Yeah, I wasn't implying the same fate will occur, just looking at the recursion of it all
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.
MP All Models 40 core ASI with PCI-E slot is around the corner!!
- Thread starter Amethyst
- Start date
- Sort by reaction score
I am afraid that you don't actually get it.
I can say how I believe the world should work. You get to say how you think the world should work.
But you don't get to say how I believe the world works.
Or perhaps like this:
I can say: "I think you should to this or that!" (that's me representing my opinion = fine)
I can't say: "You did or felt this or that!" (that's me representing your actions or opinions = not fine)
I am absolutely fine with, and agree, that you are 100% entitled to your opinion. That is not the same as agreeing with that opinion. And by the way, in that last bit of your post, can you see how you did that again: you are telling me what I'm doing? It looks like this:
"I'm also fine that you do not extend the same courtesy to others." (you representing me)
That is not OK. You should have written:
"It seems like you don't extend the same courtesy to others." (you representing you)
The last time we had this kind of discussion, in the thread you quote from, you ended up being banned. Let's not do that again.
During the last two weeks, I've come to realize a couple of things about the forum. I don't expect to keep trying to raise the bar in these threads. It would be good if the content of this post sinks in. These seemingly small changes in phrasing make all the difference.
It's OK to be opinionated—we have all been. Let's stick to our own opinions and keep the language civilized.
All of Gurman's very early chip leaks panned out on Apple's eventual delivery. I hope he's wrong now, or perhaps only has half the picture, but I don't see any rationale for him to suddenly drop into pure speculation on the extreme being cancelled. He admits to educated guesses in some of his commentary, but what, educated or not, would have caused him to U-turn on the earlier info?
Infinite irony machine.
I get it now. You get to tell people how they should do things. You tell us how WE get to think about your and others' posts. Got it.
Last edited:
Guys, I get that it’s all very important (to you 
), but the thread is about future Mac Pro…
Shall we perhaps, get back on topic (just saying)?
), but the thread is about future Mac Pro…Shall we perhaps, get back on topic (just saying)?
As soon as possible, I assure you.Shall we perhaps, get back on topic (just saying)?
Are you really sure? Because...
You actually have my quote just above in the same post where I say that I get to express my opinion and you get to express your opinion.
I hope you understand that it is completely impossible to "discuss" things where words lose all meaning. I'm starting to look for hidden cameras in my apartment because this sure feels like a joke.
But I'll say this: I'll exit this discussion now and let our respective posts just sit there for a while, doing their own thing.
You do seem to have some support for your arguments from a small circle of people, whereas no one agrees with me. It's a bit of a head-scratcher seeing as I just want constructive and polite discussions, but I guess if I'd bring a Tomahawk steak to a vegan cookout, it would be similarly awkward. Which isn't the same as saying Tomahawks suck—just stating the obvious.
I think that's it.
I don't know, the basic problem with Gurman letters is he actually have no cues on how hardware is developed and manufactured, to begin with that, Gurman's rationale even lacks any logic (not releasing m2 extreme on silicon availability), I need a lot of salt to eat that (didn't you?).
I could legit fear on Apple having a technical challenge going from 2 to 4 way UltraFusion fabric interconnect, but as I work in the area unless apple hired incompetent engineers (discard it, current staff it's All-stars) not in knowledge on how to scale a CPU fabric before sending it to the foundries, no I can't believe apple sent M2 Max to production having issues with a 4 Way Ultra Fusion.... Do you believe it could happen between an All-stars r&d engineers team repeating the work they previously done at AMD, Intel etc? No I don't buy neither apple cancelled m2
Maybe Gurman got an fake tip he can't filter given his lack of background on hardware r&d.
I'm quite confident on apple delivering a quad m2 max aka M2 extreme Mac Pro. Also discret DDR5 ram (no cue if DIMM, SO-DIMM or custom socket).
A separate thing that is under radar is m2 pro also seems has inbuilt ultra fusion (two ways), maybe apple plan is to differentiate even more Mac studio m2 Ultra from M2 Extreme (I'll begin naming it M2 modular as seem it will grow in x1,x2, x3,X4 steps, and support discreet PCIe and Ram what's is not part of m1 ultra (Mac studio) non modular.
About GPU is where I felt short, is quite confusing, since AMD Radeon keeps r&d teams going to Cupertino the same also seems apple is developing an huge PCIe x16 peripherals (an afterburner 2.0? Or an GPU accelerator build on defective m2 max with disabled CPU cores but working usable GPU cores?), Given the resources apple has on hands, only thing safe to discard is disappointement.
That's my mostly safe speculation (I'm no one at Apple, neither I got in touch with Apple staff leaking información, all this comes from my imagination fueled by pot beers and unhealthy food, lets set it clear I'm not a leaker for all legal purposes and liabilities, I just write here a bunch of ideas if coincidence happens its merely luck, no insider información -hi Apple legal-).
Regards from wizard's land.
Last edited:
Just to remember I posted it here Way before Gurman, Proser etc
militaries? well maybe apple needs an easy way to nuke storage / change the boot disk.
and no DFU mode that needs an 2th system may not fit that bill. (in some cases need to have an disk that can be destroyed) DID you know that dell and others have an warranty that does let you destroy an failed disk and still get an warranty replacement + on site service. Will apple ever do that?
lol, well I certainly hope your version of realty comes to pass, it sounds a lot better than Gurman's current line.
Yes, but I assume he acquires snippets of detail bobbing around within Apple's food chain, from which he attempts to construct his own narratives. The narratives, or reasoning, wrapped around whatever nuggets he receives are likely complete fiction, and pretty obvious fiction at times, but if the details of his sources turn out correct, that's what counts on his credibility scorecard.
I would suspect the actual reasoning or motivations behind Apple's executive decisions rarely, if ever, trickle far from their origin, but the results of those decisions ultimately have to become more widely known, and that's where you'd expect the leaks come from.
I started a thread about the possibility of a dual M2 Pro chip, so it’s pretty interesting that you think this a real thing. Is there any factual information confirming that this a thing…?
I also have to agree with you on the big chip, whatever they call it being real. I’m betting it’s literally going to be a shock to the (current) system!
Only information safe to share with you is watch at the transistor count rise relative to prior generation, then compare it with M2 Max transistor count rise, evidently something big was added.
Maybe I know more about, but nothing I could comment here.
Found this today and thought it was rather interesting.
I recommend scrolling down and start reading from: "In 2019..."
I recommend scrolling down and start reading from: "In 2019..."
commented enough this thread, remember by Gurman, M2 Max MBP should not be shipping now, neither anytime soon. He's speculating without actual cues, and doing little to none curation on the rumors he gather (if the case he got something actually).
And then there's the possibility of something like this . . .
. . . which @jmho posted in another thread, hope you don't mind me re-posting here 🙂.
If Gurman's info on the M2 extreme is correct, he may only have half the picture. It could be because the economics of producing MPX like modules of M2 Ultra and/or M2 Ultra duo makes more sense than producing an extreme version under whatever the current manufacturing reality is.
I think it's interesting that the M1 Ultra Studio still has a 3-4 week lead time ten months on from release and for most the time since release, has been sitting at between 8-12 weeks. I assume the bottleneck, for some reason, has to be the production of the M1 Ultra chip.
. . . which @jmho posted in another thread, hope you don't mind me re-posting here 🙂.
If Gurman's info on the M2 extreme is correct, he may only have half the picture. It could be because the economics of producing MPX like modules of M2 Ultra and/or M2 Ultra duo makes more sense than producing an extreme version under whatever the current manufacturing reality is.
I think it's interesting that the M1 Ultra Studio still has a 3-4 week lead time ten months on from release and for most the time since release, has been sitting at between 8-12 weeks. I assume the bottleneck, for some reason, has to be the production of the M1 Ultra chip.
Last edited:
Reality is A bunch of enthusiast directly opted by the M1 Ultra.
The Shortcomings from a Lego-SOC scalability issues are enough topic for a whole thread, Machine Learning with Mac Daisy-Chained is nothing new, I used some frameworks doing the same thing a decade ago, it dont pay as much as having a true 4P machine loaded with True Accelerators.
Two M2 Ultra barely matches a Single Amd MI200 you can install 2 of these in the cheese grater 7,1; I'm confident Apple either to enable AMD dGPU in M2 Mac Pro, or even manufacture its own compute accelerators, maybe recycling M2 Max/ Pro Chips with defective cores disabled and custom firmware to behave as an slave peripheral, it would be really smart, saving Silicon and Power the same time keeps the Mac GPU proprietary, or maybe Apple just build and updated Afterburner 2.0 based on commercial ASICs related to codec transcoding and ML.
Thanks.
just checked, and there is a ~5% increase vs the M2 Max difference. That is its 20% increase in transistors vs the 15% increase for M2 Max vs M1 Max. Yeah it’s a lot more than what would be need for the two low power cores and three GPU cores…even with the new neural engine. A lot to ponder.
If you don’t mind answering another question, are the weird holes seen in the die images possible TSV holes…? As in can silicon be added to the top of the new dies such as more memory…?
Apple is using TSMC inFo ls this time as cowos technically is 3D packaging but actually used as bridges, cowos implies that chiplets are placed on top a single big Base chip, while inFo-ls only uses an Small silicon bridge, this is meaningful as CoWoS-S besides expensive restricts the architecture arrangements as the lineal arrangements required by m2 'stacked' at edges would require extensive silicon waffers, meanwhile inFo only silicon is the bridge itself, both placed on top of either copper or other substrates, it allows bigger linear chiplets arrays.
The 8,1 will be a modernized take on the 6,1 that rises like a Phoenix from its thermal corner and arrives fully loaded with up to four custom M3X Ultras (80 CPU cores, 256 GPU cores, 1TB RAM), two each on two of the sides of the thermal core, the third side accommodating the optional dedicated Afterburner and Raytracing packages, Thunderbolt 4, eGPU support, HDMI, headphone jack, and SD Card slot.
Between now and WWDC, I’ll share all DMs I receive from Apple’s legal department.
Between now and WWDC, I’ll share all DMs I receive from Apple’s legal department.
