I’m not expecting to see a new larger 27in. iMac, but there are rumours it could still be happening.
And if they skip M2 for iMac 24 and Mac Studio then the M3 versions just might still happen by the end of 2023.
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.
More evidence that Apple plans to update Apple Silicon yearly
- Thread starter senttoschool
- Start date
- Sort by reaction score
And if they skip M2 for iMac 24 and Mac Studio then the M3 versions just might still happen by the end of 2023.
I'm actually hoping that given the rumoured issues with the A16 that they skip the A16 derived M series and move straight to the A17 derived chip.
At full load a top end MBP 16" M1 Max will have a power input approaching 140W. There is allowance for charger efficiency as it will not actually output the full 140W constantly.
Similar to PC tower PSUs being rated for 1.5kW but it does output that much power for any long duration of time due to efficiencies
Intel came out with the 1st 24 CPU core laptop chip that requires as 330W charger.
M2 Ultra would have 24 CPU cores as well from two M1 Max chips that each has 12 CPU cores. The M2 Ultra would be tailored to the 240W USB PD charger.
Would not be surprised that the M2 Ultra would outperform the Intel part at sustained full load and yet not throttle because of better thermal output.
This assumes yield of that part will be better than the M1 Ultra. I suspect that was 1 factor as to why a MBP 16" M1 Ultra was not offered even after the 240W charger was announced last May 2022.
Seeming the Mac mini M2 Pro SKU got released then a MBP 16" M2 Ultra SKU would not be that far fetched.
Yep! It was actually a post of your's that made me think about this, and combining other bits of evidence together (like the published patents which Apple usually times to go out close to hardware availability) I grow increasingly convinced that this is what might be happening.
I happen to use one of those and I can ensure you it's very far from 140 watts. The charger is 140 watts, for fast charging. The SoC uses up to 40 watt for the CPU and up to 40 watts for the GPU, combined max SoC power usage is under 80 watts, just like with any other Mac laptop ever. The display can be power hungry, sure, and supporting circuitry also draws some power, so you might be able to get it to 120 watts or so if you put it into a torture test at maximal brightness, but that's not normal usage. Anyway, this is the TDP Apple has been using since the first Intel MBP and I see no reason to believe that they intend to triple it.
And by the way, if Apple decides that they want a faster laptop they will add another P-core cluster to it. UltraFusion comes with a power consumption overhead which makes it less suitable for laptops. Intel can use it because their per-tile bandwidth is much lower.
On the Mac Studio M1 Max to M1 Ultra CPU Max difference is 100W
So if that is actually lower then 240W charger has a lot of overhead for a M2 Ultra that does not actually use the full extra 100W overhead.
The point is Intel proved demand for a 24 CPU core chip that requires a 330W charger in a laptop with a near 100Whr battery.
If that did not exist then you'd have a point but with it out there then Apple may want to address that new demand.
If a MBP 16" M2 Ultra has a battery life 1/2 of up to 15hrs wireless web 22hrs Apple TV video playback then it is still a winner compared to any other laptop with that many CPU cores. It would be a class leader.
Last edited:
If they do move to a yearly cycle,
How much sooner will we start to see Apple move these M processors to obsolete?
How much sooner will we start to see Apple move these M processors to obsolete?
M2 ultra into an 16" laptop? no thank you, and it will never happen, why you like to talk about something that will never happen, its waste of space and time
That it is another "desktop" in a laptop form like so many from windows world....if you want that go with windows
M1 max to M1 ultra 100W , yes but look at the thermal evenlope between the mac studio and the laptop....physics never lie .. and i dont believe that Apple will bring back the 17" to somehow compensate that
14" with M Max is at its limits....16" with M ultra exceeds that, stop dreaming...Apple is not Intel, and Intel is not Apple
Intel can have whatever they want because they play in the windows world where you can find almost everything, from a good thinking device to an abstract one
This is what i had to say about something that not exist or will ever. People should not pay for extra performance that goes down after 4-5 min under sustain load
Now back on topic, off topics should be erased
PS: after i saw the internals from that book 3 ultra...i bet on my projects that i9+4070 will behave after 4 minutes like the base model...so why people should pay 1 more cent for something that perform similar ?! only for burst, fast tasks maybe
Whats next from you guys? place the M Max into the 13 Macbook Air
Last edited:
There is "ready" and there is READY....
Remember those benchmarks for an "M1x" (aka M1Pro) which were leaked as early as late 2020 with them being used in an early 2021 MacBookPro reusing the shell from the last Intel ones?
Maybe it was all just rumors, or maybe something did not go as planned.
My guess is that they just ran into issues when growing those chips beyond M1 size. Bad yields, not reaching frequency targets, memory controller overwhelmed by so many cores etc etc...
It is also quite clear that they did plan for something worthy an MacPro with the M2 or maybe even the M1 generation, but right now it is not even sure if they will have anything fitting that bill anytime soon.
Of course there is a market. For big gaming laptops. It’s not a new category. People have been putting desktop CPUs in laptops for a while. But it’s a tiny market and people looking to buy one of those are not interested in Macs.
Why would anything change here? It will still be 5-7 years, like with any other Apple device.
If I remember it correctly the main problem were supply issues with miniLED panels. I suppose Apple didn’t want to reuse the old chsssis.
Not for the late 2021 launch that happened, but if they had planned an early 2021 launch that old chassis would have been a given.
Apple execs on M2 chips, winning gamers and when to buy a Mac | TechCrunch
Apple's M series chips were incredibly well telegraphed when they arrived in late 2020. Apple had been designing its own silicon since the A4 appeared intechcrunch.com
That article has other context to that doesn't really point to exactly annual updates at all. Same article.
"...
We don’t want to leave them wondering…do they not care about us? A new phone shipped last year. Why didn’t the Mac get the love?”
“We want to reset to the technology curve and then we want to live on it. We don’t want the Mac to stray too far away from it.” ..."
Apple intends to follow the technology curve... not some arbitrary, non technical, iPhone release date schedule.
A major chunk of that technology curve is fab node updates. And that doesn't come in exactly 12 month increments.
If it is available Apple will take. If not they wait until it is ready. That strategy is not going get them exactly 12 month cycles.
The M-series chips are larger than the iPhone chip and sell at order of magnitude less volume ( at best if not 2 or 3 orders smaller). Apple is charging lots more to offset that but Apple 'killing off' Mx Ultra SoC every single year. Not even Intel or AMD does that in a sizably larger market.
You are very highly likely not going to see "across the whole SoC 20%" improvements every year. There could be specialized subsystems that move forward in years the CPU/GPU don't, but this is deeply flawed expectation trying to set here that much larger and harder to developer SoCs are coming to come at the same pace as monolithic dies that are 3-6x smaller.
Can already see cracks forming in that non technical iPhone every 12 month dogma where only the top end iPhone Max is actually getting any SoC increment and the mainstream phone is getting a 'de-binned' SoC update.
That quote you are pulling out is far more so a veiled jab at Intel. At one point Intel did have fab tech and design updates that they sat on for years because they were the dominate player and AMD was very busying shooting themselves in the foot. It step back from the Apple cheerleading / fanboy stage it is pretty is not the 'years ahead of the rest of the pack player here'. They don't have the single thread performance crown. They don't have the top end, power is no object graphics crown. On same fab node level their perf/watt gap over AMD or Nvidia isn't as big as it was back in 2020.
What is being said here is that Apple isn't in 'rest on your laurels" mode. That is fine but that is just going to keep them competitive. Apple really doesn't have a choice of squatting on advancements. They will roll them out as they achieve them but in no way shape or form means they are going to roll out on a constantly rigid time schedule. It is not all entirely under Apple's control.
P.S. There is only pragmaticaly only one iPhone SoC. ( the rest in the line up are 'last' or 'year before that' models). There are going to be at least four , if not five , M-Series models. As long as Apple drops one of those five in a year with a 'new' iPhone SoC then they are essentially 'showing love' to the Mac line also. Each line gets 'something' in a year.
There is around two or three orders of magnitude more products that those processors go into than what Apple has. There are just Mx , Mx Pro , Mx Max , Mx Ultra , and maybe Mx Extreme processor SKU produced by Apple. Intel and AMD do 3-5x that number of SKUs depending upon how much of their processor products want to cover. The system vendors collectively do 2, if not 3, orders of magnitude more product offerings that Apple does ( around 8-10 : MBA , MBP13" ,MBP 14/16" , Mac Studio , Mac Pro, Mini / Mini Pro )
The general PC market sells gobs and gobs and gobs over overlapping products. They are trying to sell everything to everything. So yeah, with a product line up as broad as the diameter of the Earth some subsection of that is going to update each year. Updating it all actually really isn't practical.
Sure they have more SQUs but they don't necessarily have more unique chips, for their desktop designs I think Intel has only 2 or 3 chips and it is likely similar for mobile. Intel die use
This means that Apple is producing roughly the same number of unique chips as Intel but Intel breaks theirs up into more different names while Apple just calls it an M(n)Max regardless of how many GPU cores it has.
Yearly upgrade cycle for M chips does not mean all devices will get updated yearly. iMac always gets the short end of the stick in that regard.
The size of the SoC isn't the issue, it's the size of the requisite cooling systems. Look at the size of the M1 Ultra heatsink - it barely fits into a Mac Studio. To make that work in a laptop form factor, Apple would either have to downclock the SoC to accomodate existing cooling specs (thereby negating a lot of the performance advantage) or run the SoC far hotter than most people would find comfortable...
Attachments
It's actually more than that when you consider the binned versions of those SoCs.
Currently:
M1 8 CPU/7 GPU
M1 8 CPU/8 GPU
M1 Max 10 CPU/24 GPU
M1 Max 10 CPU/32 GPU
M1 Ultra 20 CPU/48 GPU
M1 Ultra 20 CPU/64 GPU
M2 8 CPU/8 GPU
M2 8 CPU/10 GPU
M2 Pro 10 CPU/16 GPU
M2 Pro 12 CPU/19 GPU
M2 Max 12 CPU/30 GPU
M2 Max 12 CPU/38 GPU
M2 Ultra 24 CPU/60 GPU (future?)
M2 Ultra 24 CPU/76 GPU (future?)
I think this is a much fairer way of counting SKUs considering that Intel does binning too but calls the binned CPUs i3, i5s, and i7s.
Is this. More efficient way to do it? Probably. My issue will be with all the accusations that the updates are “iterative” and not revolutionary. It’s the same issue of yearly iphone updates. Every year its comparing 13 to 14 rather than say 11 to 14. If new M chips came out every 3 years they would feel more revolutionary even though it would be 3 iterations otherwise.
Or make it sound like a jet engine (the option most PC makers choose).
But....
One of Apple's main marketing point with MacBooks is that you can forget the charger and will be fine even on a multi day trip. Thats why I don't expect them to ever again a laptop that will use even close to those power numbers.
Intel has been known to be on a "tick-tock" cycle of CPU updates, where ticks are basically no more than spec bumps and tocks are the big jumps in performance. I think Apple will eventually settle into their own version of that cycle with respect to the M-series SoCs, but I think that it might be broken down in terms of Mx and Mx Pro/Max/Ultra SKUs rather than M3 vs M4.
At full load, it comsumes less than 100 watts and that also includes the screen etc. Not just the chip itself. I do have an M1 Max MacBook Pro 16" and, connected to my Apple Studio Display which provides 96 watts of power, it has no trouble providing sustained full load.
Probably the main reason has nothing to do with being able to keep max power consumption below 240 watts but because it would make a loud machine with significant performance drop while used on battery, which goes against everything Apple Silicon MacBooks are good for.
You mean tick-tick-tick-tick-tick-tock cycles right? 😉
No real life task will put a full load on every part of the SoC. The CPU, the GPU, ML, media encoders and whats not.
The charger should be bigger cos you still want to recharge the battery while putting a (not) full load on the SoC and you might even have some power draining devices connected via USB/thunderbolt.
It was never going to be a 2 year cycle anyway, if they kept to the iPad Pro (AX chip) release cadence it would've been more like 16 months between releases, and every few generations you'd have skipped over an architecture (e.g. there was no A11X or A13X) so it would have evened out overall either way. IMO that was always going to be too complicated to keep up when it wasn't just one chip one product anymore, but a family of chips going into multiple devices.
My MacBook Pro M1 Max 16 runs fine on a 65 W charger for the heaviest of my use cases. I rarely use a 140W charger, mostly if I need quick charging. I don’t want Apple getting into those ridiculously powered laptops. I tried a few of those intel ones; they are like mini furnaces and throttle a lot.
Buy a PC laptop if you need those high-powered heat furnaces.
Last edited: