Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Intel Says 10nm Chip Development is On Track
- Thread starter MacRumors
- Start date
- Sort by reaction score
The flip side is looking at how quickly Apple dropped support for PowerPC once they moved to Intel.
I do feel like a move to Apple CPUs is still a few years off at least, so I think this 2018 will be supported for at least 5 years.
You can be sure if even Apple decides to move to in home silicon, it will take a lot of time and resources, and don't forget that you have to convince major software players to move to ARM, a not so easy task. I just updated my MacBook Pro and I'm pretty sure it will server me for several good years...
We’ve already reached diminishing returns. Leakage currents become a bigger problem.
So, correct me if I'm wrong, but AMD's zen chips are already manufactured on a 10nm process, right? AMD doesn't directly own foundries anymore, if I'm recalling this correctly, and since they're clearly not using Intel's foundries, who are they using? TSMC?
Anywho, tangental questions aside, some of the most fun I've ever had as a tech enthusiast has been watching Intel scramble to compete with AMD's latest CPU offerings. While Intel still maintains the raw power advantage, zen closed that gap so quickly and offered so much more value in the process. The fact that getting a mainstream chip from Intel that beats the 2700x in productivity workloads requires more than 50% additional cash... It's great. And it's competition Intel sorely needed to motivate them out of the rut they've been in.
Anywho, tangental questions aside, some of the most fun I've ever had as a tech enthusiast has been watching Intel scramble to compete with AMD's latest CPU offerings. While Intel still maintains the raw power advantage, zen closed that gap so quickly and offered so much more value in the process. The fact that getting a mainstream chip from Intel that beats the 2700x in productivity workloads requires more than 50% additional cash... It's great. And it's competition Intel sorely needed to motivate them out of the rut they've been in.
Know the facts before posting that. You are mixing things...
Considering how slow Apple has been to update Macs over the past several years, I'm kinda wondering how committed they would be to putting the necessary resources into creating pro level desktop and laptop CPUs comparable to Intel.
It's not like they could just slap 4 A12s into a MacBook Pro and call it a day.
Run a complex multi-linked sql analysis using that chip and come back to me. A certain levels not everything is equal, anything with wheels can move you from place A to B.
AMD has very good chips for consumer and small server but playing with big names is another league. The same happens with ARM.
So far Intel still has the 90% o more of the market for some reason...
The thing changing and Intel knows very good, is people modifying the way they interact with computers, iPads, iPhones and smart devices, and here Intel is to much of a gorilla for these uses and ARM a more suited for them. But still those devices need cloud servers, which happens to be the big Intel business.
Last edited:
Aaaaah, that's right. And Global Foundries is basically who they spun their foundry business off into, wasn't it. It's coming back to me.
You are confusing cause and effect. The paltry performance and power improvements by intel over the last several years have given Apple little reason to put in the effort to update macs.
[doublepost=1540244755][/doublepost]
Correct.
Very true, but the other part of the problem is that the improvement gains haven't been significant between releases. That probably plays into it more than anything. Macs are lasting a lot longer because hardware isn't advancing as quickly. If you look at Geekbench multicore scores for the 2017 high end quad core MacBook Pro, it was 15528 vs. 11631 for the 2012 high end quad core MacBook Pro. That's five years difference. If you then compare that 2012 MBP at 11631 to the 2374 score for the high end MacBook Pro from 2007, another five years apart, the difference is staggering. Only in 2018 have we seen another core count increase and the Geekbench jump up by a significant amount to 22538. It's just not worth it to Apple to upgrade for 5% gains every year. Now that's obviously not true for the Mac Pro, but they just royally screwed up with that machine and are redesigning the whole thing from scratch now.
Hardware support lasted until being declared "obsolete" in 2013, 7 years after the last PPC machine. I see that as an emminently reasonable amount of time.
Uh, as I said, Intel still maintains the power advantage, and I fully accept that AMD basically doesn't exist in the server business (as far as I'm aware). But if you're looking at a productivity desktop, or heck, even a gaming desktop (despite the single-core performance deficit) and you're not at least considering AMD's offerings, you're either looking for a very specific type of processing power or you're a little crazy.
What?
Apple shipped MacOS "Tiger" in spring of 2005 - that was the OS containing Rosetta, allowing PowerPC programs to run on Intel hardware. That continued until "Lion" was released in summer of 2011 (6 years), but "Snow Leopard", with Rosetta, wasn't officially moved to unsupported until 2014 (9 years).
You have no clue that the EPYC processors run circles around anything Intel can offer. With EPYC 2 this February it gets even wider in gap.
Agree with you 100%. I am always open to other options. Former Intel eng. but not married with them.
But GF just announced it is ceasing development on anything below 12nm, so pretty sure AMD will be moving to TSMC.
Intel lost their entire advantage in chip fabrication tech with this single process fiasco. They are on the verge of becoming completely irrelevant, IMO. IF they actually manage to stick to their current timeline, they'll be 3 years back; hard to believe.
And you need to know that for using the full CPU power you need the corresponding chipset and motherboard, and the developer modify the program to use the CPU. It's not so easy to make just a very powerful in raw and don't have the infrastructure to take advantage of it.
I don't know anything about EPYC, so I can't really talk about the chip, but in a few years we all will see if it worked or not, depending the market it achieves.
from wikipedia: Initial reception to Epyc was generally positive.[8] Epyc was generally found to outperform Intel CPUs in cases where the cores could work independently, such as in high performance computing and big data applications. Epyc fell behind in database tasks compared to Intel's Xeon parts due to higher cache latency.[8]
Nothing is perfect
Last edited:
This is one of my major problems with an ARM switch. Nobody that has to rely on a decent level of Windows/x86 support will ever trust Apple's method of x86 emulation or how long they'll support it. And corporations simply do not take those kinds of risks in buying equipment. It's not the the availability for Windows 10 on ARM. It's all the x86 pre-existing legacy software. Marketshare will plummet across the board for Macs.
There would certainly be more short term profit for Apple as Apple is never going to get into a price war with Intel PCs regardless. I see certain failure in this.
You really think Windows users will accept that? And I doubt anyone would believe Apple would support x86 that long. I think you made my point.
Intel still uses hundreds of millions dollars in R&I, as most smart companies do, they don't rest sleeping on their laurels. Be sure they are cooking something...