Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Right? We were expecting Intel's benchmarks to show what we already all knew, which is that the M1 smokes Intel chips. Then they'd unceremoniously admit defeat and curl up and die...

Or in the real world, where every company does this, it really isn't that shocking. Apple is one of the kings of hyperbole and overstatements. Intel's just trying to be geeky about it. Problem is, facts are harder to fake than grand and vague statements about magic and snappiness.
Oh everybody expected Intel to try to spin this - they just didn't expect them to do such a transparent effort that a blind person could see through it. In short it was not that Intel tried to spin it but they did such a crappy job that has everyone talking.
 
That a sea of Windows users are going to go buy Apple devices and learn a completely new environment to theoretically get a 20% performance gain on MS Office or web surfing or streaming? Oh, and while paying a lot more money for that Apple laptop or desktop.
They not only get that, but also an increase by 200-400 % in battery life. Big deal, coz now you really do not have to bring your charger.
Plus the M1s actually are competitive price wise
 
While I agree with most of the comments here about Intel getting beat for battery and performance, at the end of the day Macs only make up, at most, about 9% currently of the personal computer market. Historically, it's hovered around 5%. Now, I'm sure Intel would like to keep Apple as a customer, but losing Apple really isn't that big a deal when Apple is selling so few machines that require a CPU that Intel could provide. Also, remember that this isn't the first time Apple has stormed away from a chip provider for another provider.
If you count mobile devices as personal computers (and at least one court does) ARM not intel dominates the landscape
Lastly, we don't have Apple's product roadmap for iOS and MacOS devices. It's quite possible that Apple was demanding too much from Intel (both pricing and re-architecting something such as cross platforming iOS and MacOS) and Intel said no.
Intel effectively when Wrongway Peachfuzz with its roadmap showing that making a roadmap and sticking to it are two wildly differently thngs, That is what ticked off Apple in the first place - Intel's roadmap said they would have a CPU of the desired wattage and dies by year x and Apple designed their future hardware around that.

From a business prospective why in the name of sanity would Apple provided a road map giving their competitor's the time needed to come up with counterarguments?

For all you claiming that Intel is worried, what is Intel worried about? That a sea of Windows users are going to go buy Apple devices and learn a completely new environment to theoretically get a 20% performance gain on MS Office or web surfing or streaming?
No but that it is clear the x86 is not the future of computing and that is about the only arrow Intel has in its quiver. AMD has an ARM processor out, NVIDIA has bought ARM, Microsoft has a working version of Windows for ARM for at least two years, more and more dataservers are going to ARM for energy reasons, and ARM is cheaper.

The only things holding ARM back was x86 emulation/translation was, to put it bluntly, crap and it used more RAM then x86. Rosetta 2 shows the first can be overcome and SSD with proper memory compression as well as dirt cheap RAM makes the second largely irrelevant.
 
  • Like
Reactions: Unregistered 4U
Right? We were expecting Intel's benchmarks to show what we already all knew, which is that the M1 smokes Intel chips. Then they'd unceremoniously admit defeat and curl up and die...

Or in the real world, where every company does this, it really isn't that shocking. Apple is one of the kings of hyperbole and overstatements. Intel's just trying to be geeky about it. Problem is, facts are harder to fake than grand and vague statements about magic and snappiness.

It's not about that, it's that Intel is being directly dishonest and tries to mislead the potential customer. All third-party benchmarks show M1 laptops confidently outperforming any Tiger Lake device in WebXPTR3 — by a considerable margin. But in benchmarks published by Intel somehow their CPU is 30% faster on average. How is this possible? Answer is (in part — check below) simple: Chrome is slow on Mac. Running the benchmark using Chrome Canary scores 244, while running it in Safari scores 284.

Some WebXPRT3 figures:

Intel i7-1185G7 — 247 (via anandtech), 240-260 (via notebookcheck)
Apple M1 — 280 (via notebookcheck), 284 (my own score for M1 Pro in Safari 14.1), 244 (my own score using Google Canary 90)

You know what's really funny though? Intel claims that their own CPU is 3 times faster in the photo enhancement test! But the scores in both Safari and Chrome are identical on my M1... so there is surely something else going on. Besides, 244 is not very different from 247 measured for the Intel chip by Anandtech... even more reason to suggest dirty play from Intel.
 
  • Like
Reactions: Maximara
It's not about that, it's that Intel is being directly dishonest and tries to mislead the potential customer. All third-party benchmarks show M1 laptops confidently outperforming any Tiger Lake device in WebXPTR3 — by a considerable margin. But in benchmarks published by Intel somehow their CPU is 30% faster on average. How is this possible?

I'm really more annoyed by the battery life thing.

They show all the performance of the 1185G7… then when it comes to battery life, switch to an entirely different model using the 1165G7.
 
I'm really more annoyed by the battery life thing.

They show all the performance of the 1185G7… then when it comes to battery life, switch to an entirely different model using the 1165G7.

And they run the video test using completely different settings, yeah... that should be considered fraudulent by any measure.
 
8th gen i7 (2+ years old). I bought it as a cheaper alternative to a MacBook Pro (with the sale I got, it was 1/2 the price of a comparable Mac). In hindsight, I might make the same choice but I miss having a Mac laptop and using macOS. Besides, a desktop fills my Windows and Linux needs (although the AMD Ryzen 5000 mobile processors keep me interested in non-Apple laptop alternatives).
here is something nice i7-1185G7 for
1149€

(replaceable pcie-4 nvme ssd, 16 gb mem, 512 ssd, 2-thunderbolt 4, 1 usb, 1 memory card slot, 14")

 
Intel effectively when Wrongway Peachfuzz with its roadmap showing that making a roadmap and sticking to it are two wildly differently thngs, That is what ticked off Apple in the first place - Intel's roadmap said they would have a CPU of the desired wattage and dies by year x and Apple designed their future hardware around that.
Yeah, they were “demanding too much”.. :) Perhaps “Honoring your own set roadmap” was really crossing the line! :D
 
Intel i7-1185G7 — 247 (via anandtech), 240-260 (via notebookcheck)
Apple M1 — 280 (via notebookcheck), 284 (my own score for M1 Pro in Safari 14.1), 244 (my own score using Google Canary 90)
So at best, the i7 should be less than 10% faster than the M1, not 32%.
It's like the Tomb Raider test where they show a 10% lead fo the M1, while LTT found a 50% lead. LTT used an i7 1165G7, but this has the same GPU as the 1185G7, with only a 50 MHz slower boost clock.
I suspect that intel used a platform (their "white box") with much higher cooling capacity than a regular laptop.
 
At this point, it might make more sense to derive from RISC-V.
With Intel they might think you mean "risky" and not do it. :D
Either way, moving archs historically isn’t their strong point. They keep going back to x86.
Because up til now doing so served them so well. Back when RAM was expensive and there was no compression for disk paging for slow HDs x86 made sense but in today's world all that is gone.
 
Here's why I think Intel could be in trouble long-term. These carefully selected and biased benchmarks aside, Intel has legitimate issues related to both process shrink and maintaining backwards compatibility with 32-bit (and even 16-bit in some cases) applications. Intel benefits from AMD using x86 due to ongoing licensing agreements between the two companies, and AMD benefits because Intel licenses x64 from Sunnyvale. If AMD was to start making ARM-based chips for Windows-based machines, it would introduce additional competition for Intel as well as competition for companies like MediaTek and Qualcomm (who are making ARM-based processors for Windows and Chrome OS alike). The advantage AMD has over Intel is the TSMC partnership. TSMC has the experience and facilities for ARM production, and even if Apple were to keep buying up the first year or two of a new process run, AMD could still stay ahead of Intel like they are now with 7nm Ryzen parts. Intel could make a switch to ARM themselves, but they would have to either outsource to a third party or retool their existing fabs to accomodate the new designs. Given that Intel can't even get everything on 10nm right now (fragmentation within the Intel product lineup right now rivals Apple at the end of Gil Amelio's run), expecting them to successfully transition to 7nm or smaller is a gamble under the best-case scenarios.
 
  • Like
Reactions: Maximara
Here's why I think Intel could be in trouble long-term. These carefully selected and biased benchmarks aside, Intel has legitimate issues related to both process shrink and maintaining backwards compatibility with 32-bit (and even 16-bit in some cases) applications. Intel benefits from AMD using x86 due to ongoing licensing agreements between the two companies, and AMD benefits because Intel licenses x64 from Sunnyvale. If AMD was to start making ARM-based chips for Windows-based machines, it would introduce additional competition for Intel as well as competition for companies like MediaTek and Qualcomm (who are making ARM-based processors for Windows and Chrome OS alike). The advantage AMD has over Intel is the TSMC partnership. TSMC has the experience and facilities for ARM production, and even if Apple were to keep buying up the first year or two of a new process run, AMD could still stay ahead of Intel like they are now with 7nm Ryzen parts. Intel could make a switch to ARM themselves, but they would have to either outsource to a third party or retool their existing fabs to accomodate the new designs. Given that Intel can't even get everything on 10nm right now (fragmentation within the Intel product lineup right now rivals Apple at the end of Gil Amelio's run), expecting them to successfully transition to 7nm or smaller is a gamble under the best-case scenarios.

Fabs don’t have to be retooled for new designs, and tsmc having experience with Arm is not an advantage for TSMC.

The way it works is that the fab provides a set of design rules, then the chip designers provide mask data that define each layer of the design, in a standard format. The fab (eg TSMC) has no idea what the design is, whether it is Arm or something else. They simply process the mask data into the appropriate masks, and run the silicon through the process.
 
  • Like
Reactions: leman
The advantage AMD has over Intel is the TSMC partnership. TSMC has the experience and facilities for ARM production, and even if Apple were to keep buying up the first year or two of a new process run, AMD could still stay ahead of Intel like they are now with 7nm Ryzen parts.

Intel is reportedly outsourcing some of their CPU production to to TSMC as well. The big question is whether TSMC has enough capacity to accommodate everyone who’s interested. Apple has an advantage here: they dint sell the CPUs, so they can afford to big higher for their production. Intel and AMD, not so much
 
And they run the video test using completely different settings, yeah... that should be considered fraudulent by any measure.
I think the settings are the same between the two computers tested - the air and the swift - just the computers themselves aren’t the same ones as in the performance measurements, which is obviously sketchy. Unless I’m missing something else?

Of course one could question the use of the swift 5 at all. While the swift 5 is not the worst comparison for the air ... it’s not great, actually it’s pretty bad. It’s driving a 1080p LCD panel with 10% bigger battery. It keeps its weight down by have a magnesium body that flexes a little under use. So, yeah, matching the M1 air driving a 2560x1600 display in battery life in a video test at 250 nits becomes *a lot* less impressive for the Intel chip.
 
Last edited:
Intel is reportedly outsourcing some of their CPU production to to TSMC as well. The big question is whether TSMC has enough capacity to accommodate everyone who’s interested. Apple has an advantage here: they dint sell the CPUs, so they can afford to big higher for their production. Intel and AMD, not so much
Also Apple’s die are on average much smaller, so they get more per wafer start.
 
Not every company does this.
Every company tries to show their product in the best light, it's just some are better at it than others, and most don't outright lie, but perhaps exaggerate. Intel definitely went too far here, but don't pretend it isn't just an overextension of a very common corporate tactic.
 
Last edited:
  • Like
Reactions: Maximara
well it was truthful and not targeted in the way the intel benchmarks are. Anyone that owns an m1
and intel know what the truth is.
Yes it was truthful. Though the important part is your correct in how it was targeted. Apple wanted to say, M1 is the best. Intel is trying to say M1 is crap because it can't do what Intel does. Totally different perspectives there.
 
Intel is reportedly outsourcing some of their CPU production to to TSMC as well. The big question is whether TSMC has enough capacity to accommodate everyone who’s interested. Apple has an advantage here: they dint sell the CPUs, so they can afford to big higher for their production. Intel and AMD, not so much
I thought they were still trying to decide between TSMC, Samsung, and Qualcomm? Apple and AMD already have most of the smaller processes booked out for at least another year, so Intel would continue to lag behind on a larger process than either competitor.
 
I thought they were still trying to decide between TSMC, Samsung, and Qualcomm? Apple and AMD already have most of the smaller processes booked out for at least another year, so Intel would continue to lag behind on a larger process than either competitor.
Qualcomm doesn‘t have a fab, so using them would be a creative choice.
 
Actually, yes. Okay, I can't comment on the exact ratio but a lot of this is due to the fact that macOS on the M1 (and iOS/iPadOS on the A-series chips too) will make *very* heavy use of the SSD as a swap and is *very* optimized for that. Obviously this has limits, but it is shocking how well it seems to work.

Maybe people didn't look that hard previously, but it doesn't seem to do it as much on Intel chips - though technically I can't think of a reason why it couldn't if the computer has a fast enough SSD (someone more knowledgeable than me can chime in here) ... so as far as I know this isn't really about the CPU itself and it's possible macOS on an Intel CPU will behave the same way in Big Sur? Dunno. Anyway, this can have negative repercussions for SSDs (important since these computers are meant to last and the SSDs are not replaceable!), so we'll just have to see SSD lifetimes on the new machines as they age.

CISC vs RISC doesn't matter for this. CISC code can be smaller by about 10-20%, but code size such is a trivial part of RAM usage these days that this doesn't really matter anymore.
Reading some more, I think x86 macs will do the same.
 
They're being delusional. If they don't change attitude and soon - like a couple months - they'll be done for good.
M1 is already amazing, and we can already imagine what future gens will be like.
Can we, really? Why can’t we equally presume we’re already seeing the primary benefits of SOC in this chip? It’s not something new since they’ve been doing it for some time on the iPads. And adding cores won’t make much difference. Possibly higher clockspeeds or more RAM might help but are we really expecting anything huge? If so why?
 
Can we, really? Why can’t we equally presume we’re already seeing the primary benefits of SOC in this chip? It’s not something new since they’ve been doing it for some time on the iPads. And adding cores won’t make much difference. Possibly higher clockspeeds or more RAM might help but are we really expecting anything huge? If so why?

Because this chip was aimed at the low end machines, and easily scales in core count, frequency, and bus bandwidth.
 
Can we, really? Why can’t we equally presume we’re already seeing the primary benefits of SOC in this chip? It’s not something new since they’ve been doing it for some time on the iPads. And adding cores won’t make much difference. Possibly higher clockspeeds or more RAM might help but are we really expecting anything huge? If so why?
Because Apple engineers are not stupid, are they?

They woulnd‘t announce the change of platform for its entire line of computers without the prospect of a proper replacement at hand. They very well know an iPad chip won‘t do, so the conjecture they are cooking up more than just the M1 seems petty valid
 
  • Like
Reactions: Maximara
Can we, really? Why can’t we equally presume we’re already seeing the primary benefits of SOC in this chip? It’s not something new since they’ve been doing it for some time on the iPads. And adding cores won’t make much difference. Possibly higher clockspeeds or more RAM might help but are we really expecting anything huge? If so why?
It’s easy to presume anything. I mean, we have folks doubting the general shape that celestial bodies form when massive enough to be strongly affected by gravity, even with volumes of data available. So yes, you can presume that the processors Apple’s making today will NOT perform any better over the coming months and years. Even when processors are released that do APPEAR to show a performance increase, you can presume that it’s because all the software companies and individuals running tests have gotten together to REALLY tweak their code specifically so it’ll run better, NOT because the chip actually performs better. When you get your hands on one and it seems it performs better, you can presume that it’s because Apple’s ONTO you, and shipped you a special version of the system that actually has an Intel chip inside. When you open it up and see it’s an M1 processor, you can presume that when you stepped away to get your tools, Apple swapped out the motherboard AND that person will be there to swap back in the faster Intel one as soon as you blink when securing the last screw upon closure.

So, you can equally presume ANYTHING you like!
 
Because this chip was aimed at the low end machines, and easily scales in core count, frequency, and bus bandwidth.

To expound on this point, think of the M1 as the baseplate for a Lego set. Apple SoCs are designed for scalability in that you can easily add additional cores, RAM, and related components (such as TB3 controllers) to the baseplate and create something much more powerful and useful in the process.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.