Having an iPad Pro with an M1, I grow increasingly frustrated each year with how incapable my iPad is compared to my Macbook/Linux laptops. iPadOS still feels like iOS with some visual tweaks to the larger screen real estate. It’s absolutely not a laptop killer for a sysadmin or developer (unless your preferred language is Swift). Maybe it’s a laptop killer for creative professionals?
I think my iPad is an excellent thin client, however — but a Pro device should be more than just a thin client for me, imo.
Some things that would be a huge improvement for me:
1. Ability to do development locally
2. Virtualization (might circumvent the need to do local dev)
3. Install/build local applications
Apple seems determined to never allow anything outside of the App Store on i[Pad]OS, and with iPadOS being over four years old and still feeling like a giant iPhone, I don’t think I could recommend anyone should buy an iPad as a power user device.
Same. Don't buy an iPad! For real. It's a ridiculous device. Effectively no more useful than any other cheap tablet. Those tempting hardware features, you won't be able to use them in the end anyway, because of software limitations. You just can't do serious work with the iPad in practice. Trust me, after a short honeymoon period, and another few weeks of frustration, it will end up collecting dust and you will only ever rarely use it to watch movies at most.
If you buy one still, don't forget to turn it off completely, or you will find it someday, the battery completely drained and you realize it must have been months. Not good for battery health.
Yes, of course if you have a concrete, specific significant use case for the pencil it may be more appealing. But the pre M1 iPad with the pencil did the job as well, for drawing and pure handwriting. Obviously the iPad is marketed to be more than a single purpose device. The processing power is completely wasted. Most people probably won't use the pencil much in the long term.
It's file management, app limitations and interoperability which kill usefulness.
I love the iPad for its stylus, it’s a great device for brainstorming even with just notes. I wish there was more creative software they took advantage of the form factor however, like a touch/stylus based programming environment.
It is marketed to be more than a drawing device tho. It fails at anything but being a glorified scratchpad. Very narrow usefulness.
I got the first iPad (2016?) with pencil, for 400€, I think. It still does completely fulfill my needs for brainstorming. For anything else, the app ecosystem and locked down system are useless.
I agree that a lot could be improved on the iPad and I would not say no to the features you suggest. But I’m not sure I understand why you would want to use an iPad instead of a laptop as a developer. For me the whole thing with the iPad is the pencil. I would never use it for development tasks (maybe like some sort of emergency device while traveling).
Not saying you are wrong to want to use it for development, just curious why!
It would be the last device I ever need, if it did work like that... It actually hurts me, how its limitations are completely arbitrary and artificial. The iPad could be the best thing ever, yesterday. So, so frustrating and disappointing.
> simply because a terminal-driven workflow via SSH is at least possible
This has been the one thing that has kept my iPad useful! I use Blink (grandfathered into their Legacy Pro plan) and it’s a very clean experience. I can’t help but feel like it’s a fragile experience having to rely on an app from the app store for my iPad to be useful — that Apple or the Developer could remove at any time. Maybe it would feel less fragile if Apple provided an official i[Pad]OS Terminal/SSH client.
1. N3B TSMC node used by M3 is reportedly low yielding. Apple likely wants to move away from it asap and move to N3E, which is the higher yielding node.
2. Rumors are that M3 Ultra is not planned due to low yields. M3 Max die shot does not have UltraFusion silicon. This supports the first theory.
3. Apple base the M chips on the A series cores. A18 Pro has already finished in design because Apple always starts making A series chips in Spring before a Fall debut. Therefore, A18 Pro is already in production using N3E node. Given enough early planning, it isn't surprising that M4 can be released early.
4. Reportedly, Apple plans to focus on GenAI at WWDC. They're also behind in GenAI. In order to catch up, they need to be aggressive. M4 is likely to have a big improvement in inference speed via the NPU and/or GPU.
I think 4 is underrated. From what I've seen Apple is making a play for on device inference. Gains on better screen, thinner phone, better camera selling points are pretty incremental now. But with on-device inference they now have access to: doubled the number, no cloud we care about privacy, on device translation capability, etc.
Think that whole major, unpatchable security flaw thing could also play a part? I could see them wanting to put distance between themselves on that fast too.
M3s have a flag to disable the caching vulnerable to GoFetch on p-cores. M1s and M2s are the ones that have no mitigation besides running workloads on slower e-cores.
That‘s not true DMP can be disabled on M2 chips and probably also on the M1 chips. This was found by Hector Martin (@marcan) and will probably supported in Asahi Linux.
I am a physicists (theory), and about a third of my work is “pencil and paper” calculations (and related notes) on the iPad Pro. I use it professionally ~20 hours per week.
May I ask if they are direct paper to screen transfers, or if there are specific features/apps that add extra value? I do a lot of work in (paper) engineering notebooks, and I'm curious about what improvements your workflow has.
I ~exclusively use ZoomNotes, which is very customizable but is quirky and comes with a steep learning curve.
For me the main benefits of the tablet are:
- Copy/paste/drag. Pretty helpful when you manipulate long expressions, or want to condense a bunch of scratch work.
- Relatedly: easy erase. When I used real paper, pencil was too low contrast, and “erasable pen” never worked well. So I used a real pen and just crossed tons of stuff out; ugly.
- Keep a decade of notes with me wherever I go. (And cloud backed up so I won’t lose them.)
- Infinite zoom and no page boundaries. Seems silly, but it’s really nice to write as big as I want without worrying about running out of room on the page.
- Text search of handwritten notes. Works surprisingly well.
- Easily switch to different colors. 99% of the time I only use 2 or 3, but I find it helpful to visual distinguish the main argument/computation from “side commentary”.
Undo is one thing you get used to very quickly. Copy paste, select and move, change colors, erase without a trace... digital handwriting has a lot to offer.
However, the app ecosystem is lacking in terms of interoperability, feature completeness within one app and export/import capabilities, and there is nothing you can do about it in iPadOS. File management is pure pain unless you fully bought into Apple already.
It's very nice for brainstorming and making transient personal illustrations while learning and working through technical problems, but you won't get the feel and certainty of a physical notebook as an institution for persistent notes.
The iPad is way too expensive for what it has to offer, in practice.
That said, without its artificial limitations it would be the last computing device you'll ever need. It could be everything.
Strong agree with the first two paragraphs. I have spent >10 working hours per week for past 4 years with an app I paid $7 for. I wish so much I could pay $7,000 for an improved version.
Disagree on permanence. My notes are useful for much longer when they can be electronically sorted, searched, backed up, and when they take up no additional space in my backpack.
Which doesn't mean parent is wrong. You can do any work on non-ideal equipment and be happy with it for various reasons.
When I saw what one of my ex was using (bio chemist growing various stuff and testing treatments to rare sickle cell anemia that pharma didn't care about due to low patient count), it was very subpar and inefficient but she was happy with it and got results nevertheless eventually. Some of her work I could make more efficient by 10x, some by 50x maybe, but ie she as hardcore unix person just plainly refused to use MS Excel and all its statistics power, to pick a random example (there were quite a few).
Not saying this is your case of course, but oh boy do small screens, limited browsers, limited power and storage speed etc. do have negative effect on work itself, mental flow, staying focused etc. There is a reason not only we IT guys want powerful machines and massive screens, it ain't just chasing shiny higher numbers.
> These are very much pro. That they're extremely functional for art is the whole point.
I'd argue that iPads being great for "art" is false. They have one killer app for one narrow specialized area (Procreate/illustration) but they aren't widely used as a main device professionally outside of that context (they are often used as a satellite device). They have adequate support for hobbyist level video editing/audio production/photography. They have minimal to non-existent support for motion graphics or anything related to 3D. Calling a device like this "extremely functional for art" I don't think is a defensible statement (unless it's specifically scoped to illustration).
In the concert hall I see more and more musicians & singers replacing paper music sheets with 12” iPad Pro’s. Mostly individual artist but recently all members of a cappella choir. Some, eg the violinists, use a small device on the floor probably connected by Bluetooth to advance to the next page with their foot. I don’t remember seeing a conductor with the full score on an iPad though, maybe one day with a foldable 2x20+”.
Art professionals usually don't use iPads although I'm sure some do. Most seem to prefer Wacom. Unlike, eg a laptop, portability isn't usually a major professional need for an art workstation.
To the contrary, art professionals use the top-tier iPads professionally more than anyone else, in my experience.
You use a Wacom for some things, but the iPad has the gigantic advantage that your image is directly underneath your Pencil. And, of course, that you can take it with you.
The iPad isn't a substitute for a workstation. But it's a huge complement -- both for creating things visually, as well as managing/showcasing all your media on the go. Professional artists are often going to a lot of meetings with different clients or team members around the city.
Wacom makes Centiqs which are very popular pen on screen monitors if you prefer that style, so the iPad doesn't bring much there.
I would say the vast majority of artists do not travel and do not need to draw mid-meeting. Certainly there are exceptions but they are just that, exceptions.
Centiqs may be great, but I've seen 100 artists using an iPad for every 1 artist I've seen using a Centiq.
I would say the vast majority of artists do have client meetings, whether they're employed full-time or freelance.
And artists need to work on the go just like the rest of us. Whether it's a dad trying to do 30 min of video editing in his car while he waits for daycare to get out, or somebody in the city doing an hour of website design in a cafe while they wait for their next meeting.
I'm baffled by your assertion that artists don't use iPads or shouldn't need them. It's just not the case, if you work with professional artists.
Interesting, I've worked in large game studios with many 2d and 3d digital artists and they just don't use iPads. It could be a specific industry thing or a regional thing but even during the early hybrid work days I would see artists haul in and pack up their Centiqs. IPads were not in the conversation.
I will say that it's a major disconnect in our experience that you would say the majority of artists are directly client facing.
What company/industry is buying their artists iPads?
iPad Pros are often used in music production too. There’s plenty of heavier duty non-casual sitting on the couch watching TV type use cases it’s used for and it’s targeted at.
Speaking of the Mac Studio M4, Apple is considering having it support up to a half-terabyte of memory. In comparison, the existing Mac Studio and Mac Pro top out at 192GB of unified memory (RAM).
(according to Tom's guide). I wonder if it will still be on-chip.
Not sure for the downvote, but here is an excerpt for the M3 MBP from the link above:
> One display up to 4K at 144Hz over HDMI
The other M series have HDMI with similar speeds as long as it isn’t driving 8K.
Thunderbolt seems to be limited to 6K and 60Hz, but HDMI isn’t similarly limited (well, it can only do 60Hz with 8K). Hopefully there will be a thunderbolt successor they fixed the problem.
At first I thought, it sounded crazy but then I realized that reports have told us that the M4 will be a key part of Apple’s reneweed AI strategy for 2024. The A18 and M4 in the iPhone 16 and the M4 MacBook Pros at the ens of this year will have much better neural engines to drive Apple’s WWDC 2024 AI features. Apple can’t leave the iPad Pro out to dry.
They already do this with fuses to lock each chip to a specific iBoot, OS, and release[0] set.
Changing the configuration of the chip is unnecessary. iPhones, iPads, and Macs use roughly the same balance of compute. The thing that does need to change is how many cores in each compute device, but they already do this with A17 vs. M3 vs. M3 Pro/Max/Ultra. That already segregates the product lines a bit: iPads aren't getting anything bigger than an M3/M4, Mac Pros aren't getting anything smaller than an M3 Max, etc.
[0] e.g. production devices vs. Apple internal devkits vs. that weird Security Research Device they released a few years ago to compete with Corellium
My guess is that M4 and forwards will not be so much about CPU and GPU performance but have some sort of local LLM in ROM. So every year you should upgrade to get the best hardware LLM ever released.
iPhone/iPad 16GB RAM + 64GB hardware local LLM ROM. Apple can phrase it as both a benefit and a reason to upgrade and reduce its costs over suddenly bumping RAM by 400%.
Out of all the hardware manufacturers, Apple is the worst candidate to pin planned obsolescence on. Its devices are supported longer than its competitors, and they get better at lasting longer over the years.
I don't know if mask ROM even exists any more or what its density would be. Realistically your choices are DRAM or flash and flash will probably never be fast enough to hold model weights.
Having an iPad Pro with an M1, I grow increasingly frustrated each year with how incapable my iPad is compared to my Macbook/Linux laptops. iPadOS still feels like iOS with some visual tweaks to the larger screen real estate. It’s absolutely not a laptop killer for a sysadmin or developer (unless your preferred language is Swift). Maybe it’s a laptop killer for creative professionals?
I think my iPad is an excellent thin client, however — but a Pro device should be more than just a thin client for me, imo.
Some things that would be a huge improvement for me:
1. Ability to do development locally
2. Virtualization (might circumvent the need to do local dev)
3. Install/build local applications
Apple seems determined to never allow anything outside of the App Store on i[Pad]OS, and with iPadOS being over four years old and still feeling like a giant iPhone, I don’t think I could recommend anyone should buy an iPad as a power user device.
Same. Don't buy an iPad! For real. It's a ridiculous device. Effectively no more useful than any other cheap tablet. Those tempting hardware features, you won't be able to use them in the end anyway, because of software limitations. You just can't do serious work with the iPad in practice. Trust me, after a short honeymoon period, and another few weeks of frustration, it will end up collecting dust and you will only ever rarely use it to watch movies at most.
If you buy one still, don't forget to turn it off completely, or you will find it someday, the battery completely drained and you realize it must have been months. Not good for battery health.
An iPad for drawing vs a cheap tablet for drawing are a mile apart in capability.
Maybe for technical work your statement is true, but for artistic work an iPad with Procreate/Dreams is absolutely a viable primary workstation.
Yes, of course if you have a concrete, specific significant use case for the pencil it may be more appealing. But the pre M1 iPad with the pencil did the job as well, for drawing and pure handwriting. Obviously the iPad is marketed to be more than a single purpose device. The processing power is completely wasted. Most people probably won't use the pencil much in the long term.
It's file management, app limitations and interoperability which kill usefulness.
I love the iPad for its stylus, it’s a great device for brainstorming even with just notes. I wish there was more creative software they took advantage of the form factor however, like a touch/stylus based programming environment.
It is marketed to be more than a drawing device tho. It fails at anything but being a glorified scratchpad. Very narrow usefulness.
I got the first iPad (2016?) with pencil, for 400€, I think. It still does completely fulfill my needs for brainstorming. For anything else, the app ecosystem and locked down system are useless.
I bought the iPad to do mostly three things: browse web, send emails, watch video, and take notes.
I also wish there were better more inventive apps.
I agree that a lot could be improved on the iPad and I would not say no to the features you suggest. But I’m not sure I understand why you would want to use an iPad instead of a laptop as a developer. For me the whole thing with the iPad is the pencil. I would never use it for development tasks (maybe like some sort of emergency device while traveling).
Not saying you are wrong to want to use it for development, just curious why!
Because the iPad is only missing a keyboard to become a laptop hardware wise. In theory. The laptop is redundant.
You can even have better ergonomics with a detachable screen. Basically a desktop setup with laptop mobility.
Fair enough, that makes sense and would be nice if it worked.
It would be the last device I ever need, if it did work like that... It actually hurts me, how its limitations are completely arbitrary and artificial. The iPad could be the best thing ever, yesterday. So, so frustrating and disappointing.
> Maybe it’s a laptop killer for creative professionals?
Nope. I'd argue developers are actually the best served professionals by iPads, simply because a terminal-drive workflow via SSH is at least possible.
I've written about the situation of creative apps not gaining traction on iPads (Procreate being the main exception) https://blog.robenkleene.com/2019/08/07/apples-app-stores-ha...
> simply because a terminal-driven workflow via SSH is at least possible
This has been the one thing that has kept my iPad useful! I use Blink (grandfathered into their Legacy Pro plan) and it’s a very clean experience. I can’t help but feel like it’s a fragile experience having to rely on an app from the app store for my iPad to be useful — that Apple or the Developer could remove at any time. Maybe it would feel less fragile if Apple provided an official i[Pad]OS Terminal/SSH client.
Apple announced the M3 on October 30 last year, roughly 6 months ago. Have they ever announced successive chip designs in such a short time frame?
Unless Mark Gurman has a record of making reliable predictions about Apple, I wouldn't give this much credence.
It's not completely insane.
1. N3B TSMC node used by M3 is reportedly low yielding. Apple likely wants to move away from it asap and move to N3E, which is the higher yielding node.
2. Rumors are that M3 Ultra is not planned due to low yields. M3 Max die shot does not have UltraFusion silicon. This supports the first theory.
3. Apple base the M chips on the A series cores. A18 Pro has already finished in design because Apple always starts making A series chips in Spring before a Fall debut. Therefore, A18 Pro is already in production using N3E node. Given enough early planning, it isn't surprising that M4 can be released early.
4. Reportedly, Apple plans to focus on GenAI at WWDC. They're also behind in GenAI. In order to catch up, they need to be aggressive. M4 is likely to have a big improvement in inference speed via the NPU and/or GPU.
I think 4 is underrated. From what I've seen Apple is making a play for on device inference. Gains on better screen, thinner phone, better camera selling points are pretty incremental now. But with on-device inference they now have access to: doubled the number, no cloud we care about privacy, on device translation capability, etc.
> They're also behind in GenAI
Where can I read up on that and what's state of the art in consumer GenAI chips?
Also thanks, all great points!
Think that whole major, unpatchable security flaw thing could also play a part? I could see them wanting to put distance between themselves on that fast too.
I don't think so. I think M4 was planned years in advance to debut around this time. You can't change this stuff in a few months.
M3s have a flag to disable the caching vulnerable to GoFetch on p-cores. M1s and M2s are the ones that have no mitigation besides running workloads on slower e-cores.
That‘s not true DMP can be disabled on M2 chips and probably also on the M1 chips. This was found by Hector Martin (@marcan) and will probably supported in Asahi Linux.
That's great news. Direct link for anyone else not up to date: https://social.treehouse.systems/@marcan/112238385679496096
I don’t recall any that affected silicon?
Key extraction from the HSM. It's unclear if this was ultimately patched in the kernel, but it appears patchable.
[1] https://gofetch.fail/
[2] https://social.treehouse.systems/@marcan/112238385679496096
Definitely not the HSM, this only affects the AP
> Unless Mark Gurman has a record of making reliable predictions about Apple, I wouldn't give this much credence.
He does have a great record.
They need to stop calling these pro devices. They are barely functional for anything beyond casual use and artwork
I am a physicists (theory), and about a third of my work is “pencil and paper” calculations (and related notes) on the iPad Pro. I use it professionally ~20 hours per week.
May I ask if they are direct paper to screen transfers, or if there are specific features/apps that add extra value? I do a lot of work in (paper) engineering notebooks, and I'm curious about what improvements your workflow has.
I ~exclusively use ZoomNotes, which is very customizable but is quirky and comes with a steep learning curve.
For me the main benefits of the tablet are:
- Copy/paste/drag. Pretty helpful when you manipulate long expressions, or want to condense a bunch of scratch work.
- Relatedly: easy erase. When I used real paper, pencil was too low contrast, and “erasable pen” never worked well. So I used a real pen and just crossed tons of stuff out; ugly.
- Keep a decade of notes with me wherever I go. (And cloud backed up so I won’t lose them.)
- Infinite zoom and no page boundaries. Seems silly, but it’s really nice to write as big as I want without worrying about running out of room on the page.
- Text search of handwritten notes. Works surprisingly well.
- Easily switch to different colors. 99% of the time I only use 2 or 3, but I find it helpful to visual distinguish the main argument/computation from “side commentary”.
Undo is one thing you get used to very quickly. Copy paste, select and move, change colors, erase without a trace... digital handwriting has a lot to offer.
However, the app ecosystem is lacking in terms of interoperability, feature completeness within one app and export/import capabilities, and there is nothing you can do about it in iPadOS. File management is pure pain unless you fully bought into Apple already.
It's very nice for brainstorming and making transient personal illustrations while learning and working through technical problems, but you won't get the feel and certainty of a physical notebook as an institution for persistent notes.
The iPad is way too expensive for what it has to offer, in practice.
That said, without its artificial limitations it would be the last computing device you'll ever need. It could be everything.
Strong agree with the first two paragraphs. I have spent >10 working hours per week for past 4 years with an app I paid $7 for. I wish so much I could pay $7,000 for an improved version.
Disagree on permanence. My notes are useful for much longer when they can be electronically sorted, searched, backed up, and when they take up no additional space in my backpack.
Maybe you should try remarkable tablet?
Why would they do that if their current system works?
Lasts much longer after recharge, better for eyes, better canvas for hand writing.
Tried it, along with like 5 other tablets before I settled on the iPad+Pencil combo
I drink coffee at work. Does that make it Coffee Pro?
[dead]
Which doesn't mean parent is wrong. You can do any work on non-ideal equipment and be happy with it for various reasons.
When I saw what one of my ex was using (bio chemist growing various stuff and testing treatments to rare sickle cell anemia that pharma didn't care about due to low patient count), it was very subpar and inefficient but she was happy with it and got results nevertheless eventually. Some of her work I could make more efficient by 10x, some by 50x maybe, but ie she as hardcore unix person just plainly refused to use MS Excel and all its statistics power, to pick a random example (there were quite a few).
Not saying this is your case of course, but oh boy do small screens, limited browsers, limited power and storage speed etc. do have negative effect on work itself, mental flow, staying focused etc. There is a reason not only we IT guys want powerful machines and massive screens, it ain't just chasing shiny higher numbers.
Apple has never used "pro" to mean professional software engineers.
It's always been meant for professional artists. Whether graphic artists, music producers, video editors, 3D artists, or what have you.
So it's strange to me that you treat "artwork" as an exception.
These are very much pro. That they're extremely functional for art is the whole point.
> These are very much pro. That they're extremely functional for art is the whole point.
I'd argue that iPads being great for "art" is false. They have one killer app for one narrow specialized area (Procreate/illustration) but they aren't widely used as a main device professionally outside of that context (they are often used as a satellite device). They have adequate support for hobbyist level video editing/audio production/photography. They have minimal to non-existent support for motion graphics or anything related to 3D. Calling a device like this "extremely functional for art" I don't think is a defensible statement (unless it's specifically scoped to illustration).
I've written about my analysis of creative of apps on iOS devices here https://blog.robenkleene.com/2019/08/07/apples-app-stores-ha...
In the concert hall I see more and more musicians & singers replacing paper music sheets with 12” iPad Pro’s. Mostly individual artist but recently all members of a cappella choir. Some, eg the violinists, use a small device on the floor probably connected by Bluetooth to advance to the next page with their foot. I don’t remember seeing a conductor with the full score on an iPad though, maybe one day with a foldable 2x20+”.
Pro != “professional”
Pro == more capable
And the iPad Pro is more capable than any other iPad.
It’s better than the iPad Con
There are professionals in artwork?
Art professionals usually don't use iPads although I'm sure some do. Most seem to prefer Wacom. Unlike, eg a laptop, portability isn't usually a major professional need for an art workstation.
To the contrary, art professionals use the top-tier iPads professionally more than anyone else, in my experience.
You use a Wacom for some things, but the iPad has the gigantic advantage that your image is directly underneath your Pencil. And, of course, that you can take it with you.
The iPad isn't a substitute for a workstation. But it's a huge complement -- both for creating things visually, as well as managing/showcasing all your media on the go. Professional artists are often going to a lot of meetings with different clients or team members around the city.
Wacom makes Centiqs which are very popular pen on screen monitors if you prefer that style, so the iPad doesn't bring much there.
I would say the vast majority of artists do not travel and do not need to draw mid-meeting. Certainly there are exceptions but they are just that, exceptions.
Centiqs may be great, but I've seen 100 artists using an iPad for every 1 artist I've seen using a Centiq.
I would say the vast majority of artists do have client meetings, whether they're employed full-time or freelance.
And artists need to work on the go just like the rest of us. Whether it's a dad trying to do 30 min of video editing in his car while he waits for daycare to get out, or somebody in the city doing an hour of website design in a cafe while they wait for their next meeting.
I'm baffled by your assertion that artists don't use iPads or shouldn't need them. It's just not the case, if you work with professional artists.
Interesting, I've worked in large game studios with many 2d and 3d digital artists and they just don't use iPads. It could be a specific industry thing or a regional thing but even during the early hybrid work days I would see artists haul in and pack up their Centiqs. IPads were not in the conversation.
I will say that it's a major disconnect in our experience that you would say the majority of artists are directly client facing.
What company/industry is buying their artists iPads?
iPad Pros are often used in music production too. There’s plenty of heavier duty non-casual sitting on the couch watching TV type use cases it’s used for and it’s targeted at.
Speaking of the Mac Studio M4, Apple is considering having it support up to a half-terabyte of memory. In comparison, the existing Mac Studio and Mac Pro top out at 192GB of unified memory (RAM).
(according to Tom's guide). I wonder if it will still be on-chip.
Maybe around M6 or M7 they'll release an external monitor that does more than 60hz.
I run dual 2k monitors at 144hz from my M2 MacBook. Works even with the laptop monitor on.
I meant the Studio Display. Most of the Pro phones/ipads/macbooks are 120hz now with high DPI. But the Studio Display is stuck at 60hz.
You can do more than 60hz if you use HDMI. See https://support.apple.com/en-us/101571
Not sure for the downvote, but here is an excerpt for the M3 MBP from the link above:
> One display up to 4K at 144Hz over HDMI
The other M series have HDMI with similar speeds as long as it isn’t driving 8K.
Thunderbolt seems to be limited to 6K and 60Hz, but HDMI isn’t similarly limited (well, it can only do 60Hz with 8K). Hopefully there will be a thunderbolt successor they fixed the problem.
Maybe never will they release a mobile computer with a non reflective screen.
At first I thought, it sounded crazy but then I realized that reports have told us that the M4 will be a key part of Apple’s reneweed AI strategy for 2024. The A18 and M4 in the iPhone 16 and the M4 MacBook Pros at the ens of this year will have much better neural engines to drive Apple’s WWDC 2024 AI features. Apple can’t leave the iPad Pro out to dry.
For the first time in a long time, consumer hardware is not even close to being fast enough for a new type of mass applications: GenAI.
People have stopped caring about consumer chips because they have been "good enough" for a long time.
Hardware is getting exciting again. We need significantly more compute, memory bandwidth, and memory capacity. There are huge bottlenecks everywhere.
For GenAI, I wonder if we need consumer hardware to be that powerful. Isn't this exactly what "the cloud" is for?
I think the best models will always require "the cloud" but there will be plenty of use cases where "good enough" local LLMs need to be on-device.
It depends. If you're OpenAI, consumer hardware should never be powerful enough to run local models. If you're Apple, it should.
https://archive.is/hWVqB
I really want a new iPad mini, way overdue for a refresh.
My bet, Apple begins designing chips specific to each product line.
Meaning,
- iPad get their own unique chip
- MacPro, unique chip
- etc
And this iPad rumor is just the first Apple Silicon unique product chip.
They already do this with fuses to lock each chip to a specific iBoot, OS, and release[0] set.
Changing the configuration of the chip is unnecessary. iPhones, iPads, and Macs use roughly the same balance of compute. The thing that does need to change is how many cores in each compute device, but they already do this with A17 vs. M3 vs. M3 Pro/Max/Ultra. That already segregates the product lines a bit: iPads aren't getting anything bigger than an M3/M4, Mac Pros aren't getting anything smaller than an M3 Max, etc.
[0] e.g. production devices vs. Apple internal devkits vs. that weird Security Research Device they released a few years ago to compete with Corellium
The M3 Pro is already a custom chip vs the M3 (base) or M3 Max.
The M3 Pro is the only chip that has 6E cores.
Both the M3 base and M3 Max only have 4E cores.
Why does the iPad Pro require a unique chip from a base Mac?
Remember that it costs tens/hundreds of millions to tape out a chip.
[dead]
My guess is that M4 and forwards will not be so much about CPU and GPU performance but have some sort of local LLM in ROM. So every year you should upgrade to get the best hardware LLM ever released.
The whole article hinges on them adding more inference power to the Neural Engine, but the is no benifit or plausable reason to burn weights into ROM.
iPhone/iPad 16GB RAM + 64GB hardware local LLM ROM. Apple can phrase it as both a benefit and a reason to upgrade and reduce its costs over suddenly bumping RAM by 400%.
Out of all the hardware manufacturers, Apple is the worst candidate to pin planned obsolescence on. Its devices are supported longer than its competitors, and they get better at lasting longer over the years.
I don't know if mask ROM even exists any more or what its density would be. Realistically your choices are DRAM or flash and flash will probably never be fast enough to hold model weights.
The toolbox ROM was a great hack in 1984 though.