Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Rumored pro chip
- Thread starter cmaier
- Start date
- Sort by reaction score
Interesting that Gurman says the new iMacs will come with "M2x." That would support my original claim that we are looking at M2s coming in the upcoming pro machines (i.e. devices with new core designs).
Of course, lots of other conflicting rumors.
Of course, lots of other conflicting rumors.
Gurman mentioned that "M1X" is the most-likely option, with "M2" ("M2X", really) a possibility.
He is likely hedging his bets - a Late 2021 release would be "M1X" whereas a mid-to-late 2022 release could support "M2X".
He is likely hedging his bets - a Late 2021 release would be "M1X" whereas a mid-to-late 2022 release could support "M2X".
I admit that I'm enjoying the attempts to decode Apple's secrecy. The Kremlinology makes it more interesting, including when the leakers bomb. The details for Alder Lake have been known for around six months, roughly speaking. There have been a few rumblings about Lunar Lake, which is allegedly Intel's design for 2025. We can't even get a handle on what Apple plans for the rest of this year. The PC guys leak all over the place, very messy. While it may be satisfying to have the complete details on the Apple Silicon plans, it would be significantly more boring.
Last edited:
The new 'unpredictable' Apple actually makes buying things more of a headache. There are and always will be just enough details leaked to make you want to wait for 'that feature', but the inaccuracy of when it's launching and in what product until near the 11th hour makes waiting for something that could be 12+ months away more likely. The vacuum of real leaks also seems to be being partially filled by educated guesses, piling up expectations Apple can't meet. In a way the pressure release valve of regular, accurate leaks was probably a net benefit to all sides.
I laugh.. there's nothing PRO about Intel's chips anymore. Their fastest processor is in #26 spot behind practically every AMD processor.
PassMark Intel vs AMD CPU Benchmarks - High End
PassMark Software - CPU Benchmarks - Over 1 million CPUs and 1,000 models benchmarked and compared in graph form, updated daily!
www.cpubenchmark.net
I'm wondering whether the new iMacs are not gonna show up before next year, in time for M2X. That would explain why the new (delayed) MBP would get M1X.
I think "M1X" was always in the pipeline and always planned for the higher-end MacBook Pros (13" 4-port / 14" and the 16" model) because those make up the significant plurality of Apple's Mac sales so they would want to move them over ASAP once they completed the M1 transition on the "consumer" models (which is now complete with the release of the 24" iMac).
It is quite possible that the "Big Brother iMac / iMac Pro" (the iMac 5K replacement) and the "Mac mini pro" and Mac Pro 9,1 were all planned to be released across 2022 on an "M2-series" SoC using the "A15" performance and efficiency cores as the basis. Where I am a bit confused, however, is the claim by Gurman that Apple is working on a multi-die version of "Jade C" for the most powerful Macs and, for all intents, "Jade C" appears to be the internal code name for "M1X".
It is possible that the multi-die Jade C models are for the "iMac Pro / BBiMac" (as either a base SoC or a BTO option) and the "Mac Pro mini" and planned for 1H 2022 and that Mac Pro 9,1 will be something else entirely in terms of SoC (be it "M2something" or a unique SoC) in 2H 2022.
One of the problems here is that the CPU guys know what they have designed and when they have taped out the designs. So they know "this chip with different cores microarchitecture has been ready for 6 months." But they don't know the marketing name for chips, and they don't know anything about when the Macs using the chips will be released. Everything is compartmentalized. So you hear things in silicon valley, but even people with direct involvement don't have all the information.
Thanks, but I misspelled "Kremlinology", which I have now corrected in the original post. So, I don't deserve the points.
I get that there is a lot of grey area in this, particularly because Apple is very much compartmentalized. However, there does seem to be something unique with this situation. Perhaps it's because Intel, AMD, and Nvidia have to share details with multiple partners, while Apple is just working with TSMC and perhaps some select folks at Foxconn or the like.
Rene Ritchie posted a nice video explaining why Apple might have both "M1X" and "M2" in different model Macs at the same time:
Pretty much reiterates what many of us have been saying.
Perhaps Apple can run this video in the Mac section of the Apple Stores so consumers would not be confused.
Pretty much reiterates what many of us have been saying.
Perhaps Apple can run this video in the Mac section of the Apple Stores so consumers would not be confused.
Thing is, even TSMC doesn't know what a given chip is - they get a gdsii file dumped on them and that's about it.
I wasn't aware of that tidbit. Thanks for the info. It does make me wonder who Gurman's source is. Most leakers get their details from the supply chain, which isn't terribly reliable. While Gurman is very careful about what he claims, he's quite specific with the names, core counts, etc. I know some people think that Apple is intentionally leaking information to him, but I'm not sure what purpose it would serve to give him such arcane details that only matter to us tech nerds. I find it more likely that he simply has a well-placed source. The only instance that gives me pause is that your former colleague at AMD was the one who was leaking highly technical details about RDNA2 to "Moore's Law is Dead", as we discussed in another thread. Of course, Apple is known for secrecy above all else, but it is still curious.
They would know who the customer is, though. And could they extrapolate the type of chip based on the contents of the file?
they know the size of the chip, they number of transistors, and they might even figure out whether the core microarchitecture is different (if they go to the effort to try and figure that out). If they really wanted to, they could use tools to come up with the entire netlist.
I seriously doubt they are doing any of that. The chances that a human is doing anything with the data other than validating it against their DRC and DFM rules using automated tools is extremely slim.
If the MBPs were being pushed to 1H 2022, maybe. But trends still point to them coming in Q4 2021 and, IMO, that means "M1X".
@cmaier apologies if this has been asked, but from your experience, how much work would be involved going from M1 -> M1X (adding more cores) vs. just going straight from M1 -> M2 chip (new design)? I assume the second scenario is more work.
Also do you still believe M1X (whatever it is called) will use a design that is not fully based on M1?
Also do you still believe M1X (whatever it is called) will use a design that is not fully based on M1?
It’s more work, but the cores will have been designed anyway for the A15. So it’s “just” a matter of transforming it into an M series chip with the necessary tweaks.
It would be tremendously more work to go to M2. But they come out with a new microarchitecture every year anyway. (I am assuming it takes them about 2 years from start to finish, so they are working on at least 2 microarchitectures at a time).
Going from M1 -> M1x (same microarchitecture but more cores, more cache, different i/o's, etc.) would be much faster. I frequently worked on multiple related chips like that simultaneously.
As for your second question, M1x would, by definition, be based on M1 (with more cores, different cache size, other comparatively minor changes). So the question is whether the bigger iMacs and MBPs coming in the next 6 months or so have M2x's or M1x's in them. Either way, we know that new core microarchitectures will be out in Septemberish for iPhone, so that work will have been done.
Cmaier's point about SOC generations is logical - we are expecting the new iPhone later this year and it should sport the next iteration of the Apple Silicon core microarchitecture. If they are indeed going to rollout M2 next it probably would sport both a new microarchitecture and increased core counts - in essence M2X would precede M2.
that’s what I’ve been assuming, but who knows - global supply chains are a mess, etc. It just didn’t make sense to me to ship MacBook pros with slower single-core speeds than iphones in september.