It bugs me that this is wasted in a freaking iPad. Put it in the damn Macs already, especially the Mac Mini because that's what I want.
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 Debuts Next-Generation M4 Chip
- Thread starter MacRumors
- Start date
- Sort by reaction score
Can't wait to get some details on the M4 chip and what it might mean for the M4 MBP lineup. If "lowest tier" M4 now has 6 efficiency cores, then what happens to M4 Pro? M3 Pro (full) has 6p + 6e, so they'll need to make a bump somewhere, so 6p+8e or, even better, 8e + 8p? The Pro chip can finally be a real upgrade and the M4 Max could be monstrous.
Makes sense, given how much they have dropped the price of the iPad Pro this generation.... oh wait.
Go to your room. You can come out when Mr Cranky Pants doesn't reappear.
How bad has it gotten? Mine still seems pretty close to original capacity in terms of how long a charge lasts (also M1 iPad Pro, 2020). But I haven’t been tracking it to see whether it has degraded (it must have to some degree, being 4 years old!).
It's not about new processors. It's about new processes, which is totally different. Or course M4 is both.
However, N3E is not nearly the hurdle that N3B was, as they're doing all the same things, just less of it. From somewhat fallible memory, N3E uses 19 EUV layers while N3B uses, um, 25? Something like that. And more importantly, N3E uses no double-patterning at all.
So the ramp for N3E is faster than the N3B ramp. Yield will start better and improve faster- I think much better and much faster, though I'm not qualified to offer that opinion with certainty.
As best as anyone has been able to determine, the idea that N3B had big yield problems is a weird internet rumor that just took on a life of its own, but has no basis in fact. It is however true that N3B is significantly more expensive than N3E, for the reasons outlined above (much greater complexity).
That's *extremely* unlikely.
There's no chance that that's due to production "problems". Look at the size of the CPUs vs. the size of the entire die. However your idea about them culling from production for Macs as well does have merit... *if* they're actually making them in that volume already, which, I don't know, would be ahead of TSMC's earlier published schedule, but it's possible.
I suspect that many of those 3-P-core chips will actually have a perfectly good 4th core that's fused off.
Because you clearly know absolutely nothing about chips. That's ludicrous. Just google for "apple m3 die shot" and do a few minutes of reading.
...at least, the pro/max part is ridiculous. The MBA M4 having more cores is highly unlikely, though maybe a little less crazy- we'll know if we get a die shot of the M4. It wouldn't make sense though - it doesn't help with product segmentation, and fusing off cores in the iPad is pointless and even counterproductive (except for segmentation *within* the pad product line).
I was clearly joking with the last part.
But as for my post, would 99% of iPad Pro users notice any difference between the M2 and M4? Would 99% of artists and content creators, even, notice a difference? Apple desperately wants the iPad to be seen as a computer, but it just isn't.
Does it mean that you don't really know what you're talking about? Apple claims the M4 is the fastest NPU in any device, which, yeah sure, is correct. For now. The Snapdragon Elite X has a more powerful NPU. And given the performance comparison Qualcomm has shown compared to the M3 and the claims Apple is making about the M4 compared to the M2, Qualcomm has nothing to worry about.
The M4 is at best roughly equal to the Snapdragon X Elite. The good news is; that's nothing for you to be defensive about. Apple can take care of itself and it *really* doesn't matter in the slightest in the first place.
The comparison Qualcomm did was comparing the chips with at totally different power level and for some reason not comparing with M3 Max.
I would say M4 is an outlier, due to process issues on TSMC's end. (And also perhaps in part because of the "AI" hype.)
How else are they going to get people to choose an iPad over a Mac.
Did you get lost in all of QC's hype and forget to look at the actual numbers? The SXE is way behind the M3 in single-core performance, nevermind the M4, and, well, choosing to compare a 12-performance-core SXE with the bottom-rung M3 (with just 4 P cores and 4 E cores) says pretty much everything that needs to be said. The quite middling M3 Pro beats it, and the Max leaves it in the dust.
It's barely possible we'll actually see M4 Pro and M4 Max shipping before the SXE. WWDC should be interesting.
As for the NPU numbers... we'll see. It may be exactly as you say, but QC has been playing fast and loose with the numbers since they first announced their chip, so I would withhold judgement until they're in the wild. (Also, once they are, how long will it be before anyone has software that takes advantage of the QC NPUs?)
And yes, I'm using GB6 numbers for everything. It's not the ideal benchmark, but it's what we've got, and it's not bad.
As I've said before, when you do a tech comparison, QC loses to Apple and it's not even close. But in the real world, actual customers aren't judging technology, they're judging tech and price. There, QC has a chance to take the fight to Apple; if they can price their chips so that laptops based on the SXE are cheaper than Apple's, with roughly comparable quality, then in some ways it *is* fair to compare the SXE to the base M3. On their track record... I don't think they will.
And that's probably not important to them anyway - their real competition is other windows laptops. How that plays out will depend on both price and how well WoA runs. We shall see.
A4 and A5 was the last time that an iPad got a new architecture before the iPhone - a while ago
Since we are not seeing Thunderbolt 5 or Wi-Fi 7 on the iPad with M4 here, I assume it would be safe to conclude that any device (e.g., Studio, MacBook Pro) coming out in the next 12 months or so with an M4 chip would also not have these given the SoC design? So earliest implementation of T5 and Wi-Fi 7 would be M5?
A10X alsoA4 and A5 was the last time that an iPad got a new architecture before the iPhone - a while ago
That's a lousy argument. We've been through this already.
It's not like they're hiding what the comparisons points are. In this case it's blindingly obvious what's going on as there is no M3 iPad. As for the M3, their point was that they don't expect most users of M2s to upgrade to M3s. But for M1 users, in many cases, it's a reasonable value proposition.
Would I like lots more hard data on the M series chips? Definitely! But Apple has actually been pretty clear about the few numbers they provide, and experience has generally supported those numbers. Sometimes they even turn out to be conservative.
Absolutely, plain as day, the M3 was 2.5x faster than the ... ?
My post and Apple's M4 slide were comparing SoCs not iPads.
And it just happens to make the speed-up look better and quite hard to compare to the previous generation. But you stick to your opinion and Apple will no-doubt have some ocean-front property in Kansas for you soon.
Since you clearly know more about binning chips than I do, you know that chips get binned for lots of reasons, like when they don’t run at the rated speed, or require more power to get there. Not just bad cores. A pretty much every chip gets binned some way. These chips will go into iPads, these chips will go into macbook airs, ”what about these? they are quite horrible” “put them in a HomePod” don’t waste anything.
They could have the same number of working cores and be binned as a lessor chip.
Think the A10 came out with the iPhone 7 in Sept 16, whereas A10X was June 2017 at WWDC! Good times...
Playing games i get about 2 hours. 3 maaaaaaybe 4 watching YouTube videos.. still works great if i just use it plugged in lol.. I‘ve pretty much ditched my laptop and this is my daily driver so it’s worth the upgrade seeing as how this is a 2021 model and it’s really a 3 generation jump in processor. I have decided I don’t need to go max with the 1TB this time though, looking back, that was overkill.
Right, because basic algebra is soooo challenging.
M3/M1 speedup = X
M2/M1 speedup = Y
M3/M2 speedup = X/Y [because (M3/M1)/(M2/M1) multiplied by M1/M1 = M3/M2]