Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
RIP Intel/AMD.

GG WP Apple.

I don't think there are ANY x86 processors that match this single-core speed.
I think you forgot the performance of Zen 3. Single core is on par with the M1, while the multi core is even better on the 6 core chip(5600x). Not to mention the 8 core chips and above. These Zen 3 chips are still on 7nm, while the M1 is on 5nm, so it is an unfair comparison. Next year Amd will also be on 5nm with at least further 20% improvement or more. These Zen 3 chips will also be in Laptops next year, so Apple is not ahead in performance. The Zen 3 chips also have incredible performance per watt at 10-15w.
 
That catchup will happen after 5 years. It is not like Intel will come up with a chip design that will match Apple’s in the next 18 months.

Apple’s iPhone chips have been smoking Qualcomm’s since iPhone 7 and the A13 is still the second fastest mass produced chip after the A14. Where are those magic chips? It has been 6 years and they have not caught up.
That catchup won't happen in 5 years, do you think Apple will be standing still in the meantime? Intel would have to leapfrog them in the next 5 years and that isn't going to be happening.
 
Definitely interested in Apple Silicon as soon as its integrated graphics handily beat the discreet graphics of the AMD Radeon Pro 5600M with 8GB of HBM2 memory.
Amd laptop gpus are rather bad. Nvidia has much better performance per watt than the 5600m. Compare it to them.
 
That catchup won't happen in 5 years, do you think Apple will be standing still in the meantime? Intel would have to leapfrog them in the next 5 years and that isn't going to be happening.

Intel has the chip designs from Jim Keller in final staging now. The same Jim Keller who led the designs of AMD's Zen Project. The fruits of that labor will begin in Fall 2021. So yes, if you think ARM architecture designed for extreme low power platforms will suddenly rule the world you're delusional.

And then the mature Zen architecture [same is being augmented for Intel] in Zen 4 will have their own AI Engine FPGA [it's in final design stages now] and much more. Apple isn't doing anything the big boys currently aren't already doing. They just wanted to do their own thing completely in-house.
 
As the general computing market shifts to a majority of SoCs in the next 5 years, either AMD and Intel recognize the shift and adjust accordingly or they fall further behind. I think AMD can make the adjustment. Intel is simply going to try and stall it by buying their way through it bribing Microsoft to not embrace Arm until they manage to get something out the door that marketing can trumpet as the next generation of SoC. Truly Intel has nothing new or innovative up their sleeves and hasn’t in at least a decade, if not longer.

The reactions here and by Linus Tech Tips, et al. are understandable given the seismic shift that has been happening over the past 5 years.
 
  • Like
Reactions: joeguide4
As the general computing market shifts to a majority of SoCs in the next 5 years, either AMD and Intel recognize the shift and adjust accordingly or they fall further behind. I think AMD can make the adjustment. Intel is simply going to try and stall it by buying their way through it bribing Microsoft to not embrace Arm until they manage to get something out the door that marketing can trumpet as the next generation of SoC. Truly Intel has nothing new or innovative up their sleeves and hasn’t in at least a decade, if not longer.
They got complacent relying on enterprise sales.
 
  • Like
Reactions: Zdigital2015
You know,
regardless of the eventual outcome, I have to believe Apple Silicone will be a win for consumers in many ways.
(munis repair/ upgrade ability)

Apple is by no stretch one of Intels biggest customers in terms of sales volumes. But Apple does capture a fair amount of mindshare and perceived value. Assuming Apple Silicon is successful then I have to believe board room discussions at Dell/Lenovo/HP/etc. could go something like. “If Apple can save $X dollars a year by having their own chip across Y number of machines being sold, how much money could we be saving from our Y * Z number of machines we sell???”

I’m not saying they would come up with solutions as great as Apple’s, but for most business use cases they wouldn’t need to. This could have much larger downstream affects.

This could give more (perceived) competition in the CPU space that we haven’t seen in decades (90s) ... if ever.

I have to hand it to Intel when they came out with the Core 2 Duo and the i-Series processors they came out roaring. And AMD has been doing great work the last while.
It will be interesting to see if they can get pull it out like that again, to stay competitive in the new market that’s coming.

side note: through a discreet GPU with some Apple Silicon and it looks like we might have a nice Gaming console contender in the Apple TV space :)
 
Intel has the chip designs from Jim Keller in final staging now. The same Jim Keller who led the designs of AMD's Zen Project. The fruits of that labor will begin in Fall 2021. So yes, if you think ARM architecture designed for extreme low power platforms will suddenly rule the world you're delusional.

And then the mature Zen architecture [same is being augmented for Intel] in Zen 4 will have their own AI Engine FPGA [it's in final design stages now] and much more. Apple isn't doing anything the big boys currently aren't already doing. They just wanted to do their own thing completely in-house.

Jim Keller didn’t do what you think he did.
 
If Apple had maintained the OpenGL API properly, things would be much easier – even if it just meant to use it as a translation layer. But noooo. Apple has a "not-invented-here syndrome".

If they're not careful, this can play out like in the 1990s, in which Apple CPUs were much faster than Intel's, but lost due to sheer inertia.

In the 90’s Apple couldn’t design and make their own processors. They can today. And in the 90’s Windows was THE dominant operating system. In business desktops it still is. But the market has fragmented and in phones and pads Apple drives sales, and with increasing tie ins between all of those devices Apple has a strong lead in 2 of them, iPhones and iPads, and the M1 is their start to make computers. With the PowerPC Apple couldn’t dictate what was done. There were 2 other companies that could disagree, IBM and Motorola. And at the time both of those companies had chip design and fabrication experience that Apple didn’t. Apple has acquired that design experience now.
 
I wonder what the benchmark would be if it was run on Rosetta.

You don’t run code “on Rosetta.” It’s a translator that does a one-time translation when an app is installed or first run, so that when you run the code it runs natively, not via an emulator.

It won’t be fully optimized, so I’d expect maybe a 10 percent penalty or so, depending on the code.

Caveat: certain types of code do have to be translated on-the-fly due to peculiarities in the x86 memory model, but it’s comparatively rare.
 
I think you forgot the performance of Zen 3. Single core is on par with the M1, while the multi core is even better on the 6 core chip(5600x). Not to mention the 8 core chips and above. These Zen 3 chips are still on 7nm, while the M1 is on 5nm, so it is an unfair comparison. Next year Amd will also be on 5nm with at least further 20% improvement or more. These Zen 3 chips will also be in Laptops next year, so Apple is not ahead in performance. The Zen 3 chips also have incredible performance per watt at 10-15w.
If it is any indication of the progress Apple made on each A series CPUs, the M series will probably be on the same trajectory. I suspect the M series CPUs for the iMac and 16” MacBook Pro to be a 12-core CPU with up to 64GB of unified RAM.
 
If Apple had maintained the OpenGL API properly, things would be much easier – even if it just meant to use it as a translation layer. But noooo. Apple has a "not-invented-here syndrome".

If they're not careful, this can play out like in the 1990s, in which Apple CPUs were much faster than Intel's, but lost due to sheer inertia.
The failure of the PowerPC experiment is a interesting business case study. However this time Apple is taking advantage of a successful existing ecosystem in ARM including manufacturing contractors with a huge economies of scale, TSMC and Samsung if needed, both are now able to beat or match Intel.

AMD has also skyrocketed when they ditched what used to be their own fab in GlobalFoundries and went with TSMC, and obviously nVidia, AMD-GPU, and Qualcomm all have been using TSMC for a while with Samsung picking up some contracts here and there.

It's actually a bit scary how much the tech world is relying on TSMC. Samsung is still investing billions of dollars into their chip business though and TSMC isn't slowing down in their spending.

Jim Keller didn’t do what you think he did.
Seeing his track record, maybe he's an exceptional recruiter and management of chip designers and has a good sense in where the industry is going? The secret of business-engineer managers are always fascinating. Then again as I saw someone seriously claim Apple's purchase of PA Semi was doomed when Dobberpuhl left 10 years ago. Maybe money is the key and managers are overrated.
 
You know,
regardless of the eventual outcome, I have to believe Apple Silicone will be a win for consumers in many ways.
(munis repair/ upgrade ability)

Apple is by no stretch one of Intels biggest customers in terms of sales volumes. But Apple does capture a fair amount of mindshare and perceived value. Assuming Apple Silicon is successful then I have to believe board room discussions at Dell/Lenovo/HP/etc. could go something like. “If Apple can save $X dollars a year by having their own chip across Y number of machines being sold, how much money could we be saving from our Y * Z number of machines we sell???”

I’m not saying they would come up with solutions as great as Apple’s, but for most business use cases they wouldn’t need to. This could have much larger downstream affects.

This could give more (perceived) competition in the CPU space that we haven’t seen in decades (90s) ... if ever.

I have to hand it to Intel when they came out with the Core 2 Duo and the i-Series processors they came out roaring. And AMD has been doing great work the last while.
It will be interesting to see if they can get pull it out like that again, to stay competitive in the new market that’s coming.

side note: through a discreet GPU with some Apple Silicon and it looks like we might have a nice Gaming console contender in the Apple TV space :)
Since Intel just announced that their next gen desktop architecture is basically tiger lake BACKPORTED to 14nm++++++++ 5+ year old tech that is going to be released next spring means their days of misery are far from over. To add salt in the wound, they couldn't even keep the same core count as 10th gen and are cutting it back down to 8 cores from 10 on their top line model...

Tiger Lake on laptops is also stuck at 4 cores for the foreseeable future at 10nm.
 
Last edited:
  • Wow
Reactions: Unregistered 4U
can someone explain the benefit of buying a MacBook pro, if MacBook air has the same chip which means same cpu/gpu etc.. and is much lighter and nice looking. Why would anyone pay more and buy a heavier MacBook pro? just for the extra battery life?
 
The failure of the PowerPC experiment is a interesting business case study. However this time Apple is taking advantage of a successful existing ecosystem in ARM including manufacturing contractors with a huge economies of scale, TSMC and Samsung if needed, both are now able to beat or match Intel.

AMD has also skyrocketed when they ditched what used to be their own fab in GlobalFoundries and went with TSMC, and obviously nVidia, AMD-GPU, and Qualcomm all have been using TSMC for a while with Samsung picking up some contracts here and there.

It's actually a bit scary how much the tech world is relying on TSMC. Samsung is still investing billions of dollars into their chip business though and TSMC isn't slowing down in their spending.


Seeing his track record, maybe he's an exceptional recruiter and management of chip designers and has a good sense in where the industry is going? The secret of business-engineer managers are always fascinating. Then again as I saw someone seriously claim Apple's purchase of PA Semi was doomed when Dobberpuhl left 10 years ago. Maybe money is the key and managers are overrated.

What’s his track record, really? Has he stayed anyplace long enough to actually have an impact? Maybe. I don’t know. But for something like Zen you are looking at probably three years of work. I worked with him at AMD in his prior stint. He was the architect in charge of the hyper transport bus. He was a smart guy and knew more about circuit design and physical design than our other architects. But then I see press reports saying he architected the whole chip, which was clearly not true (Fred Weber was the driving force).
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.