It’s naive to think Apple knows more about chip making than Intel. Very naive To assume Intel is behind. History tells this tale often with one standard outcome.
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.
Anyone else used to be very pro-ARM but now having second thoughts?
- Thread starter Spungoflex
- Start date
- Sort by reaction score
Currently, Intel is undeniably behind. Designing an architecture takes years and years, and Intel is still using an evolution of a 15+ year old design, which means they don’t have anything better ready. We don’t know what they are working on in the background, maybe they have a new grat. thing up their sleeve, but it won’t debut any time soon. So in the next couple of years, Apple has the lead. Intel or someone else might get ahead in 5 years or so however.
A big advantage in know how that Apple has is their vast experience with power management. They have been building hardware fir all kinds of power consumption brackets and they gut really good at optimizing for efficiency. All these years of data on mobile phone, watches etc. will be difficult to overcome by any other chip designer.
They are behind. What's naive to think is that they will stay behind. They have already started working on AI acceleration.
Well on consumer chips Intel ARE behind.
Intel is behind both Apple and AMD right now. For anything below top end server chips anyway.Don't forget Apple is a lot bigger than Intel. They can poach all the talent they want and chip design talent moves around quite a lot as it is.
Dunno about that actually. There are reports of a "culture war" goin on behind the scenes at Intel that's the root of their issues. A lot of old time "blue tag" guys are being replaced with contract "green tag" workers and they're feeling threatened and are retaliating by harassing them. I don't need to explain how having infighting in a company hurts the company as a whole.
Also, that aside, historically Intel's performance advantage was facilitated by their aggressive node shrinks. They had the best manufacturing, and now they don't and have been floundering. I think it's gonna take a monumental effort to get Intel back on track, let alone get them back into top-dog spot.
You still don't name any. Other than a processor that you don't know or have any details about other than the obvious "an iPhone with it is coming out this fall" and "it beats out all predecessor chips".
Nor do you say anything about why this year's iPad Pros had an SoC based on the A12 architecture and not the A13 architecture. It's pretty telling that the best they could put in an iPad Pro release a year and a half afterwards was the same SoC as before with more RAM (on non-1TB models) and a GPU core activated that was turned off. We could've had an A13X in both the iPad Pros and the DTK, yet, we didn't. You don't say anything about that which shows that you honestly haven't thought this all through.
Considering most professionals will not be ready to make the switch to Apple Silicon by the time that first 13" MacBook Pro with Apple Silicon sees the light of day due to the myriad of plug-ins not being Apple Silicon native yet, I'd say that is pretty moot because that Apple Silicon 13" Pro will never beat out the Intel 16" Pro with Rosetta benchmarks. (Furthermore, it's not like there won't be a need for Apple to keep that machine in the lineup for those mobile professionals that still need Intel software that won't run well in Rosetta.)
But even so, your theory is predicated on the notion that Apple has a chip ready to put into a 16" MacBook Pro today, and AGAIN, you have nothing to point at as a credible source to prove that. It's only speculation based on what you think will happen. I do agree that if they have the chip to put in there, it doesn't make sense to not launch it (especially since it is due for a refresh anyway), but where we disagree is that I don't think that they have that chip yet and you, based on nothing other than the age of the predecessor to the current iPad Pro SoC, believe that they do.
Right, if your point is that they're advancing in chip technology, then my response is a giant "no **** Sherlock!". No one is saying that they won't be able to replace every Intel Mac with an Apple Silicon one within the 2-year timeframe. All I'm saying is that you're not getting a 16" MacBook Pro first and you're not getting it in 2020. Considering Big Sur has to come out before any Apple Silicon Macs see the light of day, we're really talking the September to November range (as Apple rarely introduces anything in December).
There's an A14 chip that's rumored to match the A12Z. That's great that they've miniaturized that level of performance. But you haven't shown me anything or even said anything to suggest that we'll get an A14-based chip that beats out an 8-core 9th Gen Intel Core i9 H-series processor handily before 2021. Tim Cook said Apple Silicon Macs are coming out this year. Show me something that actually points to Apple releasing a processor that beats said i9 THIS CALENDAR YEAR, and I'll actually take your claims seriously. I don't think you can though because you haven't yet.
I don't care about Ming Chi Kuo or put as much stock in him as you seem to want to disprove everyone's consideration of what he says. People are most interested in the Mac notebook line right now and how that will go through the transition, therefore, it's logical to put a second note out that talks about those products. He does that stuff all the time.
I get that, but it makes your previous claim that it was first out the gate - something that you use repeatedly to justify that it will be first out the gate this time - WRONG.
It's not a wound on my pride. I don't care what you believe or don't believe. Nor do I care about Ming Chi Kuo other than being a mostly reliable source of indications of what's to come. For some reason I can't seem to get you to understand that. If you're not on the same page as the rest of reality, that's totally fine and doesn't bother me at all...that is until you insist that everyone else is wrong with nothing to prove it. That's when you start to waste my time. But it's quarantine and I'm unemployed, so please, waste as much time on your baseless accusations and theories as you want! I'm game!
You actually did say that once before. And, if you'd like, you can say it a third time. Without a source to point to in order to prove it, it's just a nice theory as to what happened. Furthermore, I'm not sure what your point is with it.
Okay, great, they have DESIGNED the chip to succeed the 8-core Core i9 H-series CPUs on the 16" MacBook Pro. Like you said, it's not popping out of the ground just because it's in the design process. They can, however, pop out of the ground a chip suitable to eclipse the performance of the 2020 13" MacBook Pros today. They've had that technology for two years. You can't make me a claim that they have something as ready to go for the 16" because there's nothing out there to suggest that is even the case other than a chip that's coming out for the iPhone in the fall. If there was, we'd be reading about it right now!
Remember what Palm, Nokia, etc. said about Apple making phones?
I assume Intel will get past its 10nm issues, perhaps by moving up 7nm. But even so, Apple has an advantage in that it can design the CPU and integrated GPU for a specific OS. They don’t need to support 32-bit content, so they don’t include it, for example.
I see a lot of posts from you , all of them are are basically the same thing , you just dont believe Apple can make better CPU`s then Intel , which is fine to believe this until shown otherwise , but it is VERY naive to think that they cant.
The fact is that Intel were the top of the food chain in the CPU design world half a decade ago , they would get the best talent out of universities and get the best talent in the industry , that world is OVER , Apple/Amazon/QC/Nvidia/Google and even AMD pays better then Intel for the same job , not to mention a TON of startups in the industry.
So its not Naive that Apple knows how to do better chips then Intel , they 100% do , because they have the best engineers in the industry , you might think of Apple as a consumer electronic company selling watch bands and Intel as a chip design company , but when you break it down to CPU design teams going head to head , its a no contest today, it really is not close , the only thing Intel has the Apple dont is its own Fab, but that seems like a burden more then an edge in recent years.
To add to my previous reply, "Apple" itself doesn't know anything. But since a lot of Intel CPU people currently work for Apple, it very well might be that Apple indeed knows more about chop making than Intel. For example, Apple's current VP of chip design is Johny Srouji, who was a senior manager at the Intel Design Center that developed the most successful Intel CPU up to date, and who was considered for Intel CEO position last year. I wouldn't be surprised if some of the most talented Intel engineers work at Apple now.
Good point! Unless Intel holds some very specific patents on killer technology that Apple can't use, it's the people who will make Apple Silicon successful or not.
TSMC (the company who makes Apple's SoC chips) is already on the 5nm process, with both 3nm and 2nm processes in development. Meanwhile, Intel has trouble with their 10nm process producing sufficient yields and delayed 7nm until 2022/23. In fact, Intel is behind AMD on the process front as well, because AMD is also using TSMC's 7nm process for its 4th gen Ryzen CPUs. Even Intel has admitted they are lagging behind their competitors (annd this was before they delayed 7nm by 2 years):
Intel Says Process Tech to Lag Competitors Until Late 2021, Will Regain Lead with 5nm
It looks like 2020 and 2021 are going to be long years for Intel.
Intel recently sacked their head of engineering and has allegedly been discussing outsourcing fab to TSMC.
Meanwhile, ARM's head of Chinese operations was also sacked for fraud or something but has refused to leave his post and has hired his own security team and will not allow ARM staff from outside China on the premises.
www.zdnet.com
Meanwhile, ARM's head of Chinese operations was also sacked for fraud or something but has refused to leave his post and has hired his own security team and will not allow ARM staff from outside China on the premises.
Arm’s fired China JV CEO refuses to leave, company reps banned from Arm China premises
The ousted executive will not leave his post, escalating tensions between the UK chip designer and its Chinese counterpart.
www.zdnet.com
My second thoughts have all been in the direction of becoming more and more excited. Learning that Tiger Lake seems to be bringing a 20% performance improvement and Apple still dumped Intel means that Apple Silicon will yield amazing performance improvements, in addition to reduction in noise and increase in battery life. Probably new form factors — thinner (but of course) and probably a number of fanless models. Apple's A-series processors allow the iPad to be very performant without a fan and in a constrained thermal environment. But with a desktop or iMac form factor large heatsinks are possible. Imagine the entire rear panel of an iMac as a passive heatsink for a 45W or 65W Apple Silicon CPU that blows the doors off of Intel's 200W processors.
When was first thought about many years ago, I figured would be more co-processors as opposed to full replacmeant, and we have seen such, ie T1/T2 chip, the afterburner card etc. W1/H1 for headphones.
now thinking that instead of traditional CPU/GPU systems that will see more processor style with dedicated silicon for specific tasks as opposed to general purpose processors.
move to own silicon should make this more practical so I suspect that the actual CPU/GPU be less performance but overall system performance better.
apple never been themselves about the spec but the performance and there own silicon should make it even easier to build high performing efficient systems, as can tailor to what they require.
I use my Mac’s pretty much with Apple software. Ie pages, numbers, iTunes and the replacements. Fcp x for video editing. Handbrake currently though looking at dvdfab as looks like using videotoolbox to accelerate now then. It worried about software compatibility as should be a simple recompile for them.
having said that will keep my mm2018 and hack (iMac 2019 spec) whilst they still work, which will give Apple time to mature the new systems. Gen1 not always the best!
now thinking that instead of traditional CPU/GPU systems that will see more processor style with dedicated silicon for specific tasks as opposed to general purpose processors.
move to own silicon should make this more practical so I suspect that the actual CPU/GPU be less performance but overall system performance better.
apple never been themselves about the spec but the performance and there own silicon should make it even easier to build high performing efficient systems, as can tailor to what they require.
I use my Mac’s pretty much with Apple software. Ie pages, numbers, iTunes and the replacements. Fcp x for video editing. Handbrake currently though looking at dvdfab as looks like using videotoolbox to accelerate now then. It worried about software compatibility as should be a simple recompile for them.
having said that will keep my mm2018 and hack (iMac 2019 spec) whilst they still work, which will give Apple time to mature the new systems. Gen1 not always the best!
What is interesting is that ARM is about to release a new instruction set update (v9). It’s possible that the first Apple Silicon will be based on v8, and that the second will wave will be based on v9. Something similar happened during the Intel switch, as the first models were 32-bit and a few months later 64-bit models were released.
Which is kind of what is happening now. At the same time, there is a finite subset of tasks that benefit from having a dedicated coprocessor. CPUs (as logic controllers with excellent random data access capability) and GPUs (as parallel processors with graphics support) are not going anywhere.
I'm now 1200% more pro ARM.
100% for each of the cores that will ship with a low end ARM.
100% for each of the cores that will ship with a low end ARM.
What's going on in this forum is sometimes way too optimistic, but this is Apple forum so it’s understandable.
IMO, I'm still skeptical of Apple's chip makikng capability for highend desktop market and GPU matching top of the line nVidia offerings.
However, I see plenty of hopes in low to mid range chips and GPU. Hence I'm positive about success of many of Apple's product line utilizing those performance level.
Let's not be too naive but also not too pessimistic about it. We still have plenty of options, and if you have to go x64 again, there are plenty of good PC vendors. I'm not bound to a specific OS for my personal needs as I enjoy having to adjust to different computing environment. I know I'm not a typical person, but atleast to me, what Apple is offering to computing market is diversity. I'm eager to find out what future it will bring! Yet,.. I still don't believe Apple will introduce Xeon and 3080ti level chips
IMO, I'm still skeptical of Apple's chip makikng capability for highend desktop market and GPU matching top of the line nVidia offerings.
However, I see plenty of hopes in low to mid range chips and GPU. Hence I'm positive about success of many of Apple's product line utilizing those performance level.
Let's not be too naive but also not too pessimistic about it. We still have plenty of options, and if you have to go x64 again, there are plenty of good PC vendors. I'm not bound to a specific OS for my personal needs as I enjoy having to adjust to different computing environment. I know I'm not a typical person, but atleast to me, what Apple is offering to computing market is diversity. I'm eager to find out what future it will bring! Yet,.. I still don't believe Apple will introduce Xeon and 3080ti level chips
Last edited:
Pretty much. Intel is getting killed because of they are still mainly relying on 5 year old technology. Last time Intel was put in a corner like this they came out with core 2 duo. One of Apple’s main advantage is access to the latest process tech with TSMC. Even then you’re not going to get 100% IPC improvement.What's going on in this forum is sometimes way too optimistic, but this is Apple forum so it’s understandable.
IMO, I'm still skeptical of Apple's chip makikng capability for highend desktop market and GPU matching top of the line nVidia offerings.
However, I see plenty of hopes in low to mid range chips and GPU. Hence I'm positive about success of many of Apple's product line utilizing those performance level.
Let's not be too naive but also not too pessimistic about it. We still have plenty of options, and if you have to go x64 again, there are plenty of good PC vendors. I'm not bound to a specific OS for my personal needs as I enjoy having to adjust to different computing environment. I know I'm not a typical person, but atleast to me, what Apple is offering to computing market is diversity. I'm eager to find out what future it will bring! Yet,.. I still don't believe Apple will introduce Xeon and 3080ti level chips
Also, most claims of A12 superiority all come out of Geekbench which doesn’t always relate to real world app speed. Once the real Apple silicon Macs come out we’ll be able to compare in real world apps.
To be fair, Geekbench on the mac has never been run on a mac that will have multiple types of engines on a SOC.
People should not think ARM chip = faster ++ (but it will)
People should think about a SOC that does neural stuff, Onboard new type of GPU, video compressor/playback workflows...and such.
Apple SOC = new levels of performance that the PC world has not really experienced saved for things in gaming and 3D like RTX.
AND
throw onto all this all the leaps and bounds of Apple adding tech to iDevices, this will correlate almost directly for the mac in new features, speed, and capabilities.
As long as Apple doesn't return to crap deceptive advertising they used to do in PPC days with using filters in Photoshop I'm fine with it.
The problem with onboard GPU's is that they will always be starved for bandwidth compared to dGPU's. Even tile based rendering will have a hard time getting past that.
I'd be surprised if Apple goes to the very expensive HBM route.