Apple Intelligence "can't" run on an iPhone 15 (with an A16 chip) but can on an M1 chip with a much weaker NPU. The only explanation is money.

[twintin]

Not sure where this 8 GB memory requirement comes from, but they have a specific model for devices like older iPhones with less RAM where they use the flash memory instead. I believe there is a possibility the iPhone 15 Pro restriction is just the first step before more older iPhones get AI. We will see in due time.

Apple Develops Breakthrough Method for Running LLMs on iPhones

Apple GPT in your pocket? It could be a reality sooner than you think. Apple AI researchers say they have made a key breakthrough in deploying large...

www.macrumors.com

 

[Ansath]

Sorry to ruin all your research but its all about RAM

Google got a HUGE backlash for announcing Gemini nano for the Pixel 8 Pro because the Pixel 8 and 8a also had 8GB RAM which is needed for on device AI.

They quickly changed course and now it's available on all 3 phones and the Pixel 9 series of phones will have 16GB RAM.

Apple did not plan this well at all and the lack of AI on the 15 and 15 plus is proof they were never going to release AI on this years phones, they were definitely pushed into announcing this earlier than they wanted.

It's been known for a while that you need 8GB RAM minimum and even the 16 series is still on 8GB RAM according to leaks, expect the 17 series to have 12GB.

Basically, Apple have been disgustingly greedy for years with RAM and now its a slap in the face of 15 owners, they deserve far more media scrutiny over this but then tech media is scared of not getting invites, t shirts, pins and selfies with Tim.

A few of us have been saying it's all about RAM since early on in the thread, but people keep disregarding/dismissing it, which sucks. Though I think that maybe they had hoped to get it into a position where it could work on devices with 6GB RAM, but just couldn't meet that goal. Maybe they will in the future - it's only in Beta when iOS 18 launches in the Fall, and features in beta in the past have had long runs being in beta phase to the consumers. We'll just have to wait and see.

I also have said Apple should have, at bare minimum, put 8GB RAM in all the phones last year. It's hugely poor planning on their part.

 

[twintin]

A few of us have been saying it's all about RAM since early on in the thread, but people keep disregarding/dismissing it, which sucks. Though I think that maybe they had hoped to get it into a position where it could work on devices with 6GB RAM, but just couldn't meet that goal. Maybe they will in the future - it's only in Beta when iOS 18 launches in the Fall, and features in beta in the past have had long runs being in beta phase to the consumers. We'll just have to wait and see.

I also have said Apple should have, at bare minimum, put 8GB RAM in all the phones last year. It's hugely poor planning on their part.

Or maybe it is not about RAM. Maybe iPhone 15 Pro users are just early adopters before it comes to the rest of us. At least one may think so after reading the article below.

Apple Develops Breakthrough Method for Running LLMs on iPhones

Apple GPT in your pocket? It could be a reality sooner than you think. Apple AI researchers say they have made a key breakthrough in deploying large...

www.macrumors.com

 

[Ansath]

Or maybe it is not about RAM. Maybe iPhone 15 Pro users are just early adopters before it comes to the rest of us. At least one may think so after reading the article below.

Apple Develops Breakthrough Method for Running LLMs on iPhones

Apple GPT in your pocket? It could be a reality sooner than you think. Apple AI researchers say they have made a key breakthrough in deploying large...

www.macrumors.com

Yea, I have said a few times earlier on this thread and others that it could be that once it comes out of beta, it might expand to devices that have 6GB RAM, including in the post of mine you quoted.

 

[twintin]

Yea, I have said a few times earlier on this thread and others that it could be that once it comes out of beta, it might expand to devices that have 6GB RAM, including in the post of mine you quoted.

It’s just that people have got hung up on 8 GB RAM as a requirement for some reason when in reality it may not be the case. Apples planning maybe is not that bad as one may think. 😊

 

[redbeard331]

Sorry to ruin all your research but its all about RAM

Google got a HUGE backlash for announcing Gemini nano for the Pixel 8 Pro because the Pixel 8 and 8a also had 8GB RAM which is needed for on device AI.

They quickly changed course and now it's available on all 3 phones and the Pixel 9 series of phones will have 16GB RAM.

That will be interesting to see if they get 16, people thought I was crazy for hoping Apple would double the ram for the 16 pro’s to 16 gb as well. If what people are saying is true, and 8gb is the minimum needed, wouldn’t that mean the 15 pro’s will be really slow once the ai rolls out?

Apple did not plan this well at all and the lack of AI on the 15 and 15 plus is proof they were never going to release AI on this years phones, they were definitely pushed into announcing this earlier than they wanted.

Which has me wondering if they decided to make a significant ram upgrade this year for the 16’s? It’s already in the iPad Pro 1 and 2 tb models. How much could it actually cost them? I myself would pay more for a phone with more ram, every time they’ve upped the ram I’ve noticed a huge increase in all around performance.

It's been known for a while that you need 8GB RAM minimum and even the 16 series is still on 8GB RAM according to leaks, expect the 17 series to have 12GB.

Basically, Apple have been disgustingly greedy for years with RAM and now its a slap in the face of 15 owners, they deserve far more media scrutiny over this but then tech media is scared of not getting invites, t shirts, pins and selfies with Tim.

I wonder if those leaks were based on early information before Apple decided to rush their AI? I can’t imagine paying 1500+ for a brand new phone in September that’s going to be hamstrung on day 1. I’m really looking forward to the pixel 9 pro, I just don’t trust Apple to push their hardware to where it should be.

 

[NT1440]

That will be interesting to see if they get 16, people thought I was crazy for hoping Apple would double the ram for the 16 pro’s to 16 gb as well. If what people are saying is true, and 8gb is the minimum needed, wouldn’t that mean the 15 pro’s will be really slow once the ai rolls out?



Which has me wondering if they decided to make a significant ram upgrade this year for the 16’s? It’s already in the iPad Pro 1 and 2 tb models. How much could it actually cost them? I myself would pay more for a phone with more ram, every time they’ve upped the ram I’ve noticed a huge increase in all around performance.




I wonder if those leaks were based on early information before Apple decided to rush their AI? I can’t imagine paying 1500+ for a brand new phone in September that’s going to be hamstrung on day 1. I’m really looking forward to the pixel 9 pro, I just don’t trust Apple to push their hardware to where it should be.

I don’t understand how anyone can look at Apple Intelligence, which clearly had years of internal development, and think that Apple “rushed”. I get that was the idea from the tech press (that Apple was “behind” somehow), but now that it’s announced you’re looking at 5+ years of development and thinking this was slapped together? Come on.

 

[Ansath]

It’s just that people have got hung up on 8 GB RAM as a requirement for some reason when in reality it may not be the case. Apples planning maybe is not that bad as one may think. 😊

Yea, we were specific about the RAM for it in its current state. It’s probably good to keep that as baseline expectation, and be clear that maybe it going to devices that have 6GB RAM as being a speculative possibility, as some people might get their hopes up.

The Android devices, they’re only running on 8GB devices too, so it does seem to be that might be the lower limit for the time being.

 

[redbeard331]

I don’t understand how anyone can look at Apple Intelligence, which clearly had years of internal development, and think that Apple “rushed”. I get that was the idea from the tech press (that Apple was “behind” somehow), but now that it’s announced you’re looking at 5+ years of development and thinking this was slapped together? Come on.

Just going by what we’ve heard, in that case doesn’t that mean there’s no excuse for a sizable ram upgrade this year?

 

[NT1440]

Just going by what we’ve heard, in that case doesn’t that mean there’s no excuse for a sizable ram upgrade this year?

We’re likely getting that anyway. Apple wants Apple Intelligence to be a business differentiator for their ecosystem. It’s coming to everything it can over the next few product cycles.

 

[vigilant]

I probably distracted everyone with the NPU mention. As someone else pointed out, the common factor between the 15 Pro and the M1+ Macs is 8+ GB of RAM.

However, that doesn't change the fact that Apple said much of the AI processing happens off-device…which means on-device compute power isn't always even needed.

Developers.

iPhone-driven sales (hardware + software) is the largest revenue source for Apple. They're not making a huge chunk of their profits from Mac sales. They want devs to use AI and figure out how to add AI features to their apps (they alluded to this in the keynote). Getting more AI features by developers who use Macs to make iOS apps makes more monetary sense to push AI features to the iPhone faster.

Meanwhile, pushing more people to get newer iPhones pushes that market segment a lot more. The iPhone is the central device of Apple for right now.

It’s worth pointing out that it’s worth reviewing the wording. I don’t recall them saying “most” or “much”. I remember them saying that Private Cloud Compute would be able to do some more complex things, with the minimal amount of data possible to complete the task and it wouldn’t be saved. Examples like “When is my mother’s flight landing” would require an intrusive level of data to be uploaded to get that response, let alone find the lunch spot from messages.

Things that make sense, and require Personal Context from on device are perfectly suited for on device work.

 

[vigilant]

I think this remains to be seen, because they also said you can use your plus account and all those features....well those features have to be stored on OpenAI servers and linked to your account, your CC or payment options. So if you upload an image and have an openAI plus account, I'm 100% sure it will be linked to your OpenAI account. Now if Apple or OpenAI offer memory features to be stored on device and uploaded encrypted to OpenAI servers, OpenAI will get 50 million plus account users.

So I think a lot is still unknown, but it looks very promising, without Apple privacy and security would not be a thing with other companies, they only try to compete with Apple. That is my #1 reason I like Apple, privacy, choice of what I allow to be used by apps.

Theres a lot of semantics that need to be worked out for sure. I do expect Apple to use iCloud Private Relay, and probably work with OpenAI to build a device specific “Pro” token that would be linked against it. How they keep the information private is going to be an interesting thing to see from a technical implementation standpoint.

 

[Ansath]

It’s worth pointing out that it’s worth reviewing the wording. I don’t recall them saying “most” or “much”. I remember them saying that Private Cloud Compute would be able to do some more complex things, with the minimal amount of data possible to complete the task and it wouldn’t be saved.

Apple have said that most, and almost all will be on device, in the keynote, in the SOTU, and in interviews.

 

[Ansath]

Theres a lot of semantics that need to be worked out for sure. I do expect Apple to use iCloud Private Relay, and probably work with OpenAI to build a device specific “Pro” token that would be linked against it. How they keep the information private is going to be an interesting thing to see from a technical implementation standpoint.

In interviews, they said that using the free access to ChatGPT will be more private, hidden IP addresses, etc. they said that if people connect their pro account, and use that, then that will have some connection for the interactions that use the pro account they connect - which makes sense, as your account is a connection. The free access remains private.

 

[shenfrey]

Ram

 

[Tagbert]

I probably distracted everyone with the NPU mention. As someone else pointed out, the common factor between the 15 Pro and the M1+ Macs is 8+ GB of RAM.

However, that doesn't change the fact that Apple said much of the AI processing happens off-device…which means on-device compute power isn't always even needed.

Developers.

iPhone-driven sales (hardware + software) is the largest revenue source for Apple. They're not making a huge chunk of their profits from Mac sales. They want devs to use AI and figure out how to add AI features to their apps (they alluded to this in the keynote). Getting more AI features by developers who use Macs to make iOS apps makes more monetary sense to push AI features to the iPhone faster.

Meanwhile, pushing more people to get newer iPhones pushes that market segment a lot more. The iPhone is the central device of Apple for right now.

Apple has stated that they expect most of the AI work to be done on-device with a minority of it passed to the PCC(Private Cloud Compute) servers. I imagine that if they opened up AI to devices that could not run on-device, they would suddenly have a lot more traffic on their PCC servers. With the scale of the iPhone market that probably would exceed what they can handle on those servers.

Maybe they will eventually get enough capacity built out to handle expanding the support to other phones. I am curious what the plans are for low power devices like Watch and HomePod. Are they going to start putting 8GB in those or will they be able to just use the PCC servers directly?

 

[Ansath]

Apple has stated that they expect most of the AI work to be done on-device with a minority of it passed to the PCC(Private Cloud Compute) servers. I imagine that if they opened up AI to devices that could not run on-device, they would suddenly have a lot more traffic on their PCC servers. With the scale of the iPhone market that probably would exceed what they can handle on those servers.

Maybe they will eventually get enough capacity built out to handle expanding the support to other phones. I am curious what the plans are for low power devices like Watch and HomePod. Are they going to start putting 8GB in those or will they be able to just use the PCC servers directly?

That and maybe they can improve on the LLM regarding RAM, could help expand down the line.

 

[aevan]

They are making the Non-Pro iPhone even less attractive to users. Who knows what may not be available on the regular iPhone 16 with iOS 19

Regular iPhone 16 will run AI 100%. Apple wants AI on as many devices as possible. It could be the RAM or something else, comparing iPhones to Macs is silly.

 

[aevan]

The iPhone 12 is just as capable of running Apple Intelligence as an M1 Pro Mac.

The engineering expertise on this forum is incredible. Apple engineers should come here regularly to take notes.

 

[contacos]

Regular iPhone 16 will run AI 100%. Apple wants AI on as many devices as possible. It could be the RAM or something else, comparing iPhones to Macs is silly.

That’s why I said by iOS 19 😉

 

[LogicalApex]

The engineering expertise on this forum is incredible. Apple engineers should come here regularly to take notes.

I am sure the suits guide more than the engineers. Apple Engineers would have ended the criminally low RAM amounts years ago.

 

[PeLaNo]

The 8GB explanation would mean that all iPhone16 models will have 8GB. I find that hard to believe.

If Apple “intelligence” doesn’t run on all iphone 16s, it would be a travesty. So something is stinkawith.

I know that Apple usually very stingy about the RAM situation but they are not dumb.

They know that this AI feature will make lots of people upgrade so they should willing to go for a bit lower margin (heck I bet it’s not that different, even in the grand scheme of things)

 

[PeLaNo]

As a base iPhone 15 user, I would rather have the AI go to the servers and actually be useful rather than spend a lot of money so I can get away from an idiotic version of Siri. The RAM shortage is just bad foresight on Apple's part in my own personal opinion.

I don’t want to sounds like Apple shills but if the rumors that Craig just realized that Apple need Apple intelligence to compete with Microsoft Co-pilot AFTER he tried it out then it seems like Apple just didn’t planned this before and just caught off guard by this.

I hope this marks the end of their stinginess with RAM and that they will now prioritize future-proofing iPhone more

 

[PeLaNo]

That sucks. But thank you! One more question: Would you recommend an M3 Macbook Air for gaming? Especially since the Macbooks have all the AI now, which could be useful for coding and etc.

If you’re not in a rush just wait for M4 tbh. It’s substantially more powerful than M3 and NPU is twice as fast so it could future proof

 

[Ansath]

I know that Apple usually very stingy about the RAM situation but they are not dumb.

They know that this AI feature will make lots of people upgrade so they should willing to go for a bit lower margin (heck I bet it’s not that different, even in the grand scheme of things)

Dead on