Today Mark Gurman was a guest on the Max Tech channel and they talked about the M2 chip, the slow SSDs and overheating controversies and upcoming Apple products. Around the 18:20 mark, Gurman says the M2 chip is stop gap chip and the M1, M2, M3 are each one year behind schedule.
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.
Mark Gurman says the M2 chip is supposed to be a stopgap chip
- Thread starter Retskrad
- Start date
- Sort by reaction score
Weird to call it a stopgap. It just seems like it was late.
Does that mean that the M3 is a stopgap for the M4?
Hmmm. stop gap? Maybe like the Tick-Tock updates schedule that Intel did/does. One is small updates to previous design; the other is a more major change that takes longer. They work on them in parallel.
Gurman's credibility is about where the credibility of the "little boy who cried wolf" was right before he go eaten... Gurman's crystal ball shattered a long time ago...
It's definitely late — it was supposed to come autumn last year. Apple is at least 6-8 months behind the schedule. But it's understandable given how messed up the world situation was/is.
Apple marketing describes the M2 as "beginning the next generation." Which, when you think about it, kinda means the same as stopgap, but with better spin.
Well, they don't describe it as the next generation, but only as beginning the next generation, i.e., it is transitional, and something transitional is stopgap. Not my opinion, btw
It's "beginning" the next generation because they still have to produce the M2 Pro, M2 Max, and M2 Ultra at least (plus whatever is going into the new Mac Pro). It is the first chip of the family.
Well surely people didn't expect the same jump in performance from M1 to M2 that we saw from Intel to M1. I mean the M1 was crushing the best i9 in the 2019 16" MBP in many tasks. The M2 looks to have a respectable jump in performance to the M1 and all the controversy seems to surround the fact that an M2 throttles down in a fan-less laptop. That's to be expected and the M1 did as well.
I don't need some youtuber to tell me it's obvious and nobody was saying otherwise.
It's the same process node with some optimizations, which happens ALL THE TIME in the chip industry, nobody can release new chips ONLY on new process nodes, makes no sense anyway.
It's the same process node with some optimizations, which happens ALL THE TIME in the chip industry, nobody can release new chips ONLY on new process nodes, makes no sense anyway.
Didn't Apple experience a lot of 'brain drain' in the chip design department? The Nuvia guys leaving, then they wooed away a bunch of chip engineers from Apple, now they're absorbed by Qualcomm..
No, as long as they have this guy it's all good.
Apple Leadership - Johny Srouji
Johny Srouji is Apple’s senior vice president of Hardware Technologies and a member of Apple’s executive leadership team. Learn more about Johny.
www.apple.com
When he said the M2 is a stopgap, he meant it won't be around as long:
Currently, the M chips are nearly year behind the A chips in 𝜇architecture: The Macs coming out now and for the rest of the year will be M2, but that's based on A15, while this November's upper-end iPhones will be based on the A16. Thus, according to Gurman, to close this gap, Apple will try to release the M3 as soon as possible. Thus the M2 won't be around very long.
I, however, am not sure if Apple wants the M-releases to catch up with the A-releases. Apple's strategy might instead be to have the iPhones lead the Macs in 𝜇architecture, while the Macs lead the iPhones in process level:
The M-chips are much bigger and more complex than the A-chips. So Apple's strategy might be for the first releases on the new 𝜇arch to be the simpler A-chips, and then gradually (as they increase their own understanding of the new 𝜇arch), release succesively larger and more complex chips. That's exactly what they did with the M1: first came the A14, then the M1, then the the M1 Pro/Max, and finally the M1 Ultra.
However, while the iPhones may lead the Macs in new 𝜇arch's, the converse may be true for process level, such that the first chips released on new process are the M-series (in particular, the higher-end M-series chips). Since initial volumes for a new process will be low, it makes sense to first use those new chips for lower-volume products and then, as production ramps up, move the highest-volume products (the iPhones) to the new process. That appears to be what they are doing now: The first devices to get the 3 nm chips will likely be the Macs released in early 2023. The iPhone (at least according to rumors) won't be getting 3 nm this year, so it will be Nov. 2023 at the earliest for it.
Currently, the M chips are nearly year behind the A chips in 𝜇architecture: The Macs coming out now and for the rest of the year will be M2, but that's based on A15, while this November's upper-end iPhones will be based on the A16. Thus, according to Gurman, to close this gap, Apple will try to release the M3 as soon as possible. Thus the M2 won't be around very long.
I, however, am not sure if Apple wants the M-releases to catch up with the A-releases. Apple's strategy might instead be to have the iPhones lead the Macs in 𝜇architecture, while the Macs lead the iPhones in process level:
The M-chips are much bigger and more complex than the A-chips. So Apple's strategy might be for the first releases on the new 𝜇arch to be the simpler A-chips, and then gradually (as they increase their own understanding of the new 𝜇arch), release succesively larger and more complex chips. That's exactly what they did with the M1: first came the A14, then the M1, then the the M1 Pro/Max, and finally the M1 Ultra.
However, while the iPhones may lead the Macs in new 𝜇arch's, the converse may be true for process level, such that the first chips released on new process are the M-series (in particular, the higher-end M-series chips). Since initial volumes for a new process will be low, it makes sense to first use those new chips for lower-volume products and then, as production ramps up, move the highest-volume products (the iPhones) to the new process. That appears to be what they are doing now: The first devices to get the 3 nm chips will likely be the Macs released in early 2023. The iPhone (at least according to rumors) won't be getting 3 nm this year, so it will be Nov. 2023 at the earliest for it.
Last edited:
The fact that Gurman went on Max Tech at all really minimizes his credibility.
Anyone following TSMC already knew this.
Apple always uses the bleeding edge node. When TSMC announced N3 was delayed last year, M2 got delayed as well to reduce complexity.
A14/M1 = October 2021 / November 2021
A15/M2 = September 2021 / June 2022
Did anyone think it really took 9 extra months for Apple to do M2 using the same N5P process as A15?
Apple always uses the bleeding edge node. When TSMC announced N3 was delayed last year, M2 got delayed as well to reduce complexity.
A14/M1 = October 2021 / November 2021
A15/M2 = September 2021 / June 2022
Did anyone think it really took 9 extra months for Apple to do M2 using the same N5P process as A15?
I’m fairly certain that M2 was ready to go by fall 2021. There were probably other supply constraints (display, yields etc) that prevented the new MBA to be released earlier.
Apples reliance on cutting edge is certainly its weak point, especially given the circumstances. Let’s see how things will develop.
I think the Macbook Pros were delayed as well.
I think Apple wanted the schedule to be:
Fall: M
Spring: M Pro/Max
Summer: Ultra/Extreme
Fall: new M
I agree. With the Pros the delay reason is clear at least — the miniLED display production.
^^ Post of the thread, currently under appreciated.
The first two cycles have been pretty rough for Apple. They decided to pair new SoCs with brand new designs which caused a lot of delays along with Covid. I fully expect M3 to go much smoother.
I do think Apple has plans to at least update the base M, Pro, Max yearly. Maybe Ultra and Extreme on a two year cadence.
M2 could be a short-lived generation. I think Apple will want to rush out M3 to try to get back to the cadence they originally wanted.
Would Apple be on schedule if it launched the M3-based 14"/16 MBP next spring and the MBA M4 next fall?