Discussion Apple Silicon SoC thread

[Eug]

M1
5 nm
Unified memory architecture - LP-DDR4
16 billion transistors

8-core CPU

4 high-performance cores
192 KB instruction cache
128 KB data cache
Shared 12 MB L2 cache

4 high-efficiency cores
128 KB instruction cache
64 KB data cache
Shared 4 MB L2 cache
(Apple claims the 4 high-effiency cores alone perform like a dual-core Intel MacBook Air)

8-core iGPU (but there is a 7-core variant, likely with one inactive core)
128 execution units
Up to 24576 concurrent threads
2.6 Teraflops
82 Gigatexels/s
41 gigapixels/s

16-core neural engine
Secure Enclave
USB 4

Products:
$999 ($899 edu) 13" MacBook Air (fanless) - 18 hour video playback battery life
$699 Mac mini (with fan)
$1299 ($1199 edu) 13" MacBook Pro (with fan) - 20 hour video playback battery life

Memory options 8 GB and 16 GB. No 32 GB option (unless you go Intel).

It should be noted that the M1 chip in these three Macs is the same (aside from GPU core number). Basically, Apple is taking the same approach which these chips as they do the iPhones and iPads. Just one SKU (excluding the X variants), which is the same across all iDevices (aside from maybe slight clock speed differences occasionally).

EDIT:



M1 Pro 8-core CPU (6+2), 14-core GPU
M1 Pro 10-core CPU (8+2), 14-core GPU
M1 Pro 10-core CPU (8+2), 16-core GPU
M1 Max 10-core CPU (8+2), 24-core GPU
M1 Max 10-core CPU (8+2), 32-core GPU

M1 Pro and M1 Max discussion here:

Page 78 - Discussion - Apple Silicon SoC thread

Page 78 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

forums.anandtech.com

M1 Ultra discussion here:

Page 109 - Discussion - Apple Silicon SoC thread

Page 109 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

forums.anandtech.com

M2 discussion here:

Page 127 - Discussion - Apple Silicon SoC thread

Page 127 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

forums.anandtech.com

Second Generation 5 nm
Unified memory architecture - LPDDR5, up to 24 GB and 100 GB/s
20 billion transistors

8-core CPU

4 high-performance cores
192 KB instruction cache
128 KB data cache
Shared 16 MB L2 cache

4 high-efficiency cores
128 KB instruction cache
64 KB data cache
Shared 4 MB L2 cache

10-core iGPU (but there is an 8-core variant)
3.6 Teraflops

16-core neural engine
Secure Enclave
USB 4

Hardware acceleration for 8K h.264, h.264, ProRes

M3 Family discussion here:

Page 215 - Discussion - Apple Silicon SoC thread

Page 215 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

forums.anandtech.com

M4 Family discussion here:

Page 263 - Discussion - Apple Silicon SoC thread

Page 263 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

forums.anandtech.com

 

[beginner99]

If you want a cheap APPLE COMPUTER buy an iPad.

tablets don't really count as "computers" to me. They are consumption devices. And don't come with your magic-keyboard non-sense. It's still an extremely limiting device in terms of connectivity and applications.

It's not Apple's fault that you insist on interpreting their computer lineup through the eyes of 2005 rather than the eyes of 2025...
(Again, this is not snark. I am trying to get you to see that Apple is skating to where the puck will be, is reinterpreting personal computing as a whole. If you insist that "they're doing it wrong" because what they're doing doesn't match what you're used to, well, complain all you like but remember the lessons of 2007.
Personal computing was reconfigured once before in the recent past -- to widespread mocking by the old guard who insisted nothing could or would change.)

So vendor lock-in is this great vision to redefine personal computing I should totally buy-in to?

I mean don't get me wrong, I'm 100% sure these ARM Macs will beat x86 in terms of performance/watt by a huge margin for many use cases. But as said tablets are already pretty limiting devices and being vendor-locked is just a no go for me. Your bascially limited to apples app-store for software and anything non-mainstream will directly be a problem.

EDIT:

And there is also this story about how apple treats their users privacy. Only a matter of time till a fixed installed keylogger that raises flags at the FBI if you type "dangerous" thoughts and no option to opt-out because closed-platform.

- Apple sends data home for every app you open, unencrypted
- Apple apps on macOS Big Sur bypass firewall and VPN connections to phone home

 

[Entropyq3]

Note regarding Cinebench 23 - it uses AVX-512, so the numbers generated will not correlate to very much.

 

[Heartbreaker]

Apple M1 vs Intel 6-core & AMD 580X in Affinity Photo:

View attachment 33718

original source

Now that is more Impressive. This is compared to an iMac, not a Macbook Pro.

I wouldn't buy an Intel Mac right now, unless you desperately need more than 16GB of Memory, and you need a new Mac now.

This is also the "slowest" ARM Mac that will ever be. It creates a very powerful, lowest common denominator of performance and functionality going forward.

I would expect there will be at least two levels above this first tier.

Tier two: Should be the highest end Macbooks, and non-Pro iMacs: This is should be 8+ performance cores and bigger memory options, and maybe an optional new Apple discrete GPU.

Tier Three: Mac Pro and and iMac Pro: This is murky. How will Apple do serious high core counts? Will get do chiplets, multiple CPUs, or a Massive Monolithic die. I expect a discrete GPUs for high end Metal compute, will be a must.

 

[awesomedeluxe]

The flood of benchmarks for the M1 have been pretty much all positive. But one small sour note seems to have gone underdiscussed.

The Mac Mini hits the same single-core performance targets as the Macbook Air in benchmarks. Of course, the Mini can doubtless sustain this performance for much longer. But I also see people making comparisons to AMD performance as if Apple could keep scaling up. "The M1 is a 15W part, Zen 3 will use such-and-such watts, imagine what Apple could do at such-and-such watts."

My imagination is failing me. What could they do? The Mac Mini used to support a TDP of 65W. If the M1 could be clocked up more, wouldn't Apple have already done so?

We know Apple is probably at the high end of the power curve and their cache sizes are already huge. Given that significant core arch changes are unlikely, what could the M1's midrange iteration do to improve single core performance? The M1's single core perf is incredible for its power budget, but it may lose a little luster if it doesn't improve much when it debuts in an iMac. It may even be... weaker than the comparable Zen 3 part

Edit: not to make this post much longer, but so no one wastes their breath - I do recognize the obvious direction to take an expanded power budget is moar cores. I am just pondering what, if anything, Apple could do to eke out a little more single core performance for parts that compete directly with Zen 3 desktop.

 

[moinmoin]

But regarding mac's and their walled garden, this is an example of a not-so-good approach:

Jeffrey Paul: Your Computer Isn't Yours

To be fair Windows 10 does that for quite some time as well with all its telemetry data. Microsoft on OS level even keeps turning it on every update, and for Office it can't be turned off completely to begin with. None of that is legal in the EU thanks to GDPR, but nobody bothers to change it or even talk about it.

Nowadays this kind of behavior seems to be widely considered acceptable when using commercial software.

 

[Eug]

Interview about M1 with:
Tim Millet, Apple VP Platform Architecture
Bob Borchers, Apple VP Worldwide Product Marketing

BTW, Bob Borchers is FaceTiming for the interview on an M1 MacBook Air.

tablets don't really count as "computers" to me. They are consumption devices. And don't come with your magic-keyboard non-sense. It's still an extremely limiting device in terms of connectivity and applications.

That's you, and in many ways, old thinking. Yes, they excel at consumption, but in 2020 for the majority of the population, consumption is the dominant computing "workload".

You have to remember, that for a large swath of the population, their main "computers" are iPads or even iPhones (or Android phones).

As for actual work, they are better at some things than others, so it really depends on your workflow. People always say traditional laptops are more flexible. Well, yes they are, but often they aren't necessarily the best tool for the job. They often are, but sometimes they aren't.

For example, I have a friend who is a professional web designer that works with design apps and Photoshop all day long on his iMac. However, when he wants to edit photos for his family and friends, he edits them on his iPad Pro. In fact, he told me one of the main reasons he needs Photoshop is not actually because he wants to edit in it. It's because people send him files that already in Photoshop format, and other applications don't always import those files properly.

Personally, for my mobile devices (not counting my iPhone) I use my iPad Pro about 95% of the time, and my MacBook about 5% of the time. It's probably more like 50/50 when I travel for work (which admittedly has been nil this year because of Covid), but note I say 50/50, meaning that I actually often take both devices with me. The MacBook is a jack of all trades, but for some stuff just isn't isn't as nice to use as the iPad Pro, despite the fact its form factor is my favourite Mac laptop form factor of all time.

The flood of benchmarks for the M1 have been pretty much all positive. But one small sour note seems to have gone underdiscussed.

The Mac Mini hits the same single-core performance targets as the Macbook Air in benchmarks. Of course, the Mini can doubtless sustain this performance for much longer. But I also see people making comparisons to AMD performance as if Apple could keep scaling up. "The M1 is a 15W part, Zen 3 will use such-and-such watts, imagine what Apple could do at such-and-such watts."

My imagination is failing me. What could they do? The Mac Mini used to support a TDP of 65W. If the M1 could be clocked up more, wouldn't Apple have already done so?

We know Apple is probably at the high end of the power curve and their cache sizes are already huge. Given that significant core arch changes are unlikely, what could the M1's midrange iteration do to improve single core performance? The M1's single core perf is incredible for its power budget, but it may lose a little luster if it doesn't improve much when it debuts in an iMac. It may even be... weaker than the comparable Zen 3 part

I'm no expert of course, but I get the impression they are not willing to bin too much based on clock speed. It seems the consensus here is that the higher power parts will add twice the number of performance cores, and increase the number of GPU cores as well. ie. More cores, not more MHz. Given that their single-core performance is already at/near the top of the heap, is that such a bad thing?

 

[xblax]

My imagination is failing me. What could they do? The Mac Mini used to support a TDP of 65W. If the M1 could be clocked up more, wouldn't Apple have already done so?

Is it already confirmed that the M1 may not take more than 15W in the Mac Mini? I'd imagine that they already increased the TDP compared to their Macbook ARM offerings.

We know Apple is probably at the high end of the power curve and their cache sizes are already huge. Given that significant core arch changes are unlikely, what could the M1's midrange iteration do to improve single core performance? The M1's single core perf is incredible for its power budget, but it may lose a little luster if it doesn't improve much when it debuts in an iMac. It may even be... weaker than the comparable Zen 3 part

Single core performance seems to be good enough already, but let's wait for more benchmarks beyond GB. The real question for me is if they can scale the number of cores with their wide architecture. Big monolithic dies on leading edge nodes have bad yield and I'm also not sure how their cache architecture and inter-core communication is able to scale.

 

[Roland00Address]

I'm no expert of course, but I get the impression they are not willing to bin too much based on clock speed. It seems the consensus here is that the higher power parts will add twice the number of performance cores, and increase the number of GPU cores as well. ie. More cores, not more MHz.

This is projection for I can not confirm this with evidence and I am too far removed to know if my mental imaginations are close to real.

But I would not be surprised this is apple trying to create a "consistent" experience and in the process giving up free performance. This is because if the developer assumes free performance for some skus and not all skus will remove the vigilance and diligence that developers should be doing to make sure their software is consistent across all devices and all skus.

Thus making a device 10% or 25% faster based off clockspeed alone merely because it is a desktop part and you can run at a higher clockspeed is not worth the "free performance" (with the only downsides for free is fan noise, power consumption, thermals, and heating the room) for if it enables sloppy developers being sloppy careless with this free performance. And this being careless can mean worse software across all devices but especially the devices that are on the lower end of the skus for they actually have to care about battery life and thermals the best. 10% to 30% faster based of clockspeed is not worth it if developers use sloppy code that performs half as well.

-----

But this can be wrong for it could merely be projection.

 

[Asterox]

As far i see, 5nm/Apple M1 is 8 Core CPU up to 3.2ghz, TDP is 15W.

We can compare it with 7nm Zen 2/Renoir APU 8/16 Ryzen 7 4800U, also 15W TDP.

Geekbench

Single-Core Score
1165
Multi-Core Score
7260

Apple M1 vs. AMD Ryzen 7 4800U - Benchmark and Specs

Apple M1 vs. AMD Ryzen 7 4800U - Geekbench 5, Cinebench R20, Cinebench R23, Cinebench R15 and FP32 iGPU (GFLOPS) benchmark results plus detailed technical data

www.cpu-monkey.com

LENOVO 82A2 - Geekbench

Benchmark results for a LENOVO 82A2 with an AMD Ryzen 7 4800U processor.

browser.geekbench.com

Is that impresive, well yes M1 IPC is very high no doubt.But this is ARM CPU for Apple OS system.

For example, Zen 3/Ryzen 7 5800U will be even more important CPU for Windows OS platform or Laptops.

In the future, we can imagine what we can expect from Zen 4/5nm Mobile 8/16 CPU or APU.

 

[moinmoin]

What are the pros and cons of either approach (width vs clocks) for a general usage in a consumer product?

Cerny for Sony at a recent PS5 event used the saying "a rising tide lifts all boats" while referring to clocks. At low clocks all the logic can be optimized to get at much performance as possible for said clock, making it wider. But that prevents the CPU from reaching higher clocks. Going back to the saying, being able to rise the clocks has the advantage that it essentially scales the performance of the whole CPU, including caches (especially L1 and L2 which are linked to clock cycles) etc., "a rising clock lifts all parts". The former is best for achieving the best possible energy efficiency, the latter is best for brute forcing performance (at the cost of energy efficiency).

Also, if somebody wanted it, could they integrate the same amount and kinds of specialist silicon in a x86 processor? Like, for video and photo processing?
Asking, because somebody implied that these specialist silicone parts account for MOST, or at least MUCH of the efficiency and performance advantage M1 seems to have, yet, I am fairly certain that vast majority of normal consumers will never use there, ever, so not sure if there is incentive for Intel/AMD to even try this approach.

The big issue with specialization is that the more specific the acceleration the less likely it is being used.

A combined hardware/software manufacturer like Apple is in the best possible position there: It can track how hardware is being used by the software and over the time accelerate all the slow parts of the software in its hardware. Since it controls all the parts, including OS, libraries and all interfaces, it can transparently introduce hardware improvements over the time even for 3rd party software which builds upon those libraries and interfaces.

If hardware, software and OS manufacturers are all separate the coordination necessary to achieve similar acceleration is much harder.

 

[Eug]

Interview about M1 with:
Tim Millet, Apple VP Platform Architecture
Bob Borchers, Apple VP Worldwide Product Marketing

BTW, Bob Borchers is FaceTiming for the interview on an M1 MacBook Air.

Heh. Not surprisingly, they wouldn't provide a number for TDP equivalency. They did mention 10 and 15 Watts though in passing after being prompted by the interviewer but didn't actually say that what the real comparable numbers were.

Wouldn't comment on clock speeds either.

Confirmed LP-DDR4, unified memory.

ML will be utilized most prominently in image processing like Photos, Photoshop, Final Cut Pro, etc. but they claim they've leveraged it throughout the OS to a lesser extent.

Battery sizes are roughly the same as previous models, so extended battery life due to new platform.

The claim that any apps that use Metal will be faster on M1 Rosetta 2 than on Intel native. So major caveat there. Implies other apps that don't use Metal will be slower on M1 Rosetta 2 (which isn't a major surprise I guess). However, one statement was that if it was running on an Intel machine last year and M1 on this year, you shouldn't notice much of a difference in regular mainstream usage. So far for example, all these apps work fine in Rosetta 2: Adobe Creative Suite, MS Office, WoW, Omni, Slack, Discord, Dota 2, Pixelmator, Tomb Raider, Dirt Rally, 1Password. (Of course, that leaves a bazillion other apps unmentioned though.)

"We're still friends with Intel."

 

[amrnuke]

I've said this a thousand times, FFS.
Better branch predictors. Better prefetching. Better cache placement/replacement algorithms. Better ISA. More aggressive instruction fusion. Better fetch engine. Better TLB reach. Better memory controller.

How can you not comprehend this? If Apple substantially exceeds AMD's PPC, doesn't that mean that they do better than AMD on pretty everything that determines CPU performance? How else do you think the numbers work out?

"They do better than AMD on pretty [much] everything that determines CPU performance" - except AMD have the core and chip design that performs the fastest at single-threaded and multi-threaded operations, correct?

I know that you're smarter than me about this, so I know that you know that a large amount of uarch design decisions are tailored to the ISA. For example if you're using an ISA that has predication then you make vastly different uarch decisions than if you're using an ISA that doesn't. You also know that most uarch design decisions are made in the context of the target market.

That they've built an excellent core around a different ISA, creating a package that you perceive to be better (again, define "better"!) does not make them any more or less "smart" than AMD, nor does it make the Firestorm core design any more or less "smart" than the Zen3 core design. They're just building around different ISAs for different (granted, converging) markets and both are doing a hell of a job doing so.

A 9-figure market wouldn't get Apple out of bed, Airpods alone are $12 billion in revenues per year.

I misspoke. It's a 12-figure market.

 

[awesomedeluxe]

I'm no expert of course, but I get the impression they are not willing to bin too much based on clock speed. More cores, not more MHz. Given that their single-core performance is already at/near the top of the heap, is that such a bad thing?

I may staring a little too hard into the void. But we're so used to seeing x86 parts just go off as their thermal headroom increases, and the conventional wisdom that while laptops are good enough for most tasks there's no replacement for being able to crank desktop power consumption through the roof as Intel and AMD processors spike to 200W for "turbo" speeds.

And we knew going into this that Apple and ARM designs were deeply rooted to a different part of the power curve and weren't going to scale up the same way. But it's still weird. The Air used to use a 9W part and the Mini used a 65W part. In my mind I'm thinking, the difference in peak performance should be enormous! But with the M1, it's zero. Maybe I'm looking at this the wrong way; it's as much the case that ARM designs don't scale up impressively as it is that x86 designs scale down poorly.

Is it already confirmed that the M1 may not take more than 15W in the Mac Mini? I'd imagine that they already increased the TDP compared to their Macbook ARM offerings.

Single core performance seems to be good enough already, but let's wait for more benchmarks beyond GB. The real question for me is if they can scale the number of cores with their wide architecture. Big monolithic dies on leading edge nodes have bad yield and I'm also not sure how their cache architecture and inter-core communication is able to scale.

I mean, I don't want to get too into the weeds with TDP. As in the interview above Apple isn't actually up front with that information, but we know the range. The important bit is that heavy benchmarks on the Mini and the Air, where the M1 is going full throttle, produce the same single core performance in both devices. So the power ceiling Apple is willing to hit doesn't get any higher with the Mini.

Which is weird, right? The Mini wouldn't be under any thermal stress in those conditions. What I assume stops Apple from going higher is that it's hit that part of the exponential power curve where increasing consumption by 20+ W just doesn't accomplish anything besides creating heat/area liabilities.

To respond to your concern about yield - I think part of the answer is right in your own paragraph - "leading edge." TSMC's new processes improve quickly, and it's a safe bet Apple had that in mind when thinking about when to start production of the larger, multicore designs. So I'm not too concerned about Apple's ability to produce an 8 perf core part next year. I do think we'll continue to see some disabled GPU cores as Apple salvages parts with non-fatal defects.

As for even larger APUs for machines with the word "Pro" but not the word "Book," I'm not sure what the approach will be. It's possible Apple will keep the die size smaller by moving the GPU off-die in a chiplet design. They might also make these on a more mature process - it's probably acceptable to have your Mac Pro APU rolling out on 5NP while the A16 is taping out on N3.

 

[LightningZ71]

Just some speculation here: Looking at some of the latencies in the AT article on the A14, could it be that some of them are so tight that they start to not be viable at clock speeds much higher than what we're seeing on the existing M1 chips now? If that's the case, then having extra TDP available won't help much at all as it won't allow higher clocks as they will be unreachable due to design latencies that are internal to the chip architecture as implemented on N5? In other words, the A14/M1 is designed around the notion that it won't clock higher than roughly 3.5Ghz ever, and, as such, has tightened down every possible latency metric / function timing in the core as much as possible to wring the maximum amount of performance from them at their target clock speeds?

I'm not implying that this is necessarily a bad thing, just a design choice given the target market of the original product. It certainly isn't hurting their performance in benchmarks and is allowing excellent power consumption metrics.

 

[Eug]

And we knew going into this that Apple and ARM designs were deeply rooted to a different part of the power curve and weren't going to scale up the same way. But it's still weird. The Air used to use a 9W part and the Mini used a 65W part. In my mind I'm thinking, the difference in peak performance should be enormous! But with the M1, it's zero. Maybe I'm looking at this the wrong way; it's as much the case that ARM designs don't scale up impressively as it is that x86 designs scale down poorly.

Yeah but Apple uses the same enclosure and cooling for vastly different CPU TDPs. This won't change.

They've maxed out the MacBook Air now, and most of think that the Air will throttle significantly under sustained load. The M1 Mac mini probably won't throttle much, as it has thermal headroom to spare, already ready for new 8 perf-core parts next year.

This has actually factored into how I've chosen my machine in the past. I am of the belief that the top end 95 W Intel desktop CPU parts are actually inappropriate for the iMac 27" enclosure, at least in terms of my preferences. The iMac Pros have great quiet cooling, but the iMacs get quite loud with the top end Intel chips Apple uses. OTOH, they have the exact same cooling for the 65 Watt parts, and the iMacs stay much quieter for much longer with those 65 Watt parts.

I bought an iMac i7-7700K in 2017, but it really bugged me that every time I had a sustained load for longer than 30 seconds, the fan would hit the maximum. It gets annoying that just exporting my kid's 1 minute birthday song video would max out the fan. So, I returned it, and bought an i5-7600, and all that annoying noise disappeared.

Apple will add more cores to the Mac mini, but will likely keep the same cooling solution. I wouldn't be surprised if they do increase the clock speed somewhat though. Not like Intel, but that's OK of course. However, if they don't increase clock speed, that's probably fine too.

tl;dr:

The Mac mini is not built for M1. It's built for next year's M1X or whatever it will be called. But the good news is that M1 likely won't throttle significantly in the Mac mini, and the M1 Mac mini will probably be a quiet machine.

 

[Doug S]

A 9-figure market wouldn't get Apple out of bed, Airpods alone are $12 billion in revenues per year.

Mac pro/iMac volumes alone can't support the R&D necessary for a bespoke cpu design, x86 or ARM, and Apple, not being Intel and AMD, don't care to design and manufacture CPUs only to sell to third parties.

Perhaps today's Mac Pro volumes can't. What if the Mac Pro is significantly faster than any x86 workstation you can buy? How many new buyers does that attract? How many 'halo' sales could those new Mac Pro buyers ultimately be responsible for - i.e. a company that switches its engineers to Mac Pros might eventually switch the rest of the company to Macs. I'm not saying Macs are going to give Windows PCs a run for their money, but one shouldn't assume that their current market share is set in stone forever.

Anyway they don't need a bespoke CPU design, the cores they have are plenty fast they just need more of them. Whether they do it via chiplets or a monolithic design remains to be seen, but this is hardly the difficult problem some people want to make it out as. It is much, much harder to design a CPU with market leading single thread performance than it is to put a lot of CPUs (of any design) onto the same chip/MCM.

 

[Doug S]

My imagination is failing me. What could they do? The Mac Mini used to support a TDP of 65W. If the M1 could be clocked up more, wouldn't Apple have already done so?

The Mac Mini is the least expensive Mac, there is such a thing as market segmentation you know.

 

[xblax]

I mean, I don't want to get too into the weeds with TDP. As in the interview above Apple isn't actually up front with that information, but we know the range. The important bit is that heavy benchmarks on the Mini and the Air, where the M1 is going full throttle, produce the same single core performance in both devices. So the power ceiling Apple is willing to hit doesn't get any higher with the Mini.

Which is weird, right? The Mini wouldn't be under any thermal stress in those conditions. What I assume stops Apple from going higher is that it's hit that part of the exponential power curve where increasing consumption by 20+ W just doesn't accomplish anything besides creating heat/area liabilities.

I assume the power ceiling could be much higher than 15W in Mac Mini and Air - just not for sustained loads. The new sustained load benchmark for Cinebench R23 was probably added exactly for that reason.

I'm very curious what power consumption will actually be measured for the Mac Mini - it should be much easier to get accurate measurements without built-in battery. Based on the data they provided I wouldn't be surprised if it's not much more efficient than Renoir in the end.

 

[name99]

You said you want real world benchmarks? OK then:

https://twitter.com/x/status/1326866126635143169

 

[Eug]

^^^ Is the CPU test pure CPU, or is it also leveraging ML?

Affinity Photo is one of the apps some of the lower end users of photo editing software are moving to, after abandoning Photoshop.

What's holding others back is that it's still a little rough around the edges, and it doesn't import complex Photoshop .psd files properly. OTOH, it's way better than something like GIMP, which IMO is horrible.

 

[Mopetar]

calling a $999 device low-end. that is the problem.

It's the low end for Apple though. Guess what a low end Ferarri costs. Bet it's more than a low end Ford.

But if their claims are to be believed then it's better than 98% of other laptops, so in a way it's also high end. It's just Apple's low end of the high end of the market.

Expecting something better than 98% of laptops for $700 is just wishful thinking. Apple wouldn't be able to supply that or would have to start building their products more like the budget or bargain basement crud that some people want to avoid.

I would have liked to se them lower prices because they save a lot on using their own chips and not paying Intel margins. However, they're also just starting off so perhaps that comes later as they ramp their production across the entire line. Either way, it's some amount of performance at some price. That's either good or bad regardless of anything else.

Also x86 isn't dying. There's one company that's competing with AMD/Intel and they aren't selling their SoCs to anyone else. Apple doesn't care about getting 50% of the market if they have to slit their wrists to do so. They'll occupy the luxury brand area and sell to 10% of consumers for $200 in profit as opposed to 90% for $20 profit.

 

[amrnuke]

You said you want real world benchmarks? OK then:

https://twitter.com/x/status/1326866126635143169

The CPU comparison is an 6-core, 6-thread i5-9600KF, correct?

 

[coercitiv]

Guess what a low end Ferarri costs. Bet it's more than a low end Ford.

Go and ask for a low-end Ferrari in a respectable shop, see how fast the door finds you.

Expecting something better than 98% of laptops for $700 is just wishful thinking. Apple wouldn't be able to supply that or would have to start building their products more like the budget or bargain basement crud that some people want to avoid.

Nobody asked Apple to sell their MBA for $700, instead somebody argued the MBA is low-end, which is not true.

 

[Roland00Address]

Not now, but I am willing to bet money in the next 5 years we are going to get $700 macbook airs, or the ipads will gain far more features with software where they are effectively mini macbook airs running iOS.

• Yes this is a large timeframe ...
• but the way I see it is 10 years+1 month ago was macbook air version 2, and the ...
• ipad 1 was 10 years+7 months ago,
• thus we are talking predictions of the 3rd era if we break up this 10 year old market into 5 year segments.

In some ways we are living the Steve Jobs dream of Mac Pro capabilities in something you can put into an envelope. (Thank You Apple Silicon!)

 

[deathBOB]

So the low end products all get the M1. Will all of the high end products share a big M1X? At what volume does it make sense for Apple to make chip just for something like the Mac Pro?