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.
Could we see M3 before A17?
- Thread starter senttoschool
- Start date
- Sort by reaction score
He is a South Korean leaker, specializing in Samsung. I don't trust anything I can't verify, but others may find it helpful in planning their purchases.
How reputable is he/she? Has this account gotten anything important correct in the past?
No idea. Unfortunately, there doesn't seem to be a leaderboard for Samsung leakers similar to the one Apple Track has for Apple leakers, so it's hard to check if this leaker is accurate or not.
It isn't a value product.
While the iPhone is probably more profitable for Apple because it generates more service revenue, the MBA most certainly has a higher average selling price than the iPhone.
Also, no one is arguing that the iPhone is less important.
If he isn't known, we should stop posting him here. Otherwise, we could be posting speculation from hundreds of Twitter accounts.
Certainly Apple is in the process of ramping yield. We know this and we know it will take several quarters until yield peaks. So, this is not news. Literally speaking, there may be something meaningful in the statement if the sources have deep insight into the supply chain for end use products of the M3. No M3 products but M3 Pro Max Ultra will likely be launched. Apple has the option to limit new computer products to M3 Max Ultra if they choose. We need to recall that Apple adjusted their product strategy (release Pro Max Ultra first) to deal with the yield ramp issue in 2022. More specifically, Apple adjusted the strategy years before but we did not see it reflected in product announcements until 2022. There is undoubtably some inertia in the supply chain that will make it difficult to delay the product launches.
M2 Pro and Max weren't released until early this year when Apple updated the 14" and 16" MBPs. The M2 dropped last fall. So what are you even talking about by "release Pro Max Ultra first"?
TSMC's current 3nm chip production not enough to satisfy Apple's needs
The iPhone 15 Pro and iPhone 15 Ultra will be the first phones to sport a 3nm chipset when released later this year.
台積電大客戶紛放緩3奈米時程 超微PC至2025年停留在4奈米
儘管台積電持續保有絕對製程技術領先優勢，但供應鏈業者表示，景氣低迷、先進製程報價昂貴，客戶若放緩投片下單，恐對台積電長期成長動能目標有所減損。
www.digitimes.com.tw
TSMC Is Unable To Meet Apple's 3nm Chip Demand For The A17 Bionic And M3
Apple’s only 3nm semiconductor partner TSMC is reportedly unable to meet the demand requirement of the A17 Bionic and M3
wccftech.com
Apple Rumored To Have Lowered A17 Bionic Performance Goal As TSMC Continues To Struggle With 3nm Chip Production
As TSMC seemingly struggles with 3nm chip production, Apple may have lowered the A17 Bionic performance goal too
wccftech.com
Geez, 3nm is already delayed and even now, TSMC cant really mass produce for iPhone chips so I highly doubt that M3 will even release in this year. And yet releasing M3 way before A17? Ridiculously impossible.
WWDC doesn't really announce new hardwares. Mac Pro might be the exception but for now, Mac Pro is rumored to delay to October.
Even throttled the M3 Pro would have a larger RAM capacity and better performance than a M3. And definitely has more net mark-up dollars attached.
In previous generations, there were Windows PC laptops that had one performance level plugged in and another when unplugged and mobile. Apple could do similar thing here where the 'dock' you plugged into on desktop context had a fan and heatsink outside. ( metal to metal coupling between dock and iPad in the correct place, it could be tractable to pull the heat out of the iPad. Effectively, you put the iPad on a chill plate.) So it is possible to get a system. It just isn't very "iPad-ish" in usage. More akin to a Surface Studio with a detachable screen.
A fan shrinks usage battery capacity in two ways. First, the fan itself consumes energy. Whatever heat the SoC is putting out that is problematically high, it is going to cost even more than that to get rid of it (actively moving air mass). Second, it soaks up space ( so area for battery smaller. Yes, you are throwing thickness at it to crank up the volume but that means more aluminum and significantly more weight. ).
The M3 Pro package size will likely eat into battery capacity also.
The M3 Pro can limbo down M3 levels just by turning the additional CPU/GPU cores off when detached and back on when plugged in. ( Apple has already had that level of power gating control in the M-series. Probably going to get better in the 3rd generation). So it is possible. May not be attractive to those who wanted more power when detached, but it works. But yes, if Apple is really focused on attacking systems like the Surface Pro and Windows Tablets with this.. a fan weaved in. The quirky thing there is why is it being called an iPad? I guess that would be the point of tacking the 'Studio' suffix on the end.
If push the iPad 14" weight up to where it is the approximately the same as a M2 MBA system or higher. ( ≥ 2.7 lbs), then it looks like a strange 'duck'. But if Apple is chasing some niche that they think a sizable number of folks will pay big bucks for a strange 'duck' then perhaps they'll pursue it.
The simpler path is to just stick to M3 all around and hope the screen size difference can do all the segmentation.
Apple didn't have to put a AC/DC power converter inside a iMac 24". (or the laptops ). The iMac 24" EtherJack was banished to the external power brick. Apple has done some strange duck stuff before. I can't see calling something with a fan inside an iPad. But if put the iPad down on a desk... does it have to solely be limited to what a tether-less/detached iPad is limited to?
TSMC's current 3nm chip production not enough to satisfy Apple's needs
The iPhone 15 Pro and iPhone 15 Ultra will be the first phones to sport a 3nm chipset when released later this year.www.phonearena.com
台積電大客戶紛放緩3奈米時程 超微PC至2025年停留在4奈米
儘管台積電持續保有絕對製程技術領先優勢，但供應鏈業者表示，景氣低迷、先進製程報價昂貴，客戶若放緩投片下單，恐對台積電長期成長動能目標有所減損。www.digitimes.com.twTSMC Is Unable To Meet Apple's 3nm Chip Demand For The A17 Bionic And M3
Apple’s only 3nm semiconductor partner TSMC is reportedly unable to meet the demand requirement of the A17 Bionic and M3Apple Rumored To Have Lowered A17 Bionic Performance Goal As TSMC Continues To Struggle With 3nm Chip Production
As TSMC seemingly struggles with 3nm chip production, Apple may have lowered the A17 Bionic performance goal too
Geez, 3nm is already delayed and even now, TSMC cant really mass produce for iPhone chips so I highly doubt that M3 will even release in this year. And yet releasing M3 way before A17? Ridiculously impossible.
It isn't impossible.
Production in Dec-Mar could have been assigned to M3. M3 , especially if confined to just one Mac model doesn't need anywhere near the production run that A17 requires for launch. But the A17 system's launch actually isn't until September. ( just need to start getting A17's to the factories in April-May ... apple can't ramp the iPhone in Jan-Mar away because rest of the system likely isn't ready. )
What Apple can't do is release 2-4 M3 high volume models at the same time doing demand bubble builds for the iPhone in May-August. The whole "is not enough" is only really relative to doing 10's of millions of dies per month. The Mac yearly sales are only in the 20-30M range. Divide by 12 months and looking at 1.6-2.5 if the entire mac line up consisted of plain Mn dies. It doesn't. If narrow the view to number of iMac 24" dies need then likely in the sub 1M/mo range. The capacity TSMC has assigned now would work for that.
The other issue is that the iPhone production lines doesn't need 10's of millions per month. The A17 is restricted to only the iPhone Pro. More than half of the new iPhones produced this year are going to be on pre-A17 SoCs. The iPhone Pro was supply constrained last Fall ... did the sky fall in and the world collapse? Nope. Apple has a scarcity meme pop up about the highest price, highest margin phone so that there is a selling frenzy whipped up around the phone.
It would be an extremely tight fit to juggle both at the same peak demand times. But if limit the M3 rollout by spreading it out over several months and doing the potentially least popular M3 first , then that reduces the bubble demand for M3 dies.
So it is possible. What is likely impossible is to make everyone happy. If Apple did a M3 iMac and held the M3 from the laptops for 4-5 months lots of folks would be grumbling. There could be 1-1.5M M3 systems shipped but lots of complaining in the forums here. Similarily if the MBP 13" got the M3 , iMac 24" and then much , much later the MBA 13". Probably lots more grumbling than "impossible to do".
IF the hardware is ready then there is no good reason to skip WWDC either. Repordedly the M2 MBA 15" is really, really late. The M2 MBP 14", 16" were time shifted out of the Fall into early this year. The 15" likewise seems to be intended to be released probably about where the MBP 14/16 finally arrived.
It also wouldn't be surprising if Apple skipped M2 iMac 24" , in part that it was just a relatively low volume , low priority system. If it is very low volume and was planned solely for M3 era then Apple could put that toward the front of the M3 rollout. The iMac being comatose since Fall 2020 is likely having some 'staleness' issues far more so than rest of "plain Mn" SoC systems.
TSMC's 3nm is not really mass producing according to their news. They cant even fulfill A17 chip so why would they bother to make M3? Even iPhone Pro sell way more than Mac.
Logically, it's impossible to release M3 ahead of A17. Apple always focus on iPhone, not Mac.
Isn't this iPad Pro update coming in early 2024? Beta iPadOS 17 comes with lower software restrictions in June WWDC. Developers spend Aug-Dec 2023 unwinding arbitrary software restrictions in their code. Feb 2024 new hardware comes.
iPad OS 16 increased memory app limit to 16GB. If now want to push concurrent very large apps that is now 32GB as floor for mega loads. M2 Pro RAM max ... 32GB. even if the M3 got to a RAM max of 32GB, that still doesn't account for kernel and nominal background app RAM utilization (i.e., need something incrementally bigger than 32GB. )
It would have been well over a year since iPad OS 16 put app developers on a path to unwinding their app limitations up to new 16GB barrier.
If the iPadOS updates take better advantages of the M1/M2 why wouldn't there be an uptick on M3 hardware ?
And if not iPadOS 17 then versions 18 , 19 ? Your $1,600 iPad get faster every year for next 2-3 years... that is going to make the outsized initial investment look better or worse?
If Apple beats the drums to encourage large scale apps on iPadOS for several years, then at some point those apps will drive usage rates right past the plain Mn SoC once throw in concurrent multitasking.
Report claims Apple's M3 Mac chip is delayed until 2024
TSMC can't make enough of them yet.
www.macworld.com
More than 50% of WWDC events the last 20 years have brought new hardware announcements during the keynote.
A17 will be on this new Tsmc...probably the new mac SoC but in October-December, otherwise Apple is facing Intel issues again
Probably they wanted to release the iMac with this, now they will delay it or release it with M2 and so on..
I still think this isnt a reliable source...from the "14" ipad pro with M3 pro" for next year it was something fisshy
Probably they wanted to release the iMac with this, now they will delay it or release it with M2 and so on..
I still think this isnt a reliable source...from the "14" ipad pro with M3 pro" for next year it was something fisshy
Doesn't really count after Tim Cook took over Apple. So far, There were only 5 hardware announcements since 2012 so what can you say, huh? It's half and half and doesn't really guarantee that WWDC is meant for hardware announcements which is proven.
The value product is the M1 MBA, not the M2 MBA. The M2 MBA starts at $1199 and quickly goes up to $1600 if you want 16/512. The profit margins in hardware would probably be better than the iPhone Pro.
M2 is probably a fairly expensive chip to make though. It’s roughly 20-30% larger and uses more expensive RAM. I kind of doubt that Apple has increased the M2 Air prices to improve their margins, it was probably to compensate the increase in the manufacturing costs.
At $1600, I'm 100% certain that the profit margins for M2 MBA hardware are better than just about any iPhone. Apple charges $400 for 8/256, which probably costs $10 in supply.
M2 costs about ~$60 per chip and has 20 billion transistors. A16 has 16 billion transistors. Just the back of the napkin math suggests that A16 costs about ~$45.
Edit: You're talking about M1 to M2. I was writing about M2 vs A16.