M4+ Chip Generation - Speculation Megathread [MERGED]

[vigilant]

Jokes aside, Apple has incentive to get off this chip ASAP. I'm predicting we'll see M4 by next fall.

If this does happen expect it to be an incremental update much like the M2 was.

We got M2 Max and M3 Max in the same year which is insane to me. I'm "afraid" that M4 Max will hit at WWDC 2023 latest.

I find this to be highly doubtful. Around this time next year? Sure. But we’re going to need something to supplement 3NM to drive that to be anything more than an incremental update.

 

[senttoschool]

Jokes aside, Apple has incentive to get off this chip ASAP. I'm predicting we'll see M4 by next fall.

I agree. I think Apple would like to transition to N3E for both iPhone and Mac chips asap.

 

[MayaUser]

M4 must support tb5 and wifi7. So if they can do that by WWDC ok(but very very low chances), if not fall 2024 it is
I guess spring will bring the same m3 family in mac mini and macbook airs

 

[dgdosen]

I agree. I think Apple would like to transition to N3E for both iPhone and Mac chips asap.

Might that me a reason for LPDDR5 instead of LPDDR5X?

If this is a 'dead end' node, why bother with cutting edge memory?

 

[altaic]

M4 must support tb5 and wifi7.

Why? I’d like that, but please explain your imperative.

 

[T'hain Esh Kelch]

Might that me a reason for LPDDR5 instead of LPDDR5X?

If this is a 'dead end' node, why bother with cutting edge memory?

Why should that have anything to do with each other?

 

[MrGunny94]

Yay new thread! Anyways, after the disappointing M3 news on the display controller front, I do hope with the M4 they finally add a second display support on the base model.

I still dream of a 15" Air with dual 5K displays support and 120hz.

 

[Frantisekj]

We got M2 Max and M3 Max in the same year which is insane to me. I'm "afraid" that M4 Max will hit at WWDC 2023 latest.

More interesting will be if we will see M3 Extreme at WWDC 2024. If we will it will be real end of Apple silicon transition.

 

[Sydde]

More interesting will be if we will see M3 Extreme at WWDC 2024. If we will it will be real end of Apple silicon transition.

I think not in the way you are envisioning. In M2 and M1, the Pro was just a cut-off Max (or the Max was an extended Pro); with M3, the Pro die has a completely different floorplan. This means there is something of a chance that M3 Ultra (or whatever they call it) will not be a fusion chip but a completely different unitary die. Having its own layout would allow for a much more efficient, neatly-clustered chip, with optimal core counts rather than mere doubling. It would be more expensive to burn, but Apple probably has their act together now.

 

[XboxEvolved]

So the M5 series (likely the M5 Pro) is when I plan on upgrading my computer, and now that we have 3 series out there seems to be a pattern emerging.

With M1 you have the 5nm architecture.

With M2 you have the M1 architecture perfected to what Apple probably thought was practical to do with that size, or the “S” generation chip.

With M3, you have some amazing gains just as you had with the introduction of the M1, thanks in part to the 3nm tech that Apple has dominion over for now.

With M4, it will likely use the more efficient 3nm process that all other companies will be using, sometime in 2024, or essentially it will be the M1 S generation, and will work out any shortcomings the M3 may have had.

The M5, which I would imagine will be released at the end of 2025/beginning of 2026 would use the 2nm process that TSMC is widely expected to begin production in the second half of 2025. So what do you guys think? I am totally good with my M1 for now, but the gains between that time is kind of hard for me to imagine.

 

[altaic]

…

 

[Mac Hammer Fan]

M4: I definitely want more GPU cores - the number of GPU cores in the M3 family is a disappointment.
Also: more performance cores for the Pro.

 

[Chuckeee]

So the M5 series (likely the M5 Pro) is when I plan on upgrading my computer, and now that we have 3 series out there seems to be a pattern emerging.

With M1 you have the 5nm architecture.

With M2 you have the M1 architecture perfected to what Apple probably thought was practical to do with that size, or the “S” generation chip.

With M3, you have some amazing gains just as you had with the introduction of the M1, thanks in part to the 3nm tech that Apple has dominion over for now.

With M4, it will likely use the more efficient 3nm process that all other companies will be using, sometime in 2024, or essentially it will be the M1 S generation, and will work out any shortcomings the M3 may have had.

The M5, which I would imagine will be released at the end of 2025/beginning of 2026 would use the 2nm process that TSMC is widely expected to begin production in the second half of 2025. So what do you guys think? I am totally good with my M1 for now, but the gains between that time is kind of hard for me to imagine.

Just in general philosophically, I prefer to not be guinea pig for new technology. Using your nomenclature I think it’s always better to go with the “S” version. I think waiting for the 2 nm version of the architecture is fine but you’d be better off waiting for the M6 and the second iteration then getting the leading edge with the M5.

 

[Frantisekj]

I think not in the way you are envisioning. In M2 and M1, the Pro was just a cut-off Max (or the Max was an extended Pro); with M3, the Pro die has a completely different floorplan. This means there is something of a chance that M3 Ultra (or whatever they call it) will not be a fusion chip but a completely different unitary die. Having its own layout would allow for a much more efficient, neatly-clustered chip, with optimal core counts rather than mere doubling. It would be more expensive to burn, but Apple probably has their act together now.

You can be right. If apple bite that bitter pill of higher cost they probably went this way as doubling chips showed troublesome.

 

[XboxEvolved]

Now that we have the M3 series out I think we can probably begin to predict when Apple will release its next chips.

The M1(5nm)series will probably become the longest generation overall without an update, and that would probably be due mostly to chip shortages at the time, but all in all that generation went about 20 months before an update, and the updates began to roll out with the M2 around WWDC 2022.

The M2 Series (5nm, improved fabrication technique),building on top of the architecture of the M1, went on for about 17 months until the M3 was introduced. A rumored “extreme“ variant was rumored but never materialized for whatever reason, if it was in fact even real.

The M3(3nm) series is unique in that pretty much all of the chip family other than the Ultra and a possible “extreme“ version was introduced. Apple also apparently has a deal with TSMC so to make sure they had all of the 3nm production and a jump start on 3nm tech, I would guess this is why they released all these machines when they did, and all next year we will see iPad and lower end Mac refreshes throughout the year using the M3. Because of the Apple Vision Pro, I can’t see Apple releasing any of these new computers until WWDC, with the last of them coming around Oct of 2024. If an ”extreme” were to be released, I wouldn’t expect it until late 2024, early 2025. All in all, I would guess this processor family goes for 16-18 months before M4 is released in 2025.

The M4 (3nm improved manufacturing process) if the M2 is like a “M1S” then the M4 would definitely be the “M2S” I think a lot of this really depends on if they have any new products that could take advantage of the 3nm chip such as an updated Vision Pro, but judging by how the M2 generation was released, we will get these chips around WWDC 2025, with the last of them probably released around October/November of 2025. I would guess this generation last even shorter Than the M2 generation did, at around a year tops as Apple likely is expecting demand to soften with these chips as well.

The M5 (“2nm” my understanding its not really 2NM just a far more dense 3NM) would at the earliest come in late 2025, but for Apple’s computers I think it is far more likely they release the computers that use these chips in 2026. My reasoning behind that is that as far as I know, first run of these chips won’t even begin until first half of 2025, and full run won’t begin until the second half, which if Apple plans to use these for iPhones and such, I wouldn’t even be surprised if there was a delay in iPhone‘s launch at the end of the year to October/November.

So what do you guys think? Does this all sound reasonable/ realistic for Apple to follow A timeline like this? Also for reference, there is a graph from Wikipedia below to show when each of the current chips were released.

 

[canyonblue737]

absolutely agree although I think everything will be on the long end of your estimates. I believe all the M3's (Ultra and anything else) will be launched by fall 2024 with the first M4 based on the new N3E enhanced 3nm process launch maybe at WWDC in 2025. That puts the first M5 at the very earliest a late, late 2026 launch or maybe spring 2027.

I think the M4's big claim to fame will be a relaunch of a larger iMac (32 inches) that has options for a M4 Pro (or maybe even Max) processor and a *shocking* update of the iMac accessories to USB-C. I think we also see the iMac updated each chip generation now, I don't think it will skip a generation like what happened with M2.

 

[XboxEvolved]

absolutely agree although I think everything will be on the long end of your estimates. I believe all the M3's (Ultra and anything else) will be launched by fall 2024 with the first M4 based on the new N3E enhanced 3nm process launch maybe at WWDC in 2025. That puts the first M5 at the very earliest a late, late 2026 launch or maybe spring 2027.

I think the M4's big claim to fame will be a relaunch of a larger iMac (32 inches) that has options for a M4 Pro (or maybe even Max) processor and a *shocking* update of the iMac accessories to USB-C. I think we also see the iMac updated each chip generation now, I don't think it will skip a generation like what happened with M2.

Well I am basing a lot of this off of the average time in between the introductions of each generation, and just when the chip technologies themselves are supposed to be ready, going by TSMC’s fab tech roadmaps. Late 2026 for M6 imo would be pushing it, making Apple late to the party, as by that point Intel, AMD, Qualcomm, etc will probably have at least released some products using that tech, unless there is a delay of some sort in the tech itself. I’d go so far as to say that by WWDC 2026, Apple will have released products using the M5, M5 Pro, and M5 Max. The M5 Ultra and M5 “Extreme” (if it ever materializes) maybe not so much.

I can definitely see them wanting to release an M5 based Vision Pro by this point, and by this point we will have had a Vision “Air” or whatever they call it that would likely use the previous year’s chips.

 

[canyonblue737]

Well I am basing a lot of this off of the average time in between the introductions of each generation, and just when the chip technologies themselves are supposed to be ready, going by TSMC’s fab tech roadmaps. Late 2026 for M6 imo would be pushing it, making Apple late to the party, as by that point Intel, AMD, Qualcomm, etc will probably have at least released some products using that tech, unless there is a delay of some sort in the tech itself. I’d go so far as to say that by WWDC 2026, Apple will have released products using the M5, M5 Pro, and M5 Max. The M5 Ultra and M5 “Extreme” (if it ever materializes) maybe not so much.

I can definitely see them wanting to release an M5 based Vision Pro by this point, and by this point we will have had a Vision “Air” or whatever they call it that would likely use the previous year’s chips.

the speed of chip launches would have to dramatically speed up to nearly 1 new generation a year to get M5 chips by summer 2026 for WWDC. I think the M3 generation is Nov 2023 -> June 2025, M4 is June 2025 -> Nov 2026, M5 is Nov 2026 -> June 2027 most likely... no M6 till 2027 IMHO. There are some hints Apple wants faster launches, my brand new M3 iMac came with OS 13.5 from JULY 2023 pre-installed, implying these computers were ready for a WWDC announcement that never came this year.

 

[XboxEvolved]

the speed of chip launches would have to dramatically speed up to nearly 1 new generation a year to get M5 chips by summer 2026 for WWDC. I think the M3 generation is Nov 2023 -> June 2025, M4 is June 2025 -> Nov 2026, M5 is Nov 2026 -> June 2027 most likely... no M6 till 2027 IMHO. There are some hints Apple wants faster launches, my brand new M3 iMac came with OS 13.5 from JULY 2023 pre-installed, implying these computers were ready for a WWDC announcement that never came this year.

When you take into account that the time between m1 and m2 can be accounted for by way of chip shortages and delays due to Covid, the new 3nm fab tech will be ready in late 2024, and the 2nm tech ready in late 2025, I’m just saying I’d be shocked if m5 generation doesn’t come by wwdc 2026. I think I read somewhere also that there is evidence of the M5 going as far back as code from the middle of this year as well. That’s over 2 and half years of two generations of processors (m3, m4)

 

[Chuckeee]

For some of us, there are alternative wish list items, not just more & faster cores:
AV1 h/w encoder (M3 had only a decoder)
H.266 h/w encoder/decoder
Higher external display frame rates
Thunderbolt 5

 

[maxorific1]

What an interesting discussion! Here I am clinging to my mid-2015 half eyeballing the m3, half skeptical that we won’t just see m4 next fall based on the recent pace of releases.

 

[Kontikonsumi]

What an interesting discussion! Here I am clinging to my mid-2015 half eyeballing the m3, half skeptical that we won’t just see m4 next fall based on the recent pace of releases.

mbp 2015, 15" dude here as well. I want to hold off until M5/M6 and until this machine gets 10+ years old. 😎

 

[cocoua]

Expect a 15-20% improvement (as much!) as M1<M2<M3

No matter what the technical specs are, Apple never would release a M version way beyond the previous one or way ahead the competition, even if they could make it 200x performance, because nobody wants to disrupt the market.

Even a 200x improve over competition only could make Apple sell a few more Macs, and this jump would hurt Apple in the mid term.

SO the technical specs would be given by a factor of desirable performance+profit+how the competition is doing it.

Intel show us this very well, x86 single core improvement was stuck until M1 came, and suddenly X86 jumped to the future in one year...

 

[cocoua]

The M2 Ultra is honestly embarrassing for a $4K desktop computer. You can get vastly more performance in a PC for that price. Apple Silicon was made for laptops and phones. There’d be no shame for Apple if they threw in the white flag and admitted that on desktop, where performance per watt is irrelevant, Apple silicon can‘t compete.

after the M3Max this is even more embarrassing, did you expend 6000USD in your Ultra this a month ago? a month after there is a Macbook, with one of the best screen out there (though 16"), same performance (not in GPU area, but remember not every app uses ultra at it highest ..) as your super new M2 Ultra at 30% less price.

 

[MRMSFC]

No matter what the technical specs are, Apple never would release a M version way beyond the previous one or way ahead the competition, even if they could make it 200x performance, because nobody wants to disrupt the market.

So you’re saying that Apple wouldn’t take the opportunity to dominate the PC market in favor of being lockstep with their competition?

I don’t think I have to explain why that logic is asinine.

Even a 200x improve over competition only could make Apple sell a few more Macs, and this jump would hurt Apple in the mid term.

Because nobody would buy a computer that performs 200x faster because… why?

Intel show us this very well, x86 single core improvement was stuck until M1 came, and suddenly X86 jumped to the future in one year...

1. Intel was dominant due to their fab lead and AMD’s architecture was not competitive.

2. AMD’s Zen architecture debuted in 2017, three years before the M1

3. Intels performance has come at the expense of power efficiency, with their processors drawing more and more power every generation to get those performance improvements.

Intel did not “jump to the future” in a single year. Intel has been making band-aid improvements to try and keep the performance crown each generation with all other factors be damned.