Apple expands M3 Mac testing ahead of rumored October launch | Hacker News Reader

Apple expands M3 Mac testing ahead of rumored October launch

9to5mac.com

44 pointsmikece28 comments

It’s crazy how the M1 chip, which was released almost 3 years ago, is still the most efficient chip on the market. Let’s all forget about the M2 family of chips and act like they never came out.

Wait what? In what way is M1 better than M2?

I think parent is saying a chip released nearly 3 years ago still hasn’t been surpassed by the (non-Apple) competition.

And yet Apple isn’t standing still and is continuing to improve on it.

In grand scheme of things it hardly matters, because Apple doesn't compete outside mobile market, they own a very tiny portion of the worldwide desktop and laptop market, and have zero presence on game consoles, server and embedded markets, and are slowly abandoning the VFX workstation market to Windows and Linux.

Now if they actually cared to compete everywhere, that would be a different matter, regarding how threating their CPUs would be for everyone else outside the mobile market.

Yeah but they are making more profits than everyone else combined. I don’t see the point of expanding into a bunch of random markets. What they’re doing is working tremendously well.

Ah, if hadn't been for Microsoft this day back in 1997.

https://twitter.com/JonErlichman/status/1688289036157243394

Still one of the most incredible turnarounds in corporate history.

There's always the paralysis of waiting for the next model. But the M2 Studio was only announced at the end of June, and now we're looking at the M3 coming out just a few months later.

That's annoying, not just selling, but announcing and releasing high end computers that you know will be superseded within a few months.

Especially since without choosing a base model, Apple was offering 4-6 week delivery windows. It's July, you buy a brand new Mac with a new processor on release day, it arrives in the middle of August and look, 6 weeks later, Apple is announcing the M3 Mac.

The M2 Ultra will not be superseded within a few months. That's not how this works. The M3 Ultra will probably be ready some time mid next year.

The M1 was released November 2020; M1 Pro/Max October 2021; M1 Ultra March 2022.

The M2 was released June 2022; M2 Pro/Max January 2023; M2 Ultra June 2023.

If you need an M2 Ultra, your needs aren't gonna be met by the base M3.

That's a good point, true. Thanks for the timeline.

Again, surprised that there are close to zero rumour on M3 / A17 benchmarks. That is rather strange and we haven't seen something like this in the past 5 years.

FWIW, from March: "A new leak has offered users a look into the performance of Apple’s forthcoming chipset, the M3. The prediction was shared by tech industry analyst Vadim Yuryev, who ran the M3 through the GeekBench 6 platform."

"According to the test, the M3 performed over 20% faster than both chips and scored 3,472 points in the single-core tests and 13,676 points in the multi-core tests. The numbers place the M3 above its predecessor, the M2 Max and M2 Pro."

https://hypebeast.com/2023/3/apple-m3-chipset-performance-es...

This looks like speculation, not a leak.

Huh what? The cited ... X (or whatever they're called) claims to be a performance estimate. I can't see where Yuryev claims to have run the M3 through GeekBench?

What a strange article.

That was the only one and within one day he claims those number was given to him and proved to be complete BS. The M2 has 2500 GB6 score single core, a 3500 is 40% increase in single core performance. Given we are not expecting much increase in clock speed, that is 40% increase in IPC, I don't remember last time the industry has had anything similar.

There is a high probability that M3 will simply be a port of M2 to N3 with some minor tweaks.

The real long node has always been N3E.

Beautiful machines. I can’t really find a purpose for them, so I’m not buying. I got a Mac Mini in 2018 and used it for Xcode Build Server, but when the project finished, it was left unused. My current project has a dedicated devops who prefers Jenkins running somewhere in the cloud.

Meanwhile Apple broke high refresh rate for external displays on Intel machines and not in a hurry to fix it:

https://news.ycombinator.com/item?id=37018902

I’m curious how long it will take Apple to switch from Lightning to USB-C across the rest of their product lines now that the iPhone has switched.

Most likely this year. The only ones left are AirPods, MagSafe battery pack I think, assuming iPhone gets it which is very likely given all the leaks.

Magic Trackpad too. Just bought a new one and it came with a usb-c to lightning cable.

Magic Mouse and Magic Keyboard (the desktop version, not the one for iPads) are Lightning as well.

Here’s Macworld on that topic: https://www.macworld.com/article/1341317/lightning-usb-c-iph...

It should happen pretty quickly. No reason not to do it now. Rumors are that Airpods case will be upgraded this fall. That mainly leaves the keyboards, mouse, and trackpads. It’s an easy switch to make.

not long. The 2018 ipad pros already had usb-c. it will just happen.

So it’s like, apple will just announce “incredible performance gains” while in reality just piggybacking an improved process (by tsmc) ?

Customers don't care where the performance gains come from. Everybody has access to TSMC so it's their fault if they don't use it.

> Everybody has access to TSMC so it's their fault if they don't use it.

Not really, Apple buys swathes of their capacity before anyone else can. They have 90% of TSMC's 3nm allocated for Apple silicon, for example.

Even further Apple's commitment to TSMC allows the financial certainty which permits the kinds of capital expenditure that TSMC makes to their bleeding edge nodes.

Apple are not altruistic but their actions are a big part of enabling these advances.

Reply on news.ycombinator.com