https://www.zdnet.com/article/my-biggest-apple-m1-question-whats-intel-been-doing-all-these-years/?ftag=TRE-03-10aaa6b&bhid=20563615010054021886152894255900&mid=13172177&cid=717419361
 
Apple came out with a new processor for its Mac line that hands down beat the equivalent Intel chip both in terms of performance and power consumption. What's Intel been doing all these years?
 
I think that it's safe to call Apple's new M1 chip "disruptive." After all, here is a first-generation product that's kicked sand into the faces of huge companies who do pretty much nothing other than make chips. I'm looking here at Intel, but it equally applies to the likes of AMD, and even Qualcomm. How did this happen? Let's start with the most obvious.
 
As big as Intel and the other chip giants are, they're not as big as Apple. Apple has a lot of spare cash to throw at R&D with a very focused goal -- a single chip to begin horning Intel aside. That money, along with the focus of manufacturing a single chip, also allowed Apple to write blank checks to TSMC to make 5-nanometer possible. Being in a position to be able to bankroll -- and make it make business sense (more on this in a moment) -- 5-nanometer architecture was a cornerstone to making M1 possible.
 
Speaking of TSMC, the company has enjoyed a close relationship with Apple since Samsung was sidelined as iPhone chipmaker back in 2014 with the iPhone 6's A8 chip. The company has a proven track record of producing tens of millions of chips every year. Being able to guarantee that sort of volume, adjusting to fluctuations in demand, and hitting the mark in terms of yearly releases, are all things that make Apple's job easier. 
 
Then there's the fact that Apple and Intel are in totally different businesses. Intel is primarily making chips for PC OEMs, and that's hardly a high-margin business. Apple, on the other hand, is making products that command high profit margins. Intel needs to satisfy an industry that doesn't just demand cheap components, but isn't in a position to pay much more. All this means that Intel's focus has been on small, incremental improvements. Same could be said of the likes of AMD's PC customers and Qualcomm's Android handset makers. The money just wasn't there to make such performance leaps worthwhile.
 
I've read a lot about internal battles between Apple and Intel over the years, and while there's no doubt that a lot has gone on between the two companies since macs shifted to Intel chips, a move this big and this complex and this expensive isn't going to have politics in the front seat. The writing has been on the wall regarding this shift for many years, and the bigger that Apple became, and more complex and timely its supply chain became, the more it made sense. And now it's happened, there's no stopping it. Intel's days in Macs are numbered. As to what this means to the wider PC industry, it's hard to tell. I've no doubt that OEMs would love chips that gave them the power that the M1 gives Macs, but whether they could afford them, or whether customers would pay for them… well, I doubt it given the trajectory of the market. But to answer the question, what has Intel been doing all these years, well, it's been making chips for an industry that's primarily price sensitive, and where tiny incremental improvements have been enough to keep everyone -- from OEMs to the end users -- happy. It's not that Intel couldn't have made a chip like the M1, it's just that we probably wouldn't have paid for it.
 
It is very interesting indeed in my opinion that Apple has made a superior CPU to Intel on it's first attempt. 
 
 

Apple hardware is 3X the price of equivalent PC hardware ...
 
Good to see Apple put their profits into R&D

It does beg that question, yes... Skylake, Coffee Lake, Rocket Lake, Alder Lake... all of these designs are still water under the Sandy Bridge.  Intel kept the same core design from 9-year-old Sandy Bridge and simply tweaked it to crank the clocks up to 5.3Ghz. The IPC jump from Zen 2 to Zen 3 alone was the equivalent of the last 9 years on Intel's side if you ignore the cranked clockspeed.
 
I wonder if Apple will have to add additional logic to its M1 design to enable x86 virtualization, it seems like one feature they can't just ignore and will need to add to M2 if they want to continue to push into the laptop/PC markets. 

kougar
I wonder if Apple will have to add additional logic to its M1 design to enable x86 virtualization, it seems like one feature they can't just ignore and will need to add to M2 if they want to continue to push into the laptop/PC markets. 

I don't see them adding anything for x86 virtualization.  Honestly typical Apple mentality, get with the lastest or get left behind is what will allow it to keep penetrating into that market alone.  People want their apps to be on the latest and greatest for apple always and will move those apps towards the new architecture, besides currently they basically VM to run current mac apps, while not perfect it's a start.  Naturally, I think other companies will start moving towards ARM architecture as well, and seeing as windows 10 already supports I wouldn't be surprised to see the likes of at least business-oriented laptops moving that way before long with apple having paved the way.
 
PC gaming though that will be another story, i don't know how that long play will work out but can't imagine it'd change much.  I'm not an Apple fan by any means but having something like the M1 even in a laptop sounds lit to me.

axtoxic
Nobody knows 14nm better than Intel

In related news, nobody knows Crayons better than kindergarteners.
 
If people remember how Apple handled PowerPC -> x86 you'll probably want to unload your x86 Macs before they are truly forced obsolescence. Product life cycles at Apple will begin to look a lot more like buying a cellphone every 2 years but for workstations. It's the Apple way.

Any benchmarks compared to amd and intel?

kougar

veganfanatic
Apple has a tool called Rosetta 2 which allows x86 software to run on the M1 CPU
 
Apple did this when they changed from the PowerPC to Intel a while ago too

 
Rosetta 2 recompiles software during install and then it's done. That is not virtualization. 
 

moose517I don't see them adding anything for x86 virtualization.  Honestly typical Apple mentality, get with the lastest or get left behind is what will allow it to keep penetrating into that market alone.  People want their apps to be on the latest and greatest for apple always and will move those apps towards the new architecture, besides currently they basically VM to run current mac apps, while not perfect it's a start.  Naturally, I think other companies will start moving towards ARM architecture as well, and seeing as windows 10 already supports I wouldn't be surprised to see the likes of at least business-oriented laptops moving that way before long with apple having paved the way.
 
PC gaming though that will be another story, i don't know how that long play will work out but can't imagine it'd change much.  I'm not an Apple fan by any means but having something like the M1 even in a laptop sounds lit to me.

That's the kicker... most tech professionals and engineers that I know of that DO own Macbooks regularly either use Bootcamp or run virtualization stuff within OSX. Apple can't afford to just give them the cold shoulder without harming adoption rates, and engineers can't afford to just wait a few years for the software to work itself out. Apple's own worst enemy may be itself unless it implements a bootcamp/virtualization option for those people. 

Most tech company give engineers macbooks as standard.

Intel has certainly perfected 14nm. Now the trouble is moving in to 7

neodinardo
Why Apple doesn’t do gaming consoles I’ll never know. They could be real competition to Nintendo, Sony and Microsoft. And now they have custom chips to do it with.

Probably they know that there is no room for another player and it won't be a financial success.
 
Happy to see that somebody is trying to be innovative after Intel being so "lazy" for the last 10 years. I don't think I will ever switch from Windows to Apple mainly because of lack of gaming part but only for work/movies/internet browsing why not.

Because building a console requires also having not only hundreds of games, but also the big name game franchises that people want. Buying a handful of studios would be uber expensive and still not be enough. Just because Apple built a console doesn't mean game studios are going to suddenly spend 1/3rd again the development costs to design the game to work on Apple's platform either, particularly when Apple's implementation is ARM-based while the Xbox and PS5 are both running AMD hardware.