne with 192 ARM cores (or rent out beefier machines from AWS and Azure). If you need to run macOS, then you are tied to Apple, but if all you want is ARM (for, say, emulated embedded hardware development), you have other options in the ARM ecosystem. I'm actually surprised Ampere maxes out at 192 cores when Intel Xeon 6+ has parts with 288 cores on a single socket (and that can go up to 4 sockets).
I wonder how many cores you'd need to make htop crash.
If your metric is single thread performance yes but on just about anything else Graviton 4 wins.
Individual benchmarks tell the bigger picture. These two are optimized for different use cases, with Apple heavily leaning towards low latency single thread throughput with low sustained power usage.
https://browser.geekbench.com/v6/cpu/compare/16833358?baseli...
EDIT: The M4 Max compares much more closely https://browser.geekbench.com/v6/cpu/compare/16834801?baseli...
https://browser.geekbench.com/v6/cpu/compare/16839304?baseli...
The M3 Ultra sacrifices a bunch of single-thread performance for not that much of a multithreaded gain:
https://browser.geekbench.com/v6/cpu/compare/16839654?baseli...
I am looking for a CPU.
I don't want to confront my users with "Please enter your Apple ID" or any other unexpected messages that I have no control over.
Is Apple M series an option for me?
For better or worse if you make a (high end) consumer CPU it will be judged against the M-series, just like if you make a high end phone it will be judged against the iPhone.
All he is saying: We currently have products in a similar product category (arm based desktop computers) that are widely used and have known benchmark scores (and general reviews) and it would make sense if I publish a new cpu for the same product category ("Reaching Desktop Performance" implies that) that I'd compare it to the known alternatives.
In the end you can just run Asahi on your macbook, the OS is not that relevant here. A comparison to macbooks running Asahi Linux would be fine.
amelius, if anyone had specific requirements, it was you with your "systems for in-flight entertainment".
OP asked a very reasonable question for a very generic comparison to the 800-pound gorilla in the consumer CPU world in general, and ARM CPU world in particular.
If the article can reference AMD's Zen 5 cores and Intel's Lion/Sunny Cove, they could have made at least a brief reference to M-series CPUs. As a reader and potential buyer of any of them, I find it would have been a very useful comparison.
This is not possible with Apple parts.
That's what my example was about. It was only specific because I wanted to have a concrete example.
Talk about specifics, eh? Didn't you just argue against an article addressing "_their_" specific usecase?
In a store people will ask "is this better than an Apple?".
And I'll tell you one more thing, when I was in the industry and taking computing parts to build products with them I did not form an opinion by reading internet reviews. I haven't met anyone who did.
"Starting with computers using macOS 28, Rosetta functionality will be available only for certain older, unmaintained games that rely on Intel-based frameworks."
https://support.apple.com/en-us/102527
And
"Beyond this timeframe, we will keep a subset of Rosetta functionality aimed at supporting older unmaintained gaming titles, that rely on Intel-based frameworks."
https://developer.apple.com/documentation/apple-silicon/abou...
Colima is backed by qemu, not Rosetta, so if Rosetta disappeared tomorrow I don't think I'd notice. I'm sure it's "better" but when the competition is "good enough" it doesn't really matter.
I don't see how that's holding you back from using these tools for your work anymore than using a Makita power tool with LXT battery pack.
I learned to live with macOS, but I also like and use Gnome, which many Linux-only people hate. I tried most WMs on Linux, like Hyprland, Sway, i3, but none ever felt worth the config hassle when compared to the sane defaults of Gnome.
I have to admit that when I read this, my eyebrows went up so far that my hat moved.
Yes, Asahi exists, and props to the developers, but I don't think I'm alone in being unwilling to buy hardware from a manufacturer who obviously is not interested in supporting open operating systems
So they don’t actively help (or event make it easy by providing clear docs), but they do still do enough to enable really motivated people
This is an industry blog, not a consumer oriented blog.
The real reason is probably because they are supported by patrons and can only get new equipment to review when people donate (either money or sometimes the hardware itself).
If you like what they do (as pretty much the last in-depth hardware reviewers), consider supporting them.
As in, they don't sell you the parts, they only sell you the entire product. If you don't want the entire package, the processors alone are irrelevant.
The tested machine is an nvidia GB10 which nvidia makes and sells as a whole unit and various vendors stick it in different devices to try to differentiate (although in the end they're all basically identical).
And yes, it is extremely weird for it to never mention the Apple chip, which has a little something to do with who they thank for lending them the device. The arbitrary claims for why they ignored the enormous, class-leading ARM processor in the space is not convincing.
I mean, the other claim that this is an "industry blog" and not a "consumer blog" was equally silly. It's basically for curious hobbyists. Zero industry insiders follow this to see about the core in the GB10. It's basically Anandtech.
A few years ago they were writing articles about Apple Silicon.
Running the SPEC benchmark interger and floating piitnt suites takes all day, but it's hard to game a benchmark with that much depth.
It's a shame that nobody has been willing to offer that level of detail.
But I did some comparisons when I tested the same Dell GB10 hardware late last year: https://www.jeffgeerling.com/blog/2025/dells-version-dgx-spa...
Your “specifically ARM cores designed by and licensable from ARM Holdings” argument doesn’t hold any water.
And Qualcomm.
Anyway, here it is in GB10 form-
https://browser.geekbench.com/v6/cpu/14078585
And here is a comparable M5 in a laptop-
https://browser.geekbench.com/macs/macbook-pro-14-inch-2025
M5 has about a 32% per core advantage, though the DGX obviously has a much richer power budget so they tossed in 10 high performance cores and 10 efficiency cores (versus the 4 performance and 6 efficiency in the latter). Given the 10/10 vs 4/6 core layouts I would expect the former to massively trounce the latter on multicore, while it only marginally does.
Samsung used the same X925 core in their Exynos 2500 that they use on a flip phone. Mediatek put it in a couple of their chips as well.
"Reaching desktop" is always such a weird criteria though. It's kind of a meaningless bar.
"Daily driver" is probably a better term, but everyone's daily usage patterns will vary. I could do my day job with a VT100 emulator on a phone for example.