The latest iMac is a pitiful computer

[za9ra22]

...They are big mains powered iPads...

This really does make me laugh. A Ford Mustang is just a faster Model T. Concorde was just a faster Wright Flyer. The ISS is just a newer tent with more sleeping compartments.

Take your pick, and whatever it is is only a .... except bigger/newer/flatter/faster/smaller... All entirely pointless comparisons, because it's how evolution works.

 

[EntropyQ3]

This really does make me laugh. A Ford Mustang is just a faster Model T. Concorde was just a faster Wright Flyer. The ISS is just a newer tent with more sleeping compartments.

Take your pick, and whatever it is is only a .... except bigger/newer/flatter/faster/smaller... All entirely pointless comparisons, because it's how evolution works.

Yeah, OK, that statement was bound to derail my point. 😀
My point being that there is nothing in the new iMacs that leverages that they are stationary, mains powered computers.
Nothing.
And that the design concept then would make more sense and be more appealing if you simply removed the computer and sold them as nicely designed and matching monitors/keyboards/mice.
If a stationary computer is a better product if you remove the actual computer, then arguably it is not an optimal incarnation of the concept.

As I said in my answer to the original question, I kinda like them anyway. They seem nice. But there it is.

 

[theluggage]

Probably never gonna happen again, the RAM on the processor die is part of the M1’s efficiency. A removable SSD would just be “dumb” memory since the M1 has the memory controller on the chip itself. Meaning standard SSDs wouldn’t work.

The justification for soldered-in RAM in laptops (which has been the case since Apple started the practice) is that they use low power LPDDR RAM which simply isn’t available in plug-in modules. Keeping the connections as short as possible is part of the power-saving game. Power usage increases drastically with speed, so reducing power and running faster are two sides of the same coin - on a desktop, you could afford to burn a bit more power and use modular, desktop RAM.

The M1 RAM isn’t on the die (I.e. the actual “silicon chip” part of the processor) - it is in regular surface-mount LPDDR chips soldered alongside the die in the processor package. This keeps the connections nice and short which helps with power vs. speed, but there is no fundamental reason why a future “desktop” processor couldn’t have conventional modular RAM at the expense of a bit more power consumption. The physical location of the RAM has nothing to do with “unified memory”.

As for SSD: yes, the existing SSD is “dumb memory” so we’re stuck with proprietary SSDs as the main system drive but they could still be on a replaceable module - which is already the case on the Mac Pro. Since SSDs have a finite life (which could possibly be eaten up by a software fault) that would be sensible even if it didn’t mean cheap upgrades. Plus, on a future processor, it should be possible to add a second slot for a standard M.2 stick for internal expansion.

The current iMac, though, can only do what the M1 can do - and the M1 is, at heart, designed for ultra low power MacBook Air-class systems and tablets - although it has really pushed the envelope as to what that class of system can do. So, given that there M1 is all Apple currently have to put in an iMac it kinda makes sense to play to the M1s strengths and make it thin’n’crispy (although I think using an external power brick was a step too far). It looks like a very capable system that will hit the spot for a lot of users.

However, I really hope that Apple will follow through and, with M1X/M2/Whatever, produce something that is designed for desktops and larger laptops, and doesn’t prioritise power consumption over all else. The only thing that concerns me is that the 24” iMac design philosophy might spill over to the replacements for the 5k/iMac Pro and higher-end Minis - and that the 16” MBP replacement might sacrifice too much for being thin’n’light.

 

[robco74]

If I were going for the new iMac, I'd probably order the vesa mount version. Then the light weight would be an asset. You could mount the previous gen, but they were still somewhat heavy. With the new one, it would mount easily and it would be great for ergonomics, and you could swing it out of the way when not needed. Hopefully we'll see a new version that replaces the higher end 21.5" configs.

 

[za9ra22]

Yeah, OK, that statement was bound to derail my point. 😀
My point being that there is nothing in the new iMacs that leverages that they are stationary, mains powered computers.
Nothing.
And that the design concept then would make more sense and be more appealing if you simply removed the computer and sold them as nicely designed and matching monitors/keyboards/mice.
If a stationary computer is a better product if you remove the actual computer, then arguably it is not an optimal incarnation of the concept.

As I said in my answer to the original question, I kinda like them anyway. They seem nice. But there it is.

I'm sorry but I don't even recognise the point you're making. To the extent that the new iMac is a 'stationary system', there is nothing about it that I see to be wrong or missing or that it should have, or be, instead of what it is.

Certainly it could be different, and if Apple were HP, or Lenovo, or Dell, it obviously would be, but in as far as the computer business goes, Apple's 'design language' is amongst the most coherent across their products, and is clearly broadly liked (because if it weren't, they would change it). To my view, what Apple have done with the iMac is take their recent design language and simply 'big it up' to a desktop computer - which is entirely consistent with their previous product transitions.

I don't see anything wrong with this at all, but then the 'slab' styling appeals to me, and I see absolutely no reason not to have an out-board power supply, or a range of colours, even as most are not necessarily appealing to me personally. But then I also see no reason why it shouldn't look like a big iPad, because the product isn't differentiated just by size, but also where it will be situated, how it is used, and for what it will be used.

The arguments that we see here (not meaning yours, but the widely differing opinions being voiced) are a pretty good example of why you can't do design by committee.

 

[za9ra22]

... and the M1 is, at heart, designed for ultra low power MacBook Air-class systems and tablets...

This is a rather specious argument when considering that the M1 is certainly designed for low-power consumption/low heat generation, which is advantageous in portable systems particularly, but is certainly not a 'MacBook Air-class' processor when considering how it performs against prior desktop-class alternatives.

On the face of it so far, it appears in fact to be an excellent first step in demonstrating that it is possible to develop systems which are both low power and high performance. In itself we should be applauding that development.

 

[Lemon Olive]

My parents usage and typical "general user" usage - one browser tab, maybe word/pages and mail. That is all. I do not know when this trend started where something isn't able to open "200 tabs".

I'm a Pro user and I rarely have more than 10 tabs open. It just isn't efficient. And I have 64 GB of RAM. And Safari still reloads and performs like garbage.

As someone who uses the high end Intel models all day long, I'm not sold on the idea that they can't be utterly replaced by an entry level Apple Silicon machine.

 

[theluggage]

but is certainly not a 'MacBook Air-class' processor when considering how it performs against prior desktop-class alternatives.

The M1 is absolutely amazing for its size and power consumption, and has raised the bar on what a "MacBook Air-class" machine can do. But don't get carried away: it is what it is, and what it is is a SoC designed for tablets and ultraportable laptops. If Apple decides to go for a one-size-fits-all processor to use in everything from the iPad to the Mac Pro then they'll be severely restricting the potential of their higher-end computers.

It doesn't consistently beat higher-end desktop x86 processors (esp. AMD ones that are currently out-performing Intel) - especially on heavily multi-threaded tasks that can exploit more than 4 high-performance cores. It "only" has the GPU performance of a 3-year-old desktop PCIe GPU, and can't compete with the latest discrete GPUs.

It uses a mixture of low-power and high-performance cores which only makes sense if you're optimising limited battery life.

It is only available with a maximum of 16GB RAM, and is physically limited to 2 LPDDR4 chips which AFAIK currently come in a maximum size of 12GB - so 24GB is the most they could squeeze on.

The GPU only supports 2 displays, total.

When Apple come to replace the 5k iMac and 16" MBP models with 8/10 core i9s and premium GPUs with 8-16GB VRAM, they'll need something that does a lot better than give last year's Intel models "a good run for their money" on selected speed tests (but don't ask about RAM and display support). More CPU cores, more GPU cores, some way of adding more than 24GB or RAM... all of which will consume more power.

They're going to need a mobile workstation/desktop-class processor for that, not the M1. It will probably use a lot less power than an i9 and save a few polar bears, but it doesn't make sense to saddle it with the strict power limits needed for a tablet or fanless ultra-mobile.

(They could go for a multiprocessor approach: i.e. the higher-end computers have multiple SoCs with GPU and RAM somehow distributed between the individual chips - but even then it doesn't make sense to saddle single-SoC tablets and ultraportables with the extra inter-processor communication junk)

 

[Jorbanead]

My point being that there is nothing in the new iMacs that leverages that they are stationary, mains powered computers.
Nothing.

Are you asking what is the point of getting an M1 iMac over an M1 MacBook Pro?

The iMac has a 24" screen, and has the ergonomics of a desktop instead of a laptop. It also has better cooling than the MBA and MBP so sustained performance should be on par with the mac mini.

And that the design concept then would make more sense and be more appealing if you simply removed the computer and sold them as nicely designed and matching monitors/keyboards/mice.

With all due respect, this makes absolutely no sense to me. So you take away the computer, and then now you need to add another piece of hardware (a computer) on your desk to accomplish the exact same thing the iMac does all in one device? "instead one of device, it would be better to split it up into two devices". How is taking away features a better product?

The whole point of iMac is that it's the most minimal desktop solution you can buy.

 

[za9ra22]

The M1 .... has raised the bar on what a "MacBook Air-class" machine can do....

Once again, a specious argument. I really don't feel the need to repeat myself.

What I will do is point out that Apple, as a company, employs quite a few, likely well talented engineers, which encourages me to think that whatever plans it has, it can work out how to achieve their desired results without my guidance. Possibly even yours!

 

[Queen6]

The M1 is absolutely amazing for its size and power consumption, and has raised the bar on what a "MacBook Air-class" machine can do. But don't get carried away: it is what it is, and what it is is a SoC designed for tablets and ultraportable laptops. If Apple decides to go for a one-size-fits-all processor to use in everything from the iPad to the Mac Pro then they'll be severely restricting the potential of their higher-end computers.

It doesn't consistently beat higher-end desktop x86 processors (esp. AMD ones that are currently out-performing Intel) - especially on heavily multi-threaded tasks that can exploit more than 4 high-performance cores. It "only" has the GPU performance of a 3-year-old desktop PCIe GPU, and can't compete with the latest discrete GPUs.

It uses a mixture of low-power and high-performance cores which only makes sense if you're optimising limited battery life.

It is only available with a maximum of 16GB RAM, and is physically limited to 2 LPDDR4 chips which AFAIK currently come in a maximum size of 12GB - so 24GB is the most they could squeeze on.

The GPU only supports 2 displays, total.

When Apple come to replace the 5k iMac and 16" MBP models with 8/10 core i9s and premium GPUs with 8-16GB VRAM, they'll need something that does a lot better than give last year's Intel models "a good run for their money" on selected speed tests (but don't ask about RAM and display support). More CPU cores, more GPU cores, some way of adding more than 24GB or RAM... all of which will consume more power.

They're going to need a mobile workstation/desktop-class processor for that, not the M1. It will probably use a lot less power than an i9 and save a few polar bears, but it doesn't make sense to saddle it with the strict power limits needed for a tablet or fanless ultra-mobile.

(They could go for a multiprocessor approach: i.e. the higher-end computers have multiple SoCs with GPU and RAM somehow distributed between the individual chips - but even then it doesn't make sense to saddle single-SoC tablets and ultraportables with the extra inter-processor communication junk)

That fact that people are even remotely talking about the M1 SOC versus high-end desktop x86 processors should make all pause for thought. M1 will be the slowest, least powerful of Apple Silicon bearing in mind that the M1 MBP easily see's off far larger Hex & Octa core notebooks without breaking sweat.

I believe that Apple will follow through. They look to be listening to the customer's concerns and getting back to under promising and over delivering. My only cavet is that Apple royally dropped the ball with the 2016 MBP etc. I want to see that the second generation of Apple silicon in a portable Mac proves to be reliable given new chassis & display technology.

For me the M1 MBP was an instant purchase; I visited our largest local Apple Store and asked to speak with someone technical not sales. 30 minutes later I left with the M1 MBP nor does it disappoint, hands down beats this very fast Asus 17.5" notebook.

I could be wrong, however I feel that Apple has made some internal adjustments. After having their entire portable Mac line up on extended warranty schemes, diametrically polarising both Pro & consumer users alike due too questionable design, topped off with poor reliability. Doubt Apple want's more of the same, so I do think that the larger Mac's will be significantly more powerful again and reliable. Is the M1 MBP perfect? No, equally it's a significant step in the right direction and the value proposition makes sense.

The only real issue is, can can you work within the environment of Apple Silicon & Big Sur.

Q-6

 

[theluggage]

Once again, a specious argument.

There's nothing "specious" about it. The M1 was designed for ultraportables and tablets. It didn't magically optimise itself as an ultra low-power mobile SoC by accident - things like mounting the RAM chips on-package, the 4 high-power/4 low-power core configuration and only supporting 2 displays are design compromises that only make sense if you're designing an ultrabook. Nobody is weeping that they can't get a 64GB Air or run 2 external displays from their iPad Pro, but those are not acceptable limitations for a high-end 16" MBP or 5k iMac, however high the Geekbench scores are.

What I will do is point out that Apple, as a company, employs quite a few, likely well talented engineers, which encourages me to think that whatever plans it has, it can work out how to achieve their desired results without my guidance. Possibly even yours!

Of course - and all the rumours say that Apple are working on new Apple Silicon chips which will go in the 14/16" MBP, with some permutation of better GPUs, more high-power CPU cores etc. I'm pretty sure that they know they'll need more RAM too. They will also know that these are not things that need to go into the next generation of MacBook Air chips when they roll around in a year or two.... Nobody is predicting that the M1 is going to be the one-size-fits-all chip for everything.

My original point was simply that there is no point in making the 24" iMac thicker (which some people have been calling for) because it is using a chip designed for ultraportable applications that is already pretty much maxed out in the 13" MBP and Mini - which have far better cooling than the Air but only run marginally faster. Ditto for adding extra ports - the Mini looks like it represents the maximum number of ports the M1 can drive.

 

[dogslobber]

It still has chunky bezels that make it look like a monitor from the late 1900s.

 

[snorkelman]

In a lot of respects I see this line up as a transitional one: So however obvious some of the other things they could do to the 24iMac, Apple aren't going to needlessly burn two or three jumps forward (and repeat sales they can squeeze from them) when one will pretty much do

Colors and basic restyle? go with that just now

Chin? Keep the option of ditching that in reserve to make a song and dance about a couple of years down the line

Bezels Leave that for a third iteration and make another song and dance about that too

Processor? establish a base line right across the whole range of mass market models, rather than needlessly segment those from day one. Which is largely what they've done; ultralight/mini/all in one? just pick the format that best appeals to you. No need to dither over relative performance of one format versus another, they're all by and large the same

Same approach for the mid range of multi port MBPs and large screen iMacs - higher core count higher RAM capacities than the mass market models (but again probably with as little performance segmentation within that line as possible too this first time around)

2022 comes around? Trickle the SOCs designed for the multi port MBPs/large screen iMac down to as many of the mass market ones as is practical to do so (24iMac and MacMini) While focusing on re-style and color options for the MBA and 2-port MBP

 

[iPadified]

There's nothing "specious" about it. The M1 was designed for ultraportables and tablets. It didn't magically optimise itself as an ultra low-power mobile SoC by accident - things like mounting the RAM chips on-package, the 4 high-power/4 low-power core configuration and only supporting 2 displays are design compromises that only make sense if you're designing an ultrabook. Nobody is weeping that they can't get a 64GB Air or run 2 external displays from their iPad Pro, but those are not acceptable limitations for a high-end 16" MBP or 5k iMac, however high the Geekbench scores are.



Of course - and all the rumours say that Apple are working on new Apple Silicon chips which will go in the 14/16" MBP, with some permutation of better GPUs, more high-power CPU cores etc. I'm pretty sure that they know they'll need more RAM too. They will also know that these are not things that need to go into the next generation of MacBook Air chips when they roll around in a year or two.... Nobody is predicting that the M1 is going to be the one-size-fits-all chip for everything.

My original point was simply that there is no point in making the 24" iMac thicker (which some people have been calling for) because it is using a chip designed for ultraportable applications that is already pretty much maxed out in the 13" MBP and Mini - which have far better cooling than the Air but only run marginally faster. Ditto for adding extra ports - the Mini looks like it represents the maximum number of ports the M1 can drive.

Actually, M1 was likely designed for those with relatively modest compute demands. The format, air, MBP, iMac, iPP, does not matter as the buyer compute demands are the same. “Modest” like a modern i7 desktop chip or i9 mobile chip and that is without the coprocessors being used!

Why the iMac24 cannot share the MX chip with MBP14/16 is strange. Technically it can but the customers do not necessarily want to pay the price for these chips as they likely will be much more expensive.

It is about perception: I feel that the iMac 24 inch is underpowered while the iPP is overpowered with the M1 chip.

 

[leman]

Actually, M1 was likely designed for those with relatively modest compute demands. The format, air, MBP, iMac, iPP, does not matter as the buyer compute demands are the same. “Modest” like a modern i7 desktop chip or i9 mobile chip and that is without the coprocessors being used!

M1 was designed first and foremost for specific power consumption and manufacture price targets. The second most important design parameter was maximizing delivered performance under these constraints.

So yes, M1 was definitely designed for ultraportables. It’s peak 15W power consumption makes it very clear. It’s just a “coincidence” that Apple‘s architecture is so far ahead of the state of the art that their low-power chip can successfully compete with more performance-oriented products from other manufacturers. This again is the reason why Apple can “get away“ with putting this chip in a desktop as well, it’s simply fast enough for that market segment. It is a smart business move after all, if not much else.

 

[skaertus]

It still has chunky bezels that make it look like a monitor from the late 1900s.

You mean, from 1909?

 

[theluggage]

Why the iMac24 cannot share the MX chip with MBP14/16 is strange.

I'm gonna go out on the limb and say that its because the M1x/M2 isn't available - or even officially announced - yet.

Even if it was ready, Apple would probably want to keep that powder dry until they launch the new 14/16 MacBook Pro which is probably the next highest profile product after the Air.

In a lot of respects I see this line up as a transitional one: So however obvious some of the other things they could do to the 24iMac, Apple aren't going to needlessly burn two or three jumps forward (and repeat sales they can squeeze from them) when one will pretty much do

I think a "transitional" iMac would have shown up in the old 21.5" enclosure, just like the other current M1 machines. Love it or loathe it, a lot of design work and tooling-up has gone into the new iMac design - including the new magnetic power/ethernet connector, new keyboards & multi-coloured mice and trackpads - so I think it will be with us for a while unless customers reject it (and although I don't personally love it, I don't see that happening unless there's a lurking design defect).

That doesn't mean it wont get better processors and/or displays in the future - although I think the smaller iMac has been one of the less frequently updated machines in the past.

Design-wise: I don't get the bezel-phobes (as long as they're a neutral colour). They've thinned them down a lot. Until/unless Apple perfect the through-the-screen webcam the top bezel is needed, and I quite like to be able to adjust my screen without getting fingerprints on the actual display area. That said, as I understand it, the bezels are just white paint on an edge-to-edge sheet of glass, so maybe they could be made smaller still without a major re-design.

The chin - that's needed for half-decent speakers if you're going to make the rest of the machine so thin. The last chinless wonder with internal speakers was the 27" LED Cinema/Thunderbolt display - and that was far thicker, plus it didn't have to squeeze in a whole computer... and those speakers always sounded like a half-decent pair of speakers shut in a tin box.

Processor-wise: at the moment, the only game in town is the M1. The next processor is most likely to be optimised for the larger MacBook Pro - and while that can afford to use more power than would be available in an MBA or iPad it's probably still going to be far lower power than the i7/i9 that it is replacing and would probably run comfortably in the 24" iMac chassis. So maybe that could be offered in the future. If not, there will doubtless be a M1 replacement along in a year or two - still optimised for Air/iPad but with updated tech squeezing out a bit more performance.

So yes, M1 was definitely designed for ultraportables. It’s peak 15W power consumption makes it very clear. It’s just a “coincidence” that Apple‘s architecture is so far ahead of the state of the art that their low-power chip can successfully compete with more performance-oriented products from other manufacturers.

Absolutely. We're really guessing at the moment what Apple's final Apple Silicon line-up is going to be. I don't think you can draw too many conclusions from the A-series, which was basically "new major revision in the flagship iPhone every (other) year, with slightly souped-up revisions for iPads, last year's processor for cheaper models". Ultimately, iPad/iPhone are just various sizes of hand-held tablet, whereas the Mac line covers everything from the Air to the Pro. My guess would be 3 "series":

1. optimised for the MacBook Air
2. optimised for the 16" MacBook Pro replacement
3. optimised for the higher-end 5k iMac/iMac Pro replacement
4. some sort of multi-processor arrangement based on (3) for the Mac Pro. Unless they produce a monolithic Xeon-W killer (or, better, AMD Threadripper-killer) - which would be humungously expensive in such small quantities.

Within that, what goes into future versions of the Mini, the 24" iMac, the 13" MBP (if it isn't replaced by the next-gen Air) might be more fluid: could be the lower model with a small boost from extra cooling, could be the higher model with some cores disabled or i/o unused... But that's just a guess, not a hill I'm going to die on.

What is not sustainable (for Apple) is the current situation where buyers with $3000 in their pocket are wondering whether a $1500 M1 system might get the job done...

 

[za9ra22]

There's nothing "specious" about it. ...

Yeah, all your points are nice and logical, but flow from a fallacious contention.

 

[dogslobber]

You mean, from 1909?

When people refer to the late 1800s then do they mean 1809?

 

[iPadified]

M1 was designed first and foremost for specific power consumption and manufacture price targets. The second most important design parameter was maximizing delivered performance under these constraints.

So yes, M1 was definitely designed for ultraportables. It’s peak 15W power consumption makes it very clear. It’s just a “coincidence” that Apple‘s architecture is so far ahead of the state of the art that their low-power chip can successfully compete with more performance-oriented products from other manufacturers. This again is the reason why Apple can “get away“ with putting this chip in a desktop as well, it’s simply fast enough for that market segment. It is a smart business move after all, if not much else.

IUltraportables have never been top of heap in terms of performance so a desired performance bracket is equally true in terms of M1. If anything the M1 demonstrate that old fashioned thinking of Desktop-, mobile , and likely workstation chip has little meaning. With the Intel chips, there was no choice as the computer was to very large extend designed around the chip heat dissipation. Now there is much more freedom in form factors and hence the form factor will be much more important in the future

With that in mind the M chips are designed for a performance bracket for computer form factor driven design and not power bracket for heat dissipation driven design.

There was no 8+4 M1X chip and not even a overclocking of the M1 chip. Why? Because the vast majority of the customers, "consumers", would not benefit from the increases in multithreaded computing.

 

[iPadified]

1. optimised for the MacBook Air
2. optimised for the 16" MacBook Pro replacement
3. optimised for the higher-end 5k iMac/iMac Pro replacement
4. some sort of multi-processor arrangement based on (3) for the Mac Pro. Unless they produce a monolithic Xeon-W killer (or, better, AMD Threadripper-killer) - which would be humungously expensive in such small quantities.

If we think less about the form factor you table will be:

1. Small compute demands
2. Medium compute demands
3+4. High compute demands

The 4 will be very interesting to see how they solve.

Within each performance bracket there could be many form factors (desktops and portables) which would increase the choice for customers.

Yes, the 24 inch might get and stronger M chip in the near future.

 

[leman]

If anything the M1 demonstrate that old fashioned thinking of Desktop-, mobile , and likely workstation chip has little meaning. With the Intel chips, there was no choice as the computer was to very large extend designed around the chip heat dissipation. Now there is much more freedom in form factors and hence the form factor will be much more important in the future

I don't disagree with your conclusion, but I can't say that I agree with your argument. Of course, there is little doubt that M1 is good enough for desktop use — it is faster than any previous 21" iMac and it can compete very well with lower-end desktop CPUs.

But it's not that M1 collapses the difference between all these different types of computers, it's simply that it is good enough for majority of users. The big difference between Apple and, say, Intel, is that Apple chip design is driven by mobile (read: smartphone) needs. They had a thermal limit (of around 5 watts) and they said, let build the fastest core that we can around this power limit. And they managed to build a core that is as fast as any 20-25 watt x86 core, which obviously gives them much more flexibility in playing the CPU game.

In the end, there is no such thing as "too much performance". The only reason why Apple can use an M1 in the iMac is because M1 is fast enough for everyday use, and not because they somehow erode the difference between mobile and desktop. An iMac would have been a better, more powerful, more useable compute if it had a more powerful chip. M1 in the Air is revolutionary, as it gives you a lightweight, passively cooled ultrabook with premium-level performance. M1 in the iMac is just "ok", but nothing special. This is not a "let's build the best possible computer with under reasonable design limitations", it's a "we have a low-power chip that's fast enough, let's build a computer around it". And again, it's not a bad thing, at all. What Apple has been doing with M1 makes a lot of sense from business perspective and I think that the new iMac makes a great home and office computer.

So yeah, I don't think that we are in disagreement, we just tell the story a bit differently. I don't really see M1 as this "big equalizer" for computer performance, it's just a damn good chip with a lot of business sense behind it. But its usage in desktop is far from revolutionary as some have hoped.

With that in mind the M chips are designed for a performance bracket for computer form factor driven design and not power bracket for heat dissipation driven design.

What's even the difference? Heat dissipation is hat dissipation and form factor is what defines heat dissipation limits in the first place.

There was no 8+4 M1X chip and not even a overclocking of the M1 chip. Why? Because the vast majority of the customers, "consumers", would not benefit from the increases in multithreaded computing.

And because it would be more expensive to make a different chip. As I wrote above, M1 is "fast enough", and that's all there is. Apple can design a wide range of machines around M1 simply because nobody else can build a CPU/GPU that performs that well at that power level.

 

[iPadified]

I don't disagree with your conclusion, but I can't say that I agree with your argument. Of course, there is little doubt that M1 is good enough for desktop use — it is faster than any previous 21" iMac and it can compete very well with lower-end desktop CPUs.

But it's not that M1 collapses the difference between all these different types of computers, it's simply that it is good enough for majority of users. The big difference between Apple and, say, Intel, is that Apple chip design is driven by mobile (read: smartphone) needs. They had a thermal limit (of around 5 watts) and they said, let build the fastest core that we can around this power limit. And they managed to build a core that is as fast as any 20-25 watt x86 core, which obviously gives them much more flexibility in playing the CPU game.

In the end, there is no such thing as "too much performance". The only reason why Apple can use an M1 in the iMac is because M1 is fast enough for everyday use, and not because they somehow erode the difference between mobile and desktop. An iMac would have been a better, more powerful, more useable compute if it had a more powerful chip. M1 in the Air is revolutionary, as it gives you a lightweight, passively cooled ultrabook with premium-level performance. M1 in the iMac is just "ok", but nothing special. This is not a "let's build the best possible computer with under reasonable design limitations", it's a "we have a low-power chip that's fast enough, let's build a computer around it". And again, it's not a bad thing, at all. What Apple has been doing with M1 makes a lot of sense from business perspective and I think that the new iMac makes a great home and office computer.

So yeah, I don't think that we are in disagreement, we just tell the story a bit differently. I don't really see M1 as this "big equalizer" for computer performance, it's just a damn good chip with a lot of business sense behind it. But its usage in desktop is far from revolutionary as some have hoped.




What's even the difference? Heat dissipation is hat dissipation and form factor is what defines heat dissipation limits in the first place.



And because it would be more expensive to make a different chip. As I wrote above, M1 is "fast enough", and that's all there is. Apple can design a wide range of machines around M1 simply because nobody else can build a CPU/GPU that performs that well at that power level.

We are in agreement about the M1 “good enough” . What you did not agree with was that I though the M1 performance was chosen to fit a certain type of users (students, office, point of sale, web based workflow) and not a certain device.

Design for performance or heat is different. Would you design an M1 for office work/student with 4+8 cores? No it would be a waste becuase office work does not utilize 8 performance cores. Single tread performance is however dominating in office work but the M1 is more than good enough in that department.

Under the Intel era, all chips was always on the verge of being too hot for nearly all Macs. Hence the performance was limited by the machine heat dissipation. I bet it was a constant tug of war between marketing and Jony Ive wanting slim portable machines and the engineering trying to shoehorn in a too hot CPU into a far too small casing. Now all M1 are cool enough ever for the slimmest and smallest device so other considerations like choice of performance to address certain market segments (performance segments) can be made without heat dissipation considerations.

We will certainly see a larger M chip for multithreaded workflows in the MBP14/16 and larger iMacs. Personally. I think whatever goes into the MBP14/16 should go into the iMac24, but perhaps that is not smart because the multithreaded performance crowd want larger screens and hence few people would buy a iMac24 with high end M chip because the screen is too small or chip is too expensive. You see, I am not even considering the heat dissipation of the M chip in this reasoning and that is the core of my argument.

 

[Apple Knowledge Navigator]

M1 in an iMac 24" starts at $1299 with only 2 USB4 ports. Man, this is just sad at this point. 8GB RAM and 256GB storage. Horrible deal.

Here's a crazy little thought; you're clearly not the target customer.

The fact that you highlight only the I/O suggests that this is the most important feature to you. Great. There are other options that meet your needs.

But for many other people, myself included, who live in pretty much a wireless and cloud-based world for work and leisure, there are few instances where we need lots of additional I/O. And when we do, the versatility of USB-C/Thunderbolt comes into play.