May 7 “Let Loose” Event - new iPads

[Andropov]

i am guessing that’s artificial and not due to fallout, but who knows.

If doing it to upsell users to more expensive tiers, I'd expect them to feature this more prominently, not in a somewhat hard to find portion of the tech specs. It's not even showcased in the configuration page like with Macs.

At this point I mostly disregard Apples own benchmark scores. It’s not that they lie, it’s that they are usually odd or poorly explained. Last time they said the M3 was around 10% faster single core and it turned out around 15-20% in Geekbench.

We can always treat them as a lower-bound of what to expect. Except for that 3090 comparison, which the Internet will never forget.

 

[Cmaier]

If doing it to upsell users to more expensive tiers, I'd expect them to feature this more prominently, not in a somewhat hard to find portion of the tech specs. It's not even showcased in the configuration page like with Macs.

Both true. But it just doesn’t make a lot of sense from a failure mode analysis. The P cores aren’t *that* big. Though. maybe it reflects failures in bigger structures like cache - if some of the cache is busted, then you allow fewer cores to access it at the same time so you don’t need as much?

 

[leman]

Both true. But it just doesn’t make a lot of sense from a failure mode analysis. The P cores aren’t *that* big. Though. maybe it reflects failures in bigger structures like cache - if some of the cache is busted, then you allow fewer cores to access it at the same time so you don’t need as much?

Apple has been using binning by power consumption in their mobile chips for a while. Maybe that’s what we see here? The binned chips could be those drawing more power, so they disable parts of them to still make the TDP envelope. With how thin that device is, and the two OLED layers, power efficiency must really be a concern.

 

[Cmaier]

Apple has been using binning by power consumption in their mobile chips for a while. Maybe that’s what we see here? The binned chips could be those drawing more power, so they disable parts of them to still make the TDP envelope. With how thin that device is, and the two OLED layers, power efficiency must really be a concern.

sure, could be that, though I wonder if it wouldn’t make more sense to allow all the cores but just use a more aggressive voltage/clock profile. In any case, Apole’s gonna Apple.

 

[rdrr]

Any reason not to get the nano-textured glass?

 

[Andropov]

Now that the N3E vs N3B is a hot topic again as the M4 has been released, does anyone know if what people are saying (that N3B is faster but more power hungry, while N3E is slower but more efficient) is true or an oversimplification? Trying to wrap my head around how that'd be the case

 

[dada_dave]

Now that the N3E vs N3B is a hot topic again as the M4 has been released, does anyone know if what people are saying (that N3B is faster but more power hungry, while N3E is slower but more efficient) is true or an oversimplification? Trying to wrap my head around how that'd be the case

N3E is less dense but faster and more efficient

TSMC Details 3nm Evolution: N3E On Schedule, N3P and N3X To Deliver 5% Performance Gains

www.anandtech.com

 

[Cmaier]

Any reason not to get the nano-textured glass?

harder to clean? it will affect sharpness/brightness a tiny bit? I just didn’t want to deal with the cleaning hassle.

 

[dada_dave]

I was wrong and Maxtech was right. A bitter pill to swallow. Now curious to see benchmarks.

Oh not so bitter, we get the M4 - besides I thought it was Gurman's prediction? I have to admit once @thenewperson shared that 9to5 mac article I leaned towards it really being the M4, though I still had a hard time totally believing it. I wonder if that benchmark leak you shared from MaxTech for the "A18pro" was actually for the M4, then again if the A18Pro exists, maybe it has a very similar CPU? We'll find out soon!

Strangely binned chips too: https://x.com/markgurman/status/1787864987110682921?s=46&t=AVo4Ae4rwcqD3xOi5WvXyg

Adding 2 Efficiency cores was something I thought they might do. I see the cut down M4 still has 8GB of RAM, hopefully the Macs will be 16GB (at least the MacBook Pro with M4 should be as the 16GB comes with the full M4, thank goodness - I mean we'll see for the Macs but that's what it should be). It is indeed odd though that they cut down a P-core and not GPU cores. What I find really interesting is that the Neural engine is still only 16 cores but much, much faster (we have to see at what precision of course, but then maybe that's part of it, more flexible precision!). @leman you mentioned patents for new neural engines?

Confused about what has or hasn’t improved vs the M3. Any ideas?

Edit: what concerns me is, if either cpu or gpu haven't improved, we have to wait another year probably.

They didn't mention any GPU feature changes (I was hoping for neural cores in the GPU) but it could have gotten a speed bump over M3 ... or not. They said the CPU was "all new" but the GPU "builds on the M3". Let's see if it's ARM v9 and if the new AMX processors are compatible with ARM's SME.

At this point I wouldn't be surprised if they jump the Mac straight to M5. Or maybe use M4 platform for the desktop. It seems that Apple is now committed to iterating faster, which makes sense for them. I think it is great for the customers.

Yeah all bets are off now. It's possible that we'll see M4 in the desktops at WWDC and M5's in the fall for the laptops. ... you know some of that is possible, not the M5, but maybe M4s ... I can't believe they'd have enough chips for minis or even worse Ultras and Extremes if those are coming but hey I didn't think they'd make the M4 available this soon either!

harder to clean? it will affect sharpness/brightness a tiny bit? I just didn’t want to deal with the cleaning hassle.

I wonder for those who are using it for note taking and drawing if it is an improvement - like those protectors to change the screen texture to make the friction more paper-like?

 

[casperes1996]

Adding 2 Efficiency cores was something I thought they might do. I see the cut down M4 still has 8GB of RAM, hopefully the Macs will be 16GB (at least the MacBook Pro with M4 should be as the 16GB comes with the full M4, thank goodness - I mean we'll see for the Macs but that's what it should be). It is indeed odd though that they cut down a P-core and not GPU cores. What I find really interesting is that the Neural engine is still only 16 cores but much, much faster (we have to see at what precision of course, but then maybe that's part of it, more flexible precision!). @leman you mentioned patents for new neural engines?

Does the base M4 not drop both a P core and a GPU core? That's what I read. A 9/9 instead of a 10/10.
But we've never seen that segmentation in iPad before, have we? With RAM sure, but SoC tiers too? It's especially weird cause it's so hidden behind the storage options

 

[Jimmyjames]

Oh not so bitter, we get the M4 - besides I thought it was Gurman's prediction? I have to admit once @thenewperson shared that 9to5 mac article I leaned towards it really being the M4, though I still had a hard time totally believing it. I wonder if that benchmark leak you shared from MaxTech for the "A18pro" was actually for the M4, then again if the A18Pro exists, maybe it has a very similar CPU? We'll find out soon!

I’d forgotten about that. That would be a nice increase.

 

[dada_dave]

Does the base M4 not drop both a P core and a GPU core? That's what I read. A 9/9 instead of a 10/10.
But we've never seen that segmentation in iPad before, have we? With RAM sure, but SoC tiers too? It's especially weird cause it's so hidden behind the storage options

No I think it's 10 GPU cores and 9 CPU cores for the cut down. But yeah I think this is all new. The association with storage tiers may possibly be trying to simplify logistics since the storage options are soldered to the motherboard (if so then seriously Apple could just adopt the slots they themselves use on the Studio/Pro and save themselves and us the headache). From Gurman:

 

[Edd]

Thinking one of the iPad Airs, tempted to go big. Pro is overkill for my needs. M3 in the Air would have been nice.

 

[casperes1996]

No I think it's 10 GPU cores and 9 CPU cores for the cut down. But yeah I think this is all new. The association with storage tiers may possibly be trying to simplify logistics since the storage options are soldered to the motherboard (if so then seriously Apple could just adopt the slots they themselves use on the Studio/Pro and save themselves and us the headache). From Gurman:

Almost definitely a logistics thing. Figure out roughly how many of each tier you need proportionally and just produce that. No built-to-order or anything.

Can't even find this on the buy page's small print at the bottom. Wild

 

[casperes1996]

Let's discuss this:

I'm curious about a few things:

• What needs does supporting the "Tandem OLED" display put on the display engine?
• What is meant by Brightness and color compensation?
• Is 10Hz a new low for ProMotion? Was lowest not 48Hz before? (On bigger screens; Watch/Phone could do 1hz or something I think)
• Is the IP block in the photo expected to be roughly to scale for how much space it takes up on the actual die? (I assume not)
• Expectations of this Display Engine supporting more external displays? Can iPad power 2 external + internal? What about in clamshell mode? (That's a joke.... I think....)

 

[theorist9]

Any reason not to get the nano-textured glass?

In comparing the nanotextured to glossy ASD's, I found the nano was less sharp for text. The difference was, for me, noticeable enough that I wouldn't want to use a nanotextured display for my work. In addition, the nanotexture creates a strong "sparkling snowfield" effect on white backgrounds that may or may not bother you (this is common with all matte-type coatings, but the effect with Apple's nanotexture is particularly strong). So if you're mostly viewing text, unless it's going to be used outdoors, I'd personally stay away.

OTOH, that sharpness reduction is not noticeable on photos and video, and it's better at killing reflections than any other surface treatement I've seen (which is probably why it also creates a strong sparking snowfield effect--it's a very strong texture). Another benefit is that other treatments seem to "dull" the screen, while the nanotexture does not. Thus if you're looking mostly at pictures or videos, and doing it in a bright environment, it would probably be superb. I assume that's whom the nano was originally designed for--people doing photo and/or video work in environments were lighting can't be easily controlled.

I didn't see the presentation, so perhaps this was addressed there, but I also wonder how nanotexture would work with a touch screen. Will it feel unpleasant to move your finger across the screen? Will there be more resistance? And willl the nanotexture remove microparticles from your skin, causing the screen to dirty more quickly than a glossy screen? "The nanotextured iPad: Now with free dermabrasion!"

EDIT: It just occurred to me that, given it's meant to be touched, it could be a significantly different type of nanotexture from that used in the ASD and XDR. So what I wrote above might not (fully) apply.

 

[Jimmyjames]

I’m curious about this...

Is this wider than the M3 or (more likely) the same?

 

[casperes1996]

I didn't see the presentation, so perhaps this was addressed there, but I also wonder how nanotexture would work with a touch screen. Will it feel unpleasant to move your finger across the screen? Will there be more resistance? And willl the nanotexture remove microparticles from your skin, causing the screen to dirty more quickly than a glossy screen?

It was not covered, but I wonder this too. And with how delicate clearning the nano texture can be, is that a good idea for a device you touch all the time?

 

[dada_dave]

In comparing the nanotextured to glossy ASD's, I found the nano was less sharp for text. The difference was, for me, noticeable enough that I wouldn't want to use a nanotextured display for my work. In addition, the nanotexture creates a strong "sparkling snowfield" effect on white backgrounds that may or may not bother you (this is common with all matte-type coatings, but the effect with Apple's nanotexture is particularly strong). So if you're mostly viewing text, unless it's going to be used outdoors, I'd personally stay away.

OTOH, that sharpness reduction is not noticeable on photos and video, and it's better at killing reflections than any other surface treatement I've seen (which is probably why it also creates a strong sparking snowfield effect--it's a very strong texture). Another benefit is that other treatments seem to "dull" the screen, while the nanotexture does not. Thus if you're looking mostly at pictures or videos, and doing it in a bright environment, it would probably be superb. I assume that's whom the nano was originally designed for--people doing photo and/or video work in environments were lighting can't be easily controlled.

I didn't see the presentation, so perhaps this was addressed there, but I also wonder how nanotexture would work with a touch screen. Will it feel unpleasant to move your finger across the screen? Will there be more resistance? And willl the nanotexture remove microparticles from your skin, causing the screen to dirty more quickly than a glossy screen? "The nanotextured iPad: Now with free dermabrasion!"

It was not covered, but I wonder this too. And with how delicate clearning the nano texture can be, is that a good idea for a device you touch all the time?

I was wondering if it might improve the Apple pencil experience, people pay a lot of money to add textured covers to their iPads to get a more paper-like writing surface.

I’m curious about this...
View attachment 29296

Is this wider than the M3 or (more likely) the same?

I was wondering the same, not clear. The only thing I can go on is that they described the CPU as "brand new" but the GPU as "building off the M3". So I lean towards a brand new CPU design, but that could just be video fluff or maybe a brand new design but still 9-wide decode. We won't know until launch.

 

[Cmaier]

I’d forgotten about that. That would be a nice increase.

around 17%? (assuming no additional bump from M4 being clocked higher than phone?) Or is my math wrong?