Intel Benchmarks Say Apple's M1 Isn't Faster (pcworld.com) 260
PCWorld reviews Intel's recently-released benchmarks claiming Apple's M1 isn't faster than their 11th gen Core i7-1185G7 processor, among other things. Here are the claims Intel makes (visit the article to read PCWorld's "take" on each claim): MacBook M1 is slower than Core i7: Intel says in the WebXPRT 3 test, using the same version of Chrome for both the Core i7 system as well as the Arm-native MacBook, Intel takes the lead. The Intel chip was largely ahead in WebXPRT 3, and the x86 chip was nearly three times faster in finishing the photo enhancement test. Intel doesn't just use WebXPRT 3, though. It also shows the Core i7 pummeling the M1 in a PowerPoint-to-PDF export, and in multiple Excel macros by a factor of 2.3x. And yes, Intel used the Arm-native versions of Office for its tests.
Core i7 Crushes M1 in AI: For content creation tasks, Intel showed the Core i7 to be about 1.12x faster than the M1 in performing a 4K AVC-to-HEVC/H.265 file conversion. In this benchmark, they had the MacBook using the M1-native version of Handbrake. But the real destruction happens once you get to Topaz Lab's Gigapixel AI and Denoise AI, with the Intel Core chip crushing the M1 in AI-based noise removal and enlargement. Or maybe "crushing" is too nice a term, as it's more like the Core i7 outpaces the M1 by so much, the M1 wishes it had never been designed.
M1 doesn't support all the features: Intel also gives itself the lead in Adobe Premiere Pro, using the beta M1 native version in Auto Reframe, exporting to H.264 and H.265. They're decent wins, but come on, the code is still in beta for the Mac. That said, Intel points out that important features like Content Aware Fill are outright disabled on the beta version, and that's a concern. If the native version of Photoshop comes out, and there are critical features missing from it, that's a huge problem for Apple (and Adobe).
You can't be faster if you can't run it: For gaming, we see a bit of a back and forth between the Apple M1 and Core i7 in games that actually work on the MacBook. Intel doesn't let it end there, though, and decides to embarrass Apple further by showing the numerous games where the MacBook scores a 0 because game support just doesn't exist. Intel points out that "countless more" games "don't run on the M1," and then for good measure, it rushes Apple's bench with a list 10 more games you can't play on the M1 MacBook: Overwatch, Crysis Remastered, Halo MCC, Red Dead Redemption 2, PUBG, Monster, Hunter World, Doom Eternal, Microsoft Flight Simulator 2020, Apex Legends, and Rainbow Six Siege.
MacBook wouldn't win Evo certification: You know that fancy Intel Evo program that tries to improve laptop performance in key areas that annoy consumers? Well, Intel pretty much says that if Apple submitted the M1 MacBook to the same program that Asus, Dell, HP, Lenovo, MSI, Acer and others go through, it would be rejected. The reason? Intel says the M1 MacBook is too slow in doing things that anger consumers, such as "switch to Calendar" in Outlook, "start video conference Zoom" and "select picture menu" in PowerPoint.
Great battery life?: Perhaps the most shocking claim Intel showed deals with battery life. While performance tests can be cherry picked by those looking to prove an outcome, battery life usually can't be disputed. Apple's official claim gives the M1 MacBook up to 18 hours of battery life using Apple TV app to watch a 1080p video with the brightness set to "8 clicks from the bottom." Apple also claims up to 15 hours browsing 25 "popular" websites with the same "8 clicks" criteria. When Intel pitted a MacBook Air M1 against an Acer Swift 5 with a Core i7-1165G7, however, it found both basically dead even. The MacBook Air came in at 10 hours and 12 minutes, and the Acer Swift 5 lasted 10 hours and 6 minutes. The difference? Intel said it used Safari to watch a Netflix stream with tabs open with the screen set to a relatively bright 250 nits. On the Acer, Safari was subbed out for Chrome, but the brightness and Netflix remained the same. Intel did add that Apple's "8 clicks up" is about 125 nits of brightness on the MacBook Air which is pretty dim.
All kinds of things just don't work on the M1: Intel didn't just get into the performance of the M1. It also said it found the MacBook Pro had serious shortcomings, such as an inability to use more than one display with a Thunderbolt dock. And while the PC can use gaming headsets, eGPUs, a third-party finger print reader, Wacom Drawing tablet and Xbox Controller, Intel said it found the MacBook Pro simply doesn't work with eGPUs, and had multiple issues with other devices. That's just hardware incompatibility. Intel's rap battle with Apple also highlights issues with plug-ins for Ableton, Bitwig Studio, Avid Pro Tools, FL Studio, Motu and many others.
Core i7 Crushes M1 in AI: For content creation tasks, Intel showed the Core i7 to be about 1.12x faster than the M1 in performing a 4K AVC-to-HEVC/H.265 file conversion. In this benchmark, they had the MacBook using the M1-native version of Handbrake. But the real destruction happens once you get to Topaz Lab's Gigapixel AI and Denoise AI, with the Intel Core chip crushing the M1 in AI-based noise removal and enlargement. Or maybe "crushing" is too nice a term, as it's more like the Core i7 outpaces the M1 by so much, the M1 wishes it had never been designed.
M1 doesn't support all the features: Intel also gives itself the lead in Adobe Premiere Pro, using the beta M1 native version in Auto Reframe, exporting to H.264 and H.265. They're decent wins, but come on, the code is still in beta for the Mac. That said, Intel points out that important features like Content Aware Fill are outright disabled on the beta version, and that's a concern. If the native version of Photoshop comes out, and there are critical features missing from it, that's a huge problem for Apple (and Adobe).
You can't be faster if you can't run it: For gaming, we see a bit of a back and forth between the Apple M1 and Core i7 in games that actually work on the MacBook. Intel doesn't let it end there, though, and decides to embarrass Apple further by showing the numerous games where the MacBook scores a 0 because game support just doesn't exist. Intel points out that "countless more" games "don't run on the M1," and then for good measure, it rushes Apple's bench with a list 10 more games you can't play on the M1 MacBook: Overwatch, Crysis Remastered, Halo MCC, Red Dead Redemption 2, PUBG, Monster, Hunter World, Doom Eternal, Microsoft Flight Simulator 2020, Apex Legends, and Rainbow Six Siege.
MacBook wouldn't win Evo certification: You know that fancy Intel Evo program that tries to improve laptop performance in key areas that annoy consumers? Well, Intel pretty much says that if Apple submitted the M1 MacBook to the same program that Asus, Dell, HP, Lenovo, MSI, Acer and others go through, it would be rejected. The reason? Intel says the M1 MacBook is too slow in doing things that anger consumers, such as "switch to Calendar" in Outlook, "start video conference Zoom" and "select picture menu" in PowerPoint.
Great battery life?: Perhaps the most shocking claim Intel showed deals with battery life. While performance tests can be cherry picked by those looking to prove an outcome, battery life usually can't be disputed. Apple's official claim gives the M1 MacBook up to 18 hours of battery life using Apple TV app to watch a 1080p video with the brightness set to "8 clicks from the bottom." Apple also claims up to 15 hours browsing 25 "popular" websites with the same "8 clicks" criteria. When Intel pitted a MacBook Air M1 against an Acer Swift 5 with a Core i7-1165G7, however, it found both basically dead even. The MacBook Air came in at 10 hours and 12 minutes, and the Acer Swift 5 lasted 10 hours and 6 minutes. The difference? Intel said it used Safari to watch a Netflix stream with tabs open with the screen set to a relatively bright 250 nits. On the Acer, Safari was subbed out for Chrome, but the brightness and Netflix remained the same. Intel did add that Apple's "8 clicks up" is about 125 nits of brightness on the MacBook Air which is pretty dim.
All kinds of things just don't work on the M1: Intel didn't just get into the performance of the M1. It also said it found the MacBook Pro had serious shortcomings, such as an inability to use more than one display with a Thunderbolt dock. And while the PC can use gaming headsets, eGPUs, a third-party finger print reader, Wacom Drawing tablet and Xbox Controller, Intel said it found the MacBook Pro simply doesn't work with eGPUs, and had multiple issues with other devices. That's just hardware incompatibility. Intel's rap battle with Apple also highlights issues with plug-ins for Ableton, Bitwig Studio, Avid Pro Tools, FL Studio, Motu and many others.
1.12 X faster is "crushing" (Score:5, Funny)
Am I missing something? Did someone misplace a decimal?
Re:1.12 X faster is "crushing" (Score:5, Insightful)
Re:1.12 X faster is "crushing" (Score:5, Insightful)
And raw performance still isn't the be all and end all of a product's worth
Intel from several months ago [extremetech.com] agrees with you:
We should see this moment [the COVID-19 pandemic] as an opportunity to shift our focus as an industry from benchmarks to the benefits and impacts of the technology we create.
Basically even if Intel wins on benchmarks, Intel is also saying that those wins are not important anyway. ;)
Re: (Score:2, Funny)
To be blunt INTEL crippled by privacy invasive Windows anal probe 10 (we control you, you do not control the computer you gave us for free they control it, they own it). Apple is not, given a choice, the CPU versus the operating system encumbered by it, Apple is taking the lead where it counts in the interface between user and hardware.
Re:1.12 X faster is "crushing" (Score:5, Insightful)
Apple is also taking a lead in privacy, despite the prevailing cult-like fake news you hear on this site. They are in a major court fight with FaceBook about blocking FaceBooks tracking, for one. Why? Because they sell hardware, not ads.
Now someone needs to make up for the loss of ad income the other companies have, which makes their product more expensive.
Apple pro level software is cheap too, and does not have any ads in it. Hardware makes up for that.
The sad reality is that, no-one wants to pay for anything and demand that other companies' ads pay for their own fun.
Privacy has become a luxury item on the Internet.
Re: (Score:3)
That's an area where Google obviously can't compete (since their main source of income is gathering all the data they can on their users and use that to sell ads) and where very likely Android will be behind iOS in the inmediate future (if not forever).
What Apple does was some
Re: (Score:2)
Re: (Score:3, Insightful)
Apple is taking the lead where it counts in the interface between user and hardware.
That, and couple of things more. Apple is truly innovating by blending domains of smartphone and computing. Adding i386 compatibility as a cherry on top.
OTOH, isn't ARM processor production more than competitive, comparing to i7? ARM used to be underdog on par with crappy Atom from Intel, low cost product. With smartphone and computing domains blending, economy of scale does play on hand to Apple in regards of manufacture expense so much. While Apple is brand of desire in both domains, as a matter of fact.
T
Re: (Score:2)
Many years ago there was an article on Slashdot. Someone toyed with Javascript and managed to vaguely simulate the Mac's dock by changing the scale of an icon if the mouse hovered over it.
Slashdot headline? "Is this the beginning of the end for Apple?"
Re:1.12 X faster is "crushing" (Score:5, Informative)
Re: (Score:2)
Re:1.12 X faster is "crushing" (Score:5, Informative)
Am I missing something?
Yes, the article.
If you'd read it, you'd see that that there are a set of tests groups under "content creation" performance. On two of those tests, the Intel CPUs are close to 6x faster than the M1.
1.12x faster is the *least* advantage Intel sees in that batch of tests.
Re: (Score:2)
Am I missing something?
Yes, your missing an ability to read an entire paragraph.
Intel's core strength (Score:3)
Intel has never been as strong at truth in benchmarking as it's been in volume from marketing.
Re: (Score:3)
well duh (Score:2)
I don't know what Apple is trying to sell us here. ARM is not a "high performance" architecture. All these "content creators" are going to hit the ARM wall HARD and find out that, yes, Premiere effects or VST plugins need to crunch a lot of numbers - and the gimmick of a H.264 hardware encoder is moot when you have something as simple as, I don't know, Warp Stabilizer, which puts my 8 core Ryzen to its knees.
Re: (Score:3, Informative)
Re:well duh (Score:5, Funny)
"ARM isn't an architecture, it's an ISA."
I know what you are trying to say, but you do know what ISA stands for right?
Re:well duh (Score:5, Funny)
It's that old 8-bit and 16-bit card connector, right?
Re: (Score:2)
ARM is not an architecture, it's an instruction set architecture? And from wikipedia
Re: (Score:2)
Re: (Score:2)
It was poorly phrased, but I'm pretty sure he was saying "ARM is not a -microarchitecture-, it's an ISA"
Re:well duh (Score:5, Insightful)
Most ARM implementations are not high performance, but that doesn't mean the architecture itself is inherently incapable of it or has some kind of design flaw preventing high performance. The current fastest supercomputer in the world is using ARM processors:
https://www.top500.org/system/... [top500.org]
If anything, ARM is a newer and cleaner architecture than x86 and requires considerably less work to upgrade into a high performance system. Similarly the lower power requirements should make it capable of clocking higher and fitting more cores into the same power budget.
Re: (Score:2)
> If anything, ARM is a newer and cleaner architecture than x86 and requires considerably less work to upgrade into a high performance system.
Yeah current Intel designs are more complicated/sophisticated than current ARM designs. I wouldn't be at surprised if 20 years from now ARM wins on outright performance per core.
So far, Intel has focused on performance (at the cost of complexity) for the last 40 years. That complexity is how we get the various versions of Meltdown and Spectre. ARM designs have fo
Re: (Score:3)
I think "20 years from now" is somewhat pessimistic. The rate that Arm is progressive I'd say "within the next few years" is much more plausible.
Intels complexity *is* the problem. Theres so much legacy cruft in that instruction set and design. Arms still simple enough of an instruction set that coders can know the entire instruction set by memory. This makes it easier to just focus on tightening up performance of specific pain points without worrying that the 250 other variations of that instruction need t
Re: (Score:2)
If anything, ARM is a newer and cleaner architecture than x86
Can't argue with cleaner, but only newer by a few years.
Re: (Score:3)
We've got legacy stretching back to the 8008 reaching market in 1972. ARM dates to 1985, but almost all the yucky legacy was shorn off when 26 bit addressing was removed.
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Beyond things like that however, ISA is fairly meaningless for performance these days. Everyone emulates
Re: (Score:2)
and requires considerably less work to upgrade into a high performance system
I think you may have that backwards. ARM requires a lot of work to upgrade into a high performance system. What it doesn't require is a lot of work to upgrade to an energy efficient system.
I mean case in point right here in the real world: Apple with years of pushing the envelope for ARM performance, and even your supercomputer example which wins exclusively in core count and is beaten by the rest of the top 10 list in terms of performance per core, in some case by a factor of 2.
And further case in point: I
Re: (Score:3)
The current fastest supercomputer in the world is using ARM processors:
That's pretty misleading.
It has 7.6 million cores.
The first x86 machine on the list (AMD part) would have roughly twice the number crunching ability with that many processors, and use only 20% more power.
cost (Score:5, Interesting)
Re: (Score:2, Funny)
And that's why the new M1 Macs are priced lower than their previous Intel counterpart!
What do you mean, "nope"?
Ah dammit, Apple is being greedy again.
Re: (Score:2)
The M1 models are much cheaper than the i5 models of the 13 inch macbook pro. They are priced at the price that i3 models were.
And the i7 that Intel used in the tests is "a tiny bit" more expensive than the i3 in the previous generation in macbooks.
Though to be fair the acer swift and the macbook pro are about the same price with the acer having more memory and storage.
So, yes you get likely better value with the acer. But Apple did definitely give better value with the m1 macbooks than the previous i3 ver
Intel resorting to fig leaves (Score:5, Interesting)
So Intel have found a few benchmarks in which they come out on top. How about an average of all benchmarks? I think they'd be left with little more than a fig leaf.
Re:Intel resorting to fig leaves (Score:5, Insightful)
If Intel spent as much money on engineering as it does in cherry-picking benchmark results, it wouldn't need to cherry pick benchmark results.
Re:Intel resorting to fig leaves (Score:5, Insightful)
So Intel have found a few benchmarks in which they come out on top.
Yeah just using Chrome, Netflix, Photoshop, Premier, or playing / encoding a video. A completely unrealistic workload for a computer. The only metric that matters is Geekbench numbers. /sarcasm
I could totally kill a mouse! (Score:3)
Native or translated code? (Score:2)
With some of these benchmarks, I wonder if they're actually testing native code... or x86 code that's being emulated/translated to run on the M1.
I'm especially curious about this with some of those "Evo" benchmarks involving MS Office, and of course any apps from large vendors likely to drag their feet on recompiling for a new architecture (especially out of cycle with their major version update releases).
Re: (Score:2)
MS have ported office to ARM, so there is a native version.
It is however the first version, so it's likely to be a straight recompile. The x86 version likely has all manner of optimisations which have not yet been carried over to ARM.
Re: (Score:2)
Yep, and since MS have been looking for a while to free themselves from the Intel yoke particualrly with their own Arm based offerings, I highly suspect MS is going to be looking pretty deeply into the problems of optimization to the Arm ecosystem.
Re: (Score:2)
MS have made a version of windows for arm as well as an emulation layer and developer tools capable of building native arm binaries, that could most definitely be seen as a first step towards lessening their dependence on x86.
cherry picking hardware (Score:5, Informative)
Re:cherry picking hardware (Score:5, Insightful)
Funny that they include "no support for boot camp" as a mark against the M1, but don't include "not able to run macos" as a black mark against the non apple laptop.
Also they seem to be comparing several different devices, one of which seems to be an internal whitebox. Intel cpus are often thermally throttled in laptops, especially thin and light laptops like those apple likes to make - if their whitebox has heavy duty cooling them sure it would outperform a thin laptop.
Re: (Score:2)
Re: (Score:3)
The M1 macbooks can run windows for arm too, again the restriction is due to licensing rather than any technical limitation.
There have already been demonstrations running linux on m1 macs.
Re: cherry picking hardware (Score:5, Insightful)
The current M1 Macs, it should be noted, are the entry level, absolutely minimum-spec Macs. Apple still sells Intel laptops and desktops at the higher-end for this generation. I think Intel wants to create the conventional wisdom that M1 is slow/inferior while they have any leg to stand on. The fact that weâ(TM)re even debating about comparing a minimum spec, feather weight laptop that gets two working days of battery life against an i7 is telling.
Re: (Score:3)
There was a pc magazine back in the day that would turn on "SANE" for Mac builds. This library, Standard Apple Numerics Environment, was a specialty library designed to make sure the last digits in floating point math came out the same on Macs and Apple IIs. As it crushed floating point operations by all this extra software overhead, needless to say it made PCs look great, their ads being 50% of the magazine space, purely coincidence I assure you.
In a free society (Score:4)
We should be able to run macOS, Windows and Linux natively on Intel, Mac and Arm systems. But we cannot, to the enduring shame of all.
Re: (Score:2)
Uhh... (Score:4, Insightful)
Holding price constant, the correct comparison is apparently M1 vs. i5-1038NG7, not an i7-1185G7, which is what Intel used.
And, most importantly, test both chips under MacOS.
Re: (Score:2)
"They should compare the M1 Mac to a similarly priced Intel Mac."
Wait, what? Apple controls both the retail price and the component manifest. I'm not sure that letting them dictate what may be compared makes any sense.
It's entirely Apples decision to still be using a 10th generation i5; and to charge 1800 for the product.
Plus Apple is practically famous for long refresh times on models, continuing to use parts that are almost deemed obsolete in the rest of the PC world while continuing to charge top dollar
Re: (Score:2)
Re:Uhh... (Score:4, Funny)
Apples vs. 10 varieties of oranges (Score:5, Insightful)
On the one hand, we have the M1, which is currently only available in Apple’s lowest-end offerings. It’s a first-generation PC-class chip and literally the worst one that Apple ever produce in terms of its raw performance.
On the other end, Intel has had decades to develop an entire range of SKUs that should all be mature, capable products if Intel has been doing its job, yet the best they can do is to liberally cherry-pick whichever SKU happens to be optimized for a particular benchmark so that they can eke out small victories that ring hollow to anyone used to looking over benchmarks.
To say there’s a whiff of desperation here is stating it lightly.
Re: Apples vs. 10 varieties of oranges (Score:5, Insightful)
Someone finally stated the crux issue. M1 is Appleâ(TM)s first chip design. And, that new chip strikes terror in Intel who has been developing processors since the 1970â(TM)s.
I suspect the prospect of the M1s successors scare the crap out of Intel.
It's not their first chip design (Score:3)
Comment removed (Score:5, Informative)
Re: (Score:3)
I suspect the prospect of the M1s successors scare the crap out of Intel.
M1's successors should scare the crap out of anybody.
M5: "This unit must survive."
Re: Apples vs. 10 varieties of oranges (Score:3)
Re:Apples vs. 10 varieties of oranges (Score:5, Informative)
Here is all you need to know [medium.com]
Intel may still be able to barely compete, but if the linear extrapolation holds, it's not going to be for long.
Re: (Score:2)
Well it's either this or perhaps x86/x64 architecture in current fab nodes reached its limits. ARM arch may have some more life in curent nodes and hence comes on top. Slightly. For now, until it also reches the arch/fab limit and the progress stalls again.
Re: (Score:3)
"It’s a first-generation PC-class chip and literally the worst one that Apple ever produce in terms of its raw performance."
Sure, but it benefits from multiple generations of processors that came before it in other applications, including some preceding Apple's acquisition of the design team. To suggest that it's a first effort is a disingenuous as it can possibly be. The M1 is a fully mature design and is also the best performing "chip" Apple has ever made.
Also, what basis are you using to assert t
Re: (Score:3)
I never said it was “a first effort” and I am well aware of Apple’s history in this space, as my comment history will make clear, hence why I deliberately chose my words. The fact that you chose to misread them does not make them disingenuous, nor am I even arguing what you seem to think I’m arguing.
And I said “Apple’s lowest-end offerings”. It’s a statement that is relative to Apple’s other offerings, not to the market as a whole. There’s nothing
Benchmarkets (Score:2)
It's all marketing, ain't it?
Lies, Damn Lies, and Benchmarks
Silly PR and Marketing Departments at Intel (Score:2)
While I do not agree with Intel comparing itself to AMD (and viseversa), I can understand why they do it.
After all, if one of my workloads can run on Intel, it may run on AMD as well (and viseversa), so they are in direct competition.
Even a comparison to Qualcomm's chips used on WinARM laptops has a little bit of sense...
But, comparing themselves to an M1? what for? Is not like people will go to computacenter and buy an in box M1 for their Asus mobo, or buy a Mac to install Windows on ARM...
In spanish we ha
I,ve seen these comparisons from Intel for 40 yrs (Score:5, Insightful)
Seriously - I'm not kidding. A quick Googling and I found this from 1982: http://www.bitsavers.org/compo... [bitsavers.org]
I know there's an Intel created benchmark for the 8086 vs 68k from around 1979 - guess who has more performance according to Intel? I have seen Intel benchmarks for the '286 or '386 vs the 68032, x86 vs AMD, x86 vs Power PC, etc.
I guess creating test cases showing your chip is superior to the competition is cheaper than actually creating a superior chip.
i7 also crushes M1 (Score:5, Funny)
I remember the P4 (Score:2)
"You're benchmarking wrong" (Score:5, Interesting)
I guess Intel would say that everyone else must be benchmarking wrong then because the Linus Tech Tips video review [youtube.com] on the Macs showed they beat the pants off similar "weight class" laptops running Intel and AMD chips. The only competitor that beat the M1s out was the beefy gaming laptop they included for comparison. They even had an XPS 2-in-1 that had a Intel 1165G7, which is only a minor step down compared to the 1185G7 that Intel is talking about in this article and the 65 still lost -- handly.
Additionally, the bit I find a little disingenuous is the battery life comparison. Apple's claims are way off, but its not like Intel has anything to stand on here. The Swift 5 they are comparing against the Air, has a 56Whr battery according to my googling. The Air has a 50Whr battery and the M1 MBP has a 58Whr battery. The benchmarking by LTT showed the M1 MBP could get nearly 20 hours of battery life compared to the XPS's 12hrs. Also it's been shown that the MBP and the Air are effectively the same system minus the thermal management system (ie: fans) between the two.
This slide deck from Intel just tells me they are running scared because Apple may successfully create a professional ARM ecosystem which could really hurt Intel (and AMD).
Re: (Score:3)
I actually think its what Qualcomm (etc) are cooking up that should have Intel worried.
For now M1 is the undisputed king of the Arm world, BUT its only available to Apple. Because Apples competitors want in on that shit, Qualcom and the like are therefore stepping up their own efforts to catch up and THAT is what Intel should be fearing, because already Microsoft has made it clear they think Arm is a perfectly viable future for the PC and Nvidia thinks that too (Although for chipsets they havent really got
Re:"You're benchmarking wrong" (Score:5, Interesting)
They even had an XPS 2-in-1 that had a Intel 1165G7, which is only a minor step down compared to the 1185G7 that Intel is talking about in this article and the 65 still lost -- handly.
They did talk about the 1165G7. They subed it in for the 1185G7 in the battery life benchmarks.
Some might consider changing chips in the middle of benchmarking may be a tad deceptive. Of course, benchmarking a retail laptop against an internal white-box device with no information about it's statistics could be a tad deceptive too. Then again, doing performance benchmarks with 16GB in your white-box and 8GB in the Mac may also be considered a tad deceptive.
Real world tests show M1 faster (Score:5, Insightful)
I don't know what benchmarks Intel is using, but I've seen plenty of real world proof saying the M1 is faster, for real tasks like compilation.
Even more importantly for me, I've seen real people who have bought the M1 say it's significantly faster for working on the kinds of things I work on (development in Xcode and other tools).
Intel can claim up and down the are faster but in the mean time Apple is shipping real world laptops with proven excellent battery life and performance.
I also have to wonder if Intel disabled the performance draining Spectre/Meltdown patches for the tests as well...
I don't need benchmarks, I need to run my job (Score:5, Interesting)
I need a job to run as fast as possible. R + C++ (covid analysis). Normally I run on a supercomputer, but for checking and debugging, i9 9900k is faster (33%) single thread. That's what I've been using till now. Now I got a mac mini m1, and it is again 25% faster. For this job, single thread. And, I have a fan directed at the mini. .
It was a pain getting everything compiled for m1, but worth it!
I don't think Intel is scared of Apple, per se (Score:4, Interesting)
They're more scared of what Apple has demonstrated - a company can create a speedy laptop that's not beholden to Intel or even x86, and consumers will line up to buy it. Yeah, technically AMD was already a non-Intel option - but, for whatever reason, that just never really took off (other than in the small BYO community).
Problem is, we already know Microsoft has also been poking around in this space... so even if the horse hasn't already escaped the barn, it's charging out through the doorway right now. If Microsoft loosens up with its ARM Windows licensing, Intel might find itself suddenly holding a mostly-empty bag.
Re: (Score:2)
Others have pointed out that this is Apple's first chip.
They will likely produce much more powerful processors in the future.
This is frightening for both obvious reasons. Both as you say (what they represent) and also just Apple themselves as a company. They have a sizable chunk of the market, and could grow it if they aren't having to pay Intel for some overpriced POS.
Whiff of desperation (Score:5, Interesting)
There are several things here which seem to be less-than honest.
First, to get it out of the way -- Apple hasn't (to my knowledge) claimed that it has the overall fastest CPU. What they do hype is the performance-per-watt, and the great thermals of their M1 CPUs. Nowhere in any of the above can I see anything about the power use in all but the battery test, nor how much heat is being emitted during the tests.
Next, when comparing CPUs , listing software that doesn't work on the OS the CPU supports as some sort of "negative" is silly. This was supposed to be about performance. I could make the same argument saying that the CPU in the PS4 is superior to the Core i7, because I can play Ghost of Tsushima on the PS4 custom CPU, but it isn't available on the Intel Core i7, so the Intel CPU gets a zero.
Next, they compare a fanless M1 MacBook Air to the Acer Swift 5, and show they get about the same battery life -- completely ignoring the fact that the Acer has a 56Wh battery, whereas the M1 MBA has a 49.9Wh battery. So the Acer has about 12% more battery capacity, and yet only does as well as the M1 MBA. That makes the Acer 12% worse on battery power consumption than the M1 MBA.
I suspect their "AI" tests have been constructed to specifically avoid using the M1's built-in AI cores -- because Intel doesn't have them.
And comparing features of a released product to a beta? That hardly has anything to do with the M1 processor itself. It seems like they can't make up their mind whether they're complying about the M1s raw capabilities, the Mac platform in general, or what third-party developers choose to do outside Apple's control.
So there's a lot of dishonesty in these results. At least compare apples-to-Apples (heh). I doubt these protestations are going to swing much of anybody in either direction -- pretty much everyone who has used an M1 MBA or MBP has been extremely impressed. Anyone who isn't is free to get the machine that is right for them. Intel is just tilting at windmills here, and it reeks of desperation to me.
Yaz
Re: (Score:3)
Next, they compare a fanless M1 MacBook Air to the Acer Swift 5, and show they get about the same battery life
This test also switched to a different Intel chip from the "performance" benchmarks.
Re: (Score:3)
Good point -- they do seem to compare a single M1 process to any processor in their processor family that can beat the M1 on a given test.
I don't know about anyone else, but for a mobile computer at least (and the M1 is a mobile processor) I want something that works well across a variety of workloads, and not have a system with five different Intel processors in it just to ensure I can best the M1 with one of them for a given task :P.
Yaz
What about security? (Score:2)
Battery life (Score:3)
Wait .. what? "Intel pitted a MacBook Air M1 against an Acer Swift 5 ..." OMG Are they fucking stupid? What if the Acer Swift 5 has a much bigger battery than the Macbook Air? Googling it. Aannd yup sure enough, the Macbook has a 50Wh battery and the Swift 5 has 56 Wh battery. That's a 12% bigger battery, while the laptop lasted around the same amount of time. In other words, the M1 CPU is better on power consumption. Intel doesn't make batteries or even sell computers, so I am not sure what a laptop with a huge battery has to do with Intel's CPU.
Re: (Score:3, Interesting)
Wait .. what? What if the Acer Swift 5 has a larger screen than the Macbook Air? Googling it. Aannd yup sure enough, it does.
It also has a dedicated graphics card. It also has other components that are different. It's also actually more expensive on the Acer store than the Macbook Air is on the Apple one.
I mean, it does support your point that it's a flawed comparison but focusing on the battery size? Broadly comparable machines, and even shitty Acer are matching Apple's mean time between plug-ins.
Re: (Score:3)
The Swift also has a bigger screen, and the brightness was measured in 'clicks from the bottom' which is about as meaningful as the numbers on a toaster dial. It's an utterly meaningless comparison. These claims are being made for the sake or marketing to the 'ordinary consumer' who doesn't have the rudimentary secondry-school knowledge of physics needed to understand the definition of a watt. They are not numbers intended for engineers.
Something is fishy here. (Score:2)
Don't talk to me about AI (Score:3)
AI on a non-GPU is not something sane people do unless they're either prototyping some code, or have bizarrely high memory requirements for their models and tons of RAM. Intel for some reason thinks people want to run Torch or Tensorflow on CPUs and wants to benchmark it. Doesn't work that way.
Re: (Score:2)
If they try to select best laptop for running AI I'd agree. But one could use AI programs as tests of general computation capabilities.
Bad Breakup Starts Slander War (Score:2)
Now, like the iPhone/iPad, Apple can steer their own PC destiny without Intel roadmaps.
And then consider that the Personal Computer is no longer a desktop or laptop computer. It's a smartphone. And having all their hardware run on ARM means you can develop apps to work on all of it with minimal re
Missed opportunity. (Score:5, Funny)
What makes M1 better in perf/watt game? (Score:2)
Is it the Arm architecture itself or some smart way apple designed the chip or?
MacBook is too slow in angering consumers? (Score:2)
Intel had how long to perfect laptop CPUs? (Score:2)
I don't care how fast M1 is, let's see M10 vs then current Intel CPU. Although competition may well force Intel to improve, then cool!
My real-world impression (Score:2)
- i7-Macbook = 5min 37 seconds
- M1-Macbook = 1min 57 seconds (!)
Second test - side by side Teams call - ±2hours
- i7-Macbook = battery down to 45% - hot case and cpu
- M1-Macbook = battery down 10% - dead cold
It's not a benchmark. It's just real life.
Not exactly unbiased reporting (Score:2)
For a company that has a history of doing dodgy things like intentionally crippling benchmarks by turning off all possible optimizations for non-Intel CPUs based on /proc/cpuinfo parsing within the compiler, I give exactly 0 weight to anything Intel has to say about anyone else's CPUs. The strength of the M1 isn't the CPU, it's the combination of this with the ML accelerators. If you're not going to at least make an effort to stress test the things it was designed to optimize, you're neither running a fair
Comment removed (Score:5, Informative)
Always wondered (Score:3)
Best ad voor Apple's M1 ever. (Score:3)
This is the best advertisement for the M1 ever. Not sure what they were thinking at Intel.
Surely enough they even include the task 'online homework' where the M1 just outperformed Intel. Guess what people doing 95% of the time at my PC: surfing the web doing things.
I may export a .pdf but i won't do that frequently and i really don't care if that take 0.3, 3 or 5 seconds. Snappiness of the desktop and browser on the other hand is the most important thing to most users.
And then what all other people here on this thread say. So yes Intel, you lost the game already, people will be buying more Mac's and i might get a mini myself.
Re: (Score:3)
Yeah, because they are not dictated to by arrogant nerds that think they are stupid and actually get stuff done. Good enough reason for me
And before you call me stupid, my last CPU I designed myself on an FPGA.
Re: Who cares (Score:5, Insightful)
What is it about Apple hardware helps you "get stuff done" better than non-Apple hardware? Especially regarding anything Apple from before M1? All Apple did was use relatively weak/slow Intel CPUs compared to what you could get in Windows-based machines of the same generation. They had some interesting CPU and GPU SKUs unique to their product lineup (such as that Intel/AMD hybrid chip, lol), but generally-speaking Macs weren't what you bought for top performance. You bought them because you liked the software ecosystem or because you were locked in somehow.
Apple easily could have used better hardware if they wanted to cut into their own margins. They knew they didn't have to compete on raw performance with Windows machines, so they didn't. Look at today's Mac Pro. It's pretty embarassing. Xeon W? When you could get a 64c Threadripper for less?
A Mac Pro from Apple with 28c Xeon W, 192GB of RAM, 1 TB storage (lol), and the stock video card (looks like a Vega20 with 32GB HBM2 - guess it's an Mi60?) is $18,799. That's without Final Cut Pro or Logic Pro pre-installed. Add $498 for those.
For just shy of $9k, you can get a 3990X Threadripper, 256GB of DDR4-3200 CAS/CL 12, a Radeon Pro VII (aww only a rebranded Mi50, shucks), an ASRock TRX40 Creator motherboard, and a 1200W Corsair HX Platinum power supply. Sure there's some other odds and ends that you need (notably monitor), but not $10k worth.
So, honestly: from a "getting things done" perspective, why in God's name would you take the Mac Pro except for the software ecosystem?
Re: Who cares (Score:4, Insightful)