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.
Apple Silicon's Next Big Upgrade Could Arrive in 2026
- Thread starter MacRumors
- Start date
- Sort by reaction score
I agree and was a bit disappointed by this.
I expected them to talk about performance gains and tons of efficiency gains.
Based on TSMC's own statements on 3nm it should be a boost of ~15% performance and 25-35% efficiency. The fact they didn't talk about how much more efficient it is and we didn't see any battery life gains is puzzling.
They spent all their marketing budget on hunting down Mother Nature and putting her in a video. Everyone loved that and no one thought it was cringe.
A 2 nanometer chip is insane! 2026 here we come!! Also, my previous comment was deleted because I guess we can't have any fun and make people laugh on this site anymore. Sad.
That question is a long way off. 3nm, 2nm, etc are brilliant marketing terms from TSMC, but doesn’t describe the size of anything in a CPU chip currently.
For more info on process node marketing watch this
Apple learned a big lesson in releasing the M1 Macs back in Nov 2020. That is, you don't allow new generations of chip to come flying out of the blocks. They quickly learned that was a hard act to follow as the M2 chip was seen as nothing more than a speed bumped M1. For M3 Apple learned the lesson so that the new fab process will not allow all the dramatic gains to go out in the first iteration. M3 chips are neutered to only a 10% speed increase so that Apple can milk this in M4 and M5 generations. Then we get to this new fab process talked about here. It won't be generational but more neutering with a 10% spec increase to save the available additives for the following year's Macs.
All told, Apple Marketing is now following the Intel playbook to stretch technology gains over the lifetime of the fabrication tech. A truly garbage move by Apple.
All told, Apple Marketing is now following the Intel playbook to stretch technology gains over the lifetime of the fabrication tech. A truly garbage move by Apple.
why…speed bumps or slightly faster chips are really all anyone has been getting for 30 years. Why would this start to be a problem. No matter what magic features the chip has it all boils down to slightly faster chips year over year.
Moving from 5nm to 3nm and eventually to 2nm is all very well and good. But the benefits to the end user should be apparent.
So far, there appear to be no dramatic leaps in performance and battery life with the A17. Improvements, yes - about on the level of what they have been with each successive generation of chips. But no real leap with the move to 3nm.
Now, I’m willing to cut Apple some slack here, as the iPhone 15 Pros use the A17 built with the early N3B process. I’ll withhold judgement until the N3E chips come out next year.
So far, there appear to be no dramatic leaps in performance and battery life with the A17. Improvements, yes - about on the level of what they have been with each successive generation of chips. But no real leap with the move to 3nm.
Now, I’m willing to cut Apple some slack here, as the iPhone 15 Pros use the A17 built with the early N3B process. I’ll withhold judgement until the N3E chips come out next year.
That is a very doubtful timeline for Intel. Arrow Lake isn't going to ship in 1H 2024. It might start HVM productions in 1H, but ship to end users in finished computers? Probably not.
Intel isn't going to ship Intel 3 ( Granite Rapids) until 1H 2024.
" ... For the remaining three nodes, I would highlight Intel 3 met defect density and performance milestones in Q2, we reached PVK 1.1 and is on track for overall yields and performance targets. We will launch Sierra Forest in first half of 24 with Granite Rapids following shortly thereafter, our lead vehicles for Intel 3. ..."
Intel Roadmap Reaffirmed: 15th Gen Arrow Lake CPUs Already Running in Labs, Launch in Late 2024 | Hardware Times
The other day Intel reported its first profitable quarter after enduring multiple losses in a row. Unlike most semiconductor companies, Team Blue relies heavily on the PC market for the lion’s share of its revenue. Consequently, a weak PC market means a weak Intel, which we’ve seen over the last...
www.hardwaretimes.com
Not particularly creditable that they will get Intel 18 moving at the same time in any sort of respectable volume. Intel simply just doesn't have that many EUV fab machines . Indeed, from the same article quoting Gelsinger's comments.
" ... On Intel 18A, we continue to run internal and external chips and we remain on track to be manufacturing-ready by the second half of 2024. ..."
If somehow Sierra Forest and Granite Rapids don't get crushed by AMD and Intel actually gets decent uptick in demand for their server chips the 18A could slide out even more.
[ P.S. Yes the slide deck in that article says 1H for 20A , but 'shocker' Intel's power point slides are not entirely accurate once again. It is 'less worse' now on the magnitude of the time error, but they are still working on overpromising and underdelivering. Cmdr Montgomery Scott they are definitely not. ]
"3nm" , "2nm" , etc isn't measuring anything directly. The structures are going to get more 3D which a unidimensional unit of measure ( 'nm' ) isn't really appropriate or technically informative. People are grasping on 'nm' for continuity reasons to compare to fab processes from a decade or more ago. That really should stop because it is becomiing more and more a distortion and outright fiction. The new stuff over the next 3-5 years is just widely different to the point comparing 'apples to oranges'.
N3B vs N3E is an interesting comparison. N3E is actually slightly less dense, but expected to make a marginal gain in efficiency or performance.
The nanometer process doesn't actually refer to physical size. From Wikipedia on the 2nm process:
"The term "2 nanometer" or alternatively "20 angstrom" (a term used by Intel) has no relation to any actual physical feature (such as gate length, metal pitch or gate pitch) of the transistors. According to the projections contained in the 2021 update of the International Roadmap for Devices and Systems published by the Institute of Electrical and Electronics Engineers (IEEE), a "2.1 nm node range label" is expected to have a contacted gate pitch of 45 nanometers and a tightest metal pitch of 20 nanometers."
"The term "2 nanometer" or alternatively "20 angstrom" (a term used by Intel) has no relation to any actual physical feature (such as gate length, metal pitch or gate pitch) of the transistors. According to the projections contained in the 2021 update of the International Roadmap for Devices and Systems published by the Institute of Electrical and Electronics Engineers (IEEE), a "2.1 nm node range label" is expected to have a contacted gate pitch of 45 nanometers and a tightest metal pitch of 20 nanometers."
Judging from sack-peak figure, they count in Angstroms: A14 = 1.4 nm, A10 = 1nm, etc.
I miss the megahertz metric. It was a number that went up in the thousands then had that confusing stage when the numbers went down at least in 2 separate generations but the speed was faster anyways because of reasons. Having a metric that goes from 3 to 2 is not as exciting.
Doubtful, but still hoping Apple will find a way to have multi-tiered GPUs so we can finally get eGPUs again.
The thing is also, measuring performance increments in terms of percentages compared to the previous generation is a little bit dodgy, because things continually get faster. It’s a logarithmic scale of improvement, which from an engineering point of view may be unattainable in the long term, although it makes sense from a consumer perspective because if the new machine is twice as fast you will want to know that.
Apple could upgrade to considerably more advanced 2-nanometer chip fabrication technology as soon as 2026, according to analyst Ming-Chi Kuo.
In a recent post on Medium, Kuo explained that Apple will likely begin to use TSMC's 2-nanometer chip fabrication technology to produce iPhone chips as early as 2026.
TSMC N2 is likely to have just as 'bumpy' a roll out as N3 has had. Going to Gate-All-Around is jump like FlexFin is.
"as early as" has lots of slop in it to slide quite deep into 2026 for Apple's usage.
TSMC already has new stuff filling the calendar out to 2025 anyway.
"..
N3P H2 2024 , N3X 2025 , ... " and N3S presumably after N3P since doesn't have a solid date yet.
TSMC's 3nm update: N3P in production, N3X on track
The last and final FinFET node from TSMC lives on.
" ... The company said it is making "solid progress" in both yield and performance of N2 silicon, and that it is expecting it to deliver a better than 1.15 times improvement in density over the enhanced N3E node that is due to enter mass production this year. It is also projected to offer a 15 percent speed improvement over N3E at the same power, or up to a 30 percent power reduction at the same speed when it goes into production in 2025. ..."
TSMC claims it's still on track for 2nm chips in 2025
World's largest semiconductor contract manufacturer also gives details on 3nm nodes
Whether people are in a super hurry to rush to TSMC N2 or Intel 18A is going to depend upon pricing. If there is a 19% increase in pricing to get 15% increase in density I suspect Apple is not going to be tripping over themselves to get to this quicker. Especially for the monolithic die only packages.
When I upgrade to iPhone 16 Pro Max next year from my iPhone X, I won't upgrading again until 2031. I might make exceptions for the 20th anniversary iPhone depending on how significant of an update it is.
You should pitch the idea of using femtometers to Apple. Then we’d still be in the millions for now.
That's because it's fake 3nm. They're using the last node and marketing it as 3nm. That's why the number don't add up