Apple M5 rumors

I don’t want to turn this into a Gurman thread and maybe it would be better to create a new thread, but as one final example of flipping a coin and calling both heads and tails.
1735845483171.png
1735845541472.png

Both unable to match the competition and also “so sophisticated”, it can turn a device into a pseudo router. Interesting way to take credit no matter what happens.
 
I am kind of curious how well a multi-die-multi-process package would work. If you put the TB/USB controllers on a N4 die, the CPU blocks, GPU blcoks and the other dedicated logic on their own N2 dies and glued them together, does the N2 grid have the right kind of resolution for the ultra-connect leads to line up for welding to the N4 die?

And the A-series are still going to be a monolithic die. Thus, the compound M-series devices will have to use the N4 library, which means they have to have two different layouts for the controllers part. Which is not beyond the pale but seems like not much cost savings.
 
I am kind of curious how well a multi-die-multi-process package would work. If you put the TB/USB controllers on a N4 die, the CPU blocks, GPU blcoks and the other dedicated logic on their own N2 dies and glued them together, does the N2 grid have the right kind of resolution for the ultra-connect leads to line up for welding to the N4 die?

And the A-series are still going to be a monolithic die. Thus, the compound M-series devices will have to use the N4 library, which means they have to have two different layouts for the controllers part. Which is not beyond the pale but seems like not much cost savings.
We don’t know that Apple will do multi-node yet, but I believe @leman has seen the patents for doing that. The cost savings comes from reuse, even across generations, and the fact that the older nodes are significantly cheaper than the most recent ones. Base M and A are indeed probably going to stay monolithic for now - are certainly rumored to be.
 
I am kind of curious how well a multi-die-multi-process package would work. If you put the TB/USB controllers on a N4 die, the CPU blocks, GPU blcoks and the other dedicated logic on their own N2 dies and glued them together, does the N2 grid have the right kind of resolution for the ultra-connect leads to line up for welding to the N4 die?

And the A-series are still going to be a monolithic die. Thus, the compound M-series devices will have to use the N4 library, which means they have to have two different layouts for the controllers part. Which is not beyond the pale but seems like not much cost savings.
the ultra connect pads are much bigger than the on-die wire widths. And there isn’t any sort of snap-grid. Won’t be a problem.
 
I am kind of curious how well a multi-die-multi-process package would work. If you put the TB/USB controllers on a N4 die, the CPU blocks, GPU blcoks and the other dedicated logic on their own N2 dies and glued them together, does the N2 grid have the right kind of resolution for the ultra-connect leads to line up for welding to the N4 die?

Here you go


From my amateur-level understanding, this side-steps the challenge of aligning “tiles” by routing the connection through an interposer. Basically you partially overlay dies and connect them to a shared routing component. Makes for a more compact package too.

And the A-series are still going to be a monolithic die. Thus, the compound M-series devices will have to use the N4 library, which means they have to have two different layouts for the controllers part. Which is not beyond the pale but seems like not much cost savings.

The calculation on these things are probably complex. Maybe they have the tooling to make this less of a pain.
 
I don’t want to turn this into a Gurman thread and maybe it would be better to create a new thread, but as one final example of flipping a coin and calling both heads and tails.

Both unable to match the competition and also “so sophisticated”, it can turn a device into a pseudo router. Interesting way to take credit no matter what happens.

But how sophisticated is that really? Isn't that effectively the same as when I create a local hotspot for my cellular connection on my phone so I can connect to it through Wi-Fi on my MacBook?
 
But how sophisticated is that really? Isn't that effectively the same as when I create a local hotspot for my cellular connection on my phone so I can connect to it through Wi-Fi on my MacBook?
I’m not sure I understand. I’m not claiming it is sophisticated. I’m pointing out that it’s odd to claim both that the new chip won’t match current capabilities and that it’s sophisticated. I don’t see how those two align. If a chip with fewer capabilities than the current one is sophisticated, then everything is sophisticated, rendering the word pointless.

We agree I hope that the current WiFi chips in Apple devices can’t act as routers or WAPs. Why would less capable ones be able to?
 
I’m not sure I understand. I’m not claiming it is sophisticated. I’m pointing out that it’s odd to claim both that the new chip won’t match current capabilities and that it’s sophisticated. I don’t see how those two align. If a chip with fewer capabilities than the current one is sophisticated, then everything is sophisticated, rendering the word pointless.

We agree I hope that the current WiFi chips in Apple devices can’t act as routers or WAPs. Why would less capable ones be able to?
I didn't mean for it to sound like a response to anything you'd claimed. I was just asking how sophisticated it really is to add that functionality if desired. I agree with the poking fun at the statement, a lot of Apple rumours hedge their bets and claim contradictory things early on, or make trivial and pointless predictions like "future iPhones to include faster chips". I was just using the statements as a springboard to say "but is that really so hard to achieve if desired?"
 
I didn't mean for it to sound like a response to anything you'd claimed.
No problem. It can be awkward to communicate via text, for me at least, as tone is difficult to articulate.
I was just asking how sophisticated it really is to add that functionality if desired.
Honestly I don’t know. If Gurman is trying to say something like an Apple TV could act as a wireless router or access point, I think that’s quite a bit more difficult. AFAIK access points use more powerful wifi chips with the capability to do things like wireless backhaul in the case of mesh routers. It’s intriguing to be honest.
 
No problem. It can be awkward to communicate via text, for me at least, as tone is difficult to articulate.

Honestly I don’t know. If Gurman is trying to say something like an Apple TV could act as a wireless router or access point, I think that’s quite a bit more difficult. AFAIK access points use more powerful wifi chips with the capability to do things like wireless backhaul in the case of mesh routers. It’s intriguing to be honest.

Sure, but none of that is "new", right? It's known tech. In principle, Apple could shove all the mesh access point hardware into an Apple TV + all the hardware already in there and have an TV+mesh AP in one box that may be bigger and draw more power but is also connected to an outlet at all times anyway. Mix with HomePods doing the same thing; I can see that being attractive to the handful of homes that are in the niche the AppleTV targets. Less niche now than in its first 2-3 iterations, but I suspect it's once again getting more and more niche and looking for more purpose as TVs come out of the box with so much similar functionality even if lower quality in a lot of metrics
 
Sure, but none of that is "new", right? It's known tech. In principle, Apple could shove all the mesh access point hardware into an Apple TV + all the hardware already in there and have an TV+mesh AP in one box that may be bigger and draw more power but is also connected to an outlet at all times anyway. Mix with HomePods doing the same thing; I can see that being attractive to the handful of homes that are in the niche the AppleTV targets. Less niche now than in its first 2-3 iterations, but I suspect it's once again getting more and more niche and looking for more purpose as TVs come out of the box with so much similar functionality even if lower quality in a lot of metrics
Depends on what is meant “new” I suppose. If Proxima is able to combine the power of a WAP WiFi chip while meeting Apple’s requirements for low power, I’d say it’s quite new. They can’t do it currently, which again points out Gurman’s confusion. I don’t think they are going to make more than one WiFi chip, so I imagine it would have to meet all requirements, including low power.
 
Depends on what is meant “new” I suppose. If Proxima is able to combine the power of a WAP WiFi chip while meeting Apple’s requirements for low power, I’d say it’s quite new. They can’t do it currently, which again points out Gurman’s confusion. I don’t think they are going to make more than one WiFi chip, so I imagine it would have to meet all requirements, including low power.

Right; I was not thinking a one chip solution really. I was just thinking of the product, using other people's chips, and Proxima being outside that product category for now
 
Right; I was not thinking a one chip solution really. I was just thinking of the product, using other people's chips, and Proxima being outside that product category for now
Fair enough. I’m quite intrigued by the idea that Apple could create a mesh network from their devices. It would be great if that’s something they are able to do and would definitely be “sophisticated” I think. After the M5 gpu, which I am hoping for big things from, this new wifi/bluetooth is the most interesting Apple thing this year for me.
 
Fair enough. I’m quite intrigued by the idea that Apple could create a mesh network from their devices. It would be great if that’s something they are able to do and would definitely be “sophisticated” I think. After the M5 gpu, which I am hoping for big things from, this new wifi/bluetooth is the most interesting Apple thing this year for me.
I think the problem is that in general a mesh network will require more power than a hub and spike topology because each device is connecting to multiple other devices. That’s not really Apple’s thing with so many of their devices relying on battery power.

More generally I think those hoping that Apple’s push into modems is about innovation are going to be disappointed. Both their WiFi/bluetooth project and the cellular modem project are much more about replacing Broadcom and Qualcomm. It’s about saving money on the BoM. I don’t see a ton of opportunity for innovation in the space.
 
I think the problem is that in general a mesh network will require more power than a hub and spike topology because each device is connecting to multiple other devices. That’s not really Apple’s thing with so many of their devices relying on battery power.

More generally I think those hoping that Apple’s push into modems is about innovation are going to be disappointed. Both their WiFi/bluetooth project and the cellular modem project are much more about replacing Broadcom and Qualcomm. It’s about saving money on the BoM. I don’t see a ton of opportunity for innovation in the space.
Perhaps, although I definitely think there is room for things like improving Airdrop, Continuity Camera etc.
 
It’s sharing a cellular connection over wifi. A wireless mesh access point connects to multiple wifi points simultaneously afaik.
From my experience with OpenWrt, AP and STA mode are different. And if an interface can act as AP, and STA, both can work simultaneously. And if I'm not wrong, Apple uses BCM WiFi and those are usually fully 802.11 compliant and can be used in routers or mesh solutions.
 
Perhaps, although I definitely think there is room for things like improving Airdrop, Continuity Camera etc.
In what ways? Lowering latency? Increasing speed? That all seems like standard incremental improvement stuff more so than any Apple magic sauce.

That said, there’s another ‘next gen’ RF chip rumored for next year which I think might be a springboard for some Apple Magic - the second gen UWB chip. I could imagine UX for AirDrop that’s essentially ‘point your phone directly at someone else’s phone and flick whatever is on screen to them.’

More generally, precision location opens up a lot of opportunities for new UX. And who knows, maybe there’s some synergy between UWB and beam forming for WiFi/bluetooth…
 
Back
Top