Apple M1 Pro Review: Is it really faster than Intel/AMD?

[m4a4]

Honestly, I'd hope they have better performance per watt, considering they are actually commercially using (buying) the smallest process size (5nm) with the best transistor density.

But that advantage can't last for much longer (in a relative sense). And I'm glad this competition is pushing Intel and AMD to do better.

 

[lonetac]

So people who have windows workflows, and Microsoft won't license windows on ARM for Macs. The only surprising point in your comment is that any company would install M1 without testing if they were compatible. The fact does remain that for a lot of software, they are great, but there are still windows workflows out there

Way to just whataboutism. This isnt about windows, this is about Apple Mac M1's and the problems they have.

Prior to M1 everything was fine, the M1 processor is causing a lot of problems. And its not just my company, its many companies out there. The mac's were tested, you failed assumptions are a complete failure. Thats how we discovered the issue and are working with Apple Dev Support on resolving their completely broken system.

If someone wanted windows they would get windows. The truth of the matter is developers prefer developing on mac's and the M1 processor is causing a lot of frustrations for those developers. Thats the point. Apple is breaking one of the largest use cases that people use mac's for.

Who cares if MS wont allow windows to install on a Mac, why would you do that? Your comment is completely non sensical.

 

[TheWatcher]

With big LITTLE architecture and 5nm process it looks like x86 can match Apple silicon. Maybe there won't be any difference if memory is put in to the chip itself.

 

[Puiu]

“comparing performance across the best laptop chips available today, all run under fair and equivalent conditions”. What laptops did you benchmark with? That is important because of cooling/throttling concerns. You only mention laptop chips are these in desktops?

“disgustingly expensive” definitely a term to expect in a fair and unbiased review

Handbrake for Apple Silicon is still beta, so still probably not optimized

“I tried but couldn’t get the MacBook Pro to compile the Windows version”. But did you try to compile the apple silicon version on windows? Should probably have left out this missive

7-Zip any intelligence on being optimized for Apple silicon, I could find nothing.

Gaming?, LOL. And now we compare some non-native games running under Rosetta 2. That was meaningless

“M1 Max is ludicrously expensive.” No offense but this is a really ill-informed statement. Upgrading to the 24 core M1 Max is only $200, and to the $32 core M1 Max is only $400 - hardly ludicrously expensive. I did a price compare of M1 based Macs to similarly configured Dell XPS laptops and the Macs blew them away on value even if you could get equivalent performance, generally you had to max spec the Dells, at ludicrously expensive prices, and they still fell behind and that is before the well publicized thermal throttling issues associated with the Dells.

What you are counting are GPU cores not CPU cores.

And it's not the SOC the problem, it's the storage and RAM. Those upgrades have insane prices.

As for your Dell comparison, on Dell's website, an 13.4inch XPS with 3.5K OLED screen, Intel i7 1195G7, 32GB LPDDR4X 4267MHz, 1TB NVME is on sale for 1770$ ($2,269.99 MSRP). The price is not even in the same ballpark.

FYI gaming is important for many.

 

[nismo91]

Nice review. I don't get all the hype surrounding the new M1 Pro chips and I was right. sure it's fast for an ARM based architecture but that's exactly the only benefit it has. being able to run fast only in select cases. and people called that revolutionary.

don't need a trillion dollar company telling me which software I can run.

 

[BadThad]

Great review, very interesting. Thanks!

 

[Sausagemeat]

This article amused me, it’s incredibly Windows focused and looks at it on a “how good is a MacBook for windows users” sort of point of view. If you currently use windows, it works for your needs and you play games then a Mac probably isn’t for you and you shouldn’t buy one. If however you use macs already and cant get the software you use on windows then the M1 will be all things to you. That being said I thought the Mac performed better in games than I expected it would, I didn’t even know these games were available and I reckon neither did a lot of Mac users! Really, to show how impressive the M1 is you need to show it compared to previous Intel based MacBooks, not Intel based Windows laptops,

The only reason it’s being compared is because the M1 has rightfully impressed the tech world with its performance per watt and some of us PC enthusiast types want to know how good it actually is by comparison to a windows based X86 machine. Its kinda like wanting to know how fast a speedboat goes by comparing it to a car.

Im not trying to disrespect Tim, he’s a journalist writing about the right things. But I do think this comparison misses the point. The real comparison is the software. No user is going to ditch the software they like because faster hardware is available to power the software they currently don’t use. What does a windows machine have to offer people who use Final cut pro for example?

P.S. FL studio was blatantly cherry picked to show windows in good light. There are several better audio editing apps available for free on Mac. Tim, you even mention that the app is broken on Mac, what exactly is the point of testing a broken app?

 

[scavengerspc]

Yeah, but its a Mac.

..and the type of person who uses Apple products tend to be a bunch of insufferable wankers, who you just want to punch after 5mins of being in the same room as them.

I wish I could give 2 likes for that.

I've said it before, but I have no doubt that if Apple put dog crap in a bag, stapled it shut and slapped it with a $150 price tag, it would sell out in seconds. Then we would start seeing posts all over the internet about how it is the BEST bag of dog crap out there.

 

[Verne Arase]

The real question is whether you tested these machines as laptops, or as portable desktops.

Can you count on these performance numbers when sitting in a Starbucks away from the wall, or do they need to be plugged in to achieve their numbers?

 

[Lionvibez]

P.S. FL studio was blatantly cherry picked to show windows in good light. There are several better audio editing apps available for free on Mac. Tim, you even mention that the app is broken on Mac, what exactly is the point of testing a broken app?

lol everyone on the youtube channel is thanking Tim for including FL but of course there is an issue.....

 

[Sausagemeat]

I wish I could give 2 likes for that.

I've said it before, but I have no doubt that if Apple put dog crap in a bag, stapled it shut and slapped it with a $150 price tag, it would sell out in seconds. Then we would start seeing posts all over the internet about how it is the BEST bag of dog crap out there.

Apple haters are offen truly disgusting human beings.

Most of them don’t understand the products and know very little about technology in general.

 

[Sausagemeat]

lol everyone on the youtube channel is saying thanking Tim for including FL but of course there is an issue.....

Yeah Tim himself wrote that the app is broken on Mac. So why test it? Unless you just want to produce a graph showing poor performance on Mac. Also it’s not been optimised for M1 yet, it’s literally being emulated by the M1 and the CPU usage being reported as very low whilst doing so. It’s blatantly chosen because it hasn’t been updated yet.

Besides YouTube comments mean nothing, it’s laughable if you actually think they do.

Anyone using a Mac will use Logic Pro or several other apps that are better than FL. You’d know this if you knew anything about macs..

 

[Lionvibez]

Yeah Tim himself wrote that the app is broken on Mac. So why test it? Unless you just want to produce a graph showing poor performance on Mac. Also it’s not been optimised for M1 yet, it’s literally being emulated by the M1 and the CPU usage being reported as very low whilst doing so. It’s blatantly chosen because it hasn’t been updated yet.

Besides YouTube comments mean nothing, it’s laughable if you actually think they do.

Anyone using a Mac will use Logic Pro or several other apps that are better than FL. You’d know this if you knew anything about macs..

So why don't you go to the youtube page then tell everyone there how their comments mean nothing. Since clearly none of them know how to use macs.

 

[scavengerspc]

Apple haters are offen truly disgusting human beings.

Most of them don’t understand the products and know very little about technology in general.

Talk about not understanding, read my post again. Slower. I don't hate Apple. I very much dislike Apple's drones. They bring absolutely nothing to the table.

Anyone using a Mac will use Logic Pro or several other apps that are better than FL.

Logic Pro is better?

 

[Bullwinkle M]

It will be interesting to see how AMD and Intel's next architectures on the same/similar nodes compare to the M1. But of course Apple will be on 3nm by then for their next design so they'll likely still be more efficient and faster in certain tasks.

So what ?
Apple could be on 0.003nm and my 12 year old Nehalem would still be faster and more efficient in certain tasks

Boot times:
Nehalem boots to Windows XP in 3 seconds flat
Meanwhile, the M1 is still wasting power after trying to boot XP for 3 weeks so far

Result:
A 12 year old Intel Chip is still faster and more efficient than M1 in certain tasks

 

[brucek]

Anyone here who can shed more light on the Chromium compile results? Like, is there something about the silicon that's really good for compiling (if so, what?); or does the Mac have a better compiler; or maybe the level of optimization settings aren't directly comparable; or is there a difference in how much code needs to be compiled (maybe MacOS includes more needed libraries natively?)

I don't know much about any of these but that particular graph seemed different enough that I got to wondering what was behind it.

 

[brucek]

I couldn't agree more.

I don't literally agree but I had to like it anyway since it made me laugh coming immediately after such a long and detailed article.

 

[hahahanoobs]

Not a shocker.
X86 is compatible with so many different systems and software than any RISC chip.
And no matter how fast Apple silicon gets, I'm still not buying it! I'd have to use MacOS.
M1 is a purpose built chip for professionals, and it's damn good at that in natively supported applications.

It's no x86 killer that's for sure.

 

[Irata]

I couldn't find the original poster. but that is an ill-informed statement. Sure you can find a windows craptop (plastic hinges, bad screen, slow SSD, bad thermals) for cheaper, but I challenge any of these windows groupies to find a comparably equipped, and built laptop with comparable performance for substantially lower prices. I compared to Dell XPS using Dell's web site and I got underperforming, thermally throttled PCs with slow SSDs, that were heavier, cost more and did not compete performance wise with the M1s. so who is priced so high?

Sure you can get high end beasts in the windows world, but you pay for it

Maybe not everyone wants / needs such a high end laptop ?

Plus, there are options besides low end crap tops and high end machines, I.e. not just extremes. A Lenovo Yoga 7 with a 5800U for around €1000 actually looks like a very nice system.

If someone is looking at a higher end Microsoft Surface otoh, those look pretty terrible next to a new MacBook Pro.

 

[Nintenboy01]

Anyone here who can shed more light on the Chromium compile results? Like, is there something about the silicon that's really good for compiling (if so, what?); or does the Mac have a better compiler; or maybe the level of optimization settings aren't directly comparable; or is there a difference in how much code needs to be compiled (maybe MacOS includes more needed libraries natively?)

I don't know much about any of these but that particular graph seemed different enough that I got to wondering what was behind it.

maybe might have something to do with being an 8-wide architecture

 

[theruck]

Comparing gigabyte laptop with apple just to have something cheaper with oled and make a pricing point is just ridiculously biased
any technology matched laptop from dell or hp would cost more and weight more

 

[torchyear]

Thanks Tim - this is a really useful comparison, especially with highlighting the applications that are lagging behind with native support. (Photoshop is a surprising one after all this time despite marked as 'native'; I've heard from early adopter pro photographer colleagues running into weird issues that aren't rectified unless using Rosetta.) Although the engineering is impressive (just compare it to the disappointing Surface Pro X), current patchy native application support still makes it a more-or-less v0.99 product for critical uses. Something also not discussed frequently (if at all) is the state of peripheral hardware driver support such as printers, scanners, and audiovisual equipment. The M1 release is reminisicent of the OS9-OSX and PowerPC-Intel transitions which broke a lot of peripherals functionality - some fixed after many months, but many left bricked.

 

[HardReset]

maybe might have something to do with being an 8-wide architecture

Probably more with integrated memory and huge memory bandwidth. It seems clear that Apple has advantage when memory bandwidth really matters. Elsewhere it's pretty meh.

No real surprises here. As expected, Apple is very little ahead x86 CPU's. Considering Apple has pretty big manufacturing node advantage and sacrifices memory expansion by integrating memory on SOC making it HUGE chip, there is nothing revolutionary on Apple's M1 CPU's. Very far from x86 killer like Apple fanboys were touting.

 

[Pastuch]

The M1 can only play 5% of the games on the market... therefore it's useless

 

[TechAgnostic]

MacBook: The Betamax of laptops.
Tiny market share compared to the competition.
Fervent, almost cult-like fan base.
Cherry-picks test results to claim "better performance"
Little to no aftermarket hardware/software support compared to the competition.
Reluctant to license technology - prefers to build everything themselves.
Wildly expensive compared to the competition.