As someone who spent many years competing against them: no it’s not.
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 Announces 11th-Generation Tiger Lake Chips as Apple Plans Transition to Arm-Based Apple Silicon
- Thread starter MacRumors
- Start date
- Sort by reaction score
Because there isn't an 11th generation Intel processor yet designed for desktop systems. That won't come out till next year. By that time iMacs will more than likely use Apple Silicon
Factor of 1.6 is nothing when you consider the difference in power consumption. Unfortunately, this is not the only limitation. Anandtech data suggest that A12/A13 simply can’t go beyond 2.7 ghz - the architecture is most likely unable to sustain high clocks. This is something Apple needs to solve it they are serious about desktop.
They could use them in the 13" MBP, which currently uses a 28W TDP CPU. I agree that it's not likely, but I'd buy it in a second if I'm wrong. Intel has also said that they will make other Tiger Lake configurations in the future, which will probably include a 6-core version.
You're not supposed to. These are just internal project names. The official name is "11th gen Core", which should be clear enough.
No, switching won't happen. Marketshares for PCs will start to decline soon, because of chrome and apple. It doesn't hurt that Intel's line up is completely incomprehensible.
Professional software is moving to services to that they can charge more. More processing will occur in the cloud, which helps all companies lock in customers. Consumer software is already there.
Performance? I though these were places intel exec's vacationed...
{Waits for next scandal to find out if my joke is true... }
Many here are calling the death of intel, I wouldn't underestimate this chips value, most consumers and enterprises will buy this chip variant than the Apple Silicon. Why? It will be cheaper and do everything most people need. Also, the graphics performance of the Xe is pretty damn good and early indications show it's faster than the new AMD 4000 APU. I will probably end up buying the Apple Silicon, but I'll give this a hard look.
Last edited:
No, it's not necessarily better to outsource technology.
If you have ever worked in the industry then you know your supplies are constrained by wafer availability and supply chain issues.
Intel has had issues with 10nm, but their 10nm falls squarely between TSMC 7nm and 5nm nodes.
The names mean nothing when comparing across processes at different silicon vendors.
Intel is one company with multiple business units that are as big or bigger than quite a few companies in the semiconductor industry.
TSMC and Samsung don't have the same process tech, so for Apple to second source means a new design from the perspective of cache/memory integration, I/O cells, PLLs, layout, etc. It's not a slam dunk.
So they would need to second source now, since trying to bring a second source for silicon including qualification on line, would be a minimum of 8 months to a year. This assumes that everything would be available in the other process technology. It's a effectively the same design but it is a different chip, with different layout, different mask set and different qualification.
Once again, to prevent a serious fab related delay, they need a parallel path and second source when they go to market.
If they have yield or supply chain problems they need a backup.
This is one reason why they typically don't adopt the latest processor from Intel.
A brand new processor comes with brand new problems and hiccups.
Says who?
They could offer MPX cards containing AS systems-on-a-chip (system-on-a-chips ?) providing a mixture of additional GPU cores and extra processor cores with shared RAM (which seems to be the new sexy for them).
...at that level, GPUs get less about getting a picture on the screen and more about GPU-based computing. Even with existing Xeons and dGPU the raw per-core performance maxes out long before you get to the top-end products, after which it's all about loading on more cores and relying on multi-threaded workloads.
...the Intel Mac Pro only came out at the end of last year, so Apple will have to support it for the next 4 years or so. That gives them a chance to try something more radical "in parallel" if they really want to think different. On the other hand, the last two Mac Pros have been abandoned for years at a time so maybe they'll just forget about the Mac pro for another 5 years. Or have another bash at replacing the Mac Pro with a sealed unit (as long as they don't make it trashcan-shaped it'll be fine, right?)
Glad we are on the same page here.
That's ok if you don't see it. I did explain it above. It might be nothing but it's very suspect that this was announced now.
Thanks for sharing this fact. Someone here might not know this.
What I said doesn't change things. Intel always announce their roadmaps a long time in advance. A roadmap is different to actual specific product announcements.
False. The broader picture does involve Apple be it working with Apple (as in the past) or as competitors (the future).
Also the conspiracy theorists could say "if it had nothing to do with Apple, it would not be on the MacRumours front page. I don't believe that but some people do.
It has similar design rules to TSMC 7. Not “between” TSMC 7 and 5 That i can see.
I assume a Tiger Lake-H will be launched sometime next year for the MBP? I'd like to buy the last Intel-based MBP so I can still run Windows for a few more years. Hopefully by then MS will have a fully-featured version of Windows 10 that supports ARM.
What a sad end to our debate. Did you run out of facts to continue it on?
Also you believe what you said too. I didn't just say "I mean… OK, if you wanna believe that". I dignified your response with a proper responce of my own. Well thanks for the debate anyway.
Mac Pro exists in alternate universe where time works differently. It gets an update every half decade or so, whether it needs it or not. OTOH you can make PCIe6 work by using 3 legos, a paperclip, and a bent fork + add turbocharger and $700 wheels.
This is some Intel CPU thread, there it is, very exciting.
I’m assuming you mean for 16-inch; the 13-inch basically has appropriate new CPUs in this very thread.
A Tiger Lake-H is probably coming, but a Rocket Lake-H as well. Rocket Lake is 14nm, but with many of Tiger Lake's improvements (such as PCIe 4) backported. Intel will probably offer Tiger Lake-H CPUs with up to six cores and Xe graphics, and Rocket Lake-H CPUs with up to eight (maybe ten) cores, but worse graphics. Given that the 16-inch has discrete graphics anyway, that seems like the better tradeoff to me, so unless Apple wants to again offer a 16-inch without dGPU (like they have pre-2016), I'm guessing they'll go Rocket Lake-H — if they do do another Intel, that is.
I think you're very mistaken if you think professionals want to switch to services and cloud and ARM based Operating Systems with less functionality. I think what we've seen if anything is people have realised they do not own their content on services.... so they want to keep everything local now.
[automerge]1599208337[/automerge]
Ok show me a benchmark of the iPhone 12 beating Threadripper?
Basically what you've said is they can compete on the low to mid end.... well that won't cut it for professionals who are all ultra high end.
iPhone 12 isn't even out yet. iPhone 11 scores 5% faster at single-threaded tasks, despite 41% lower clock.
In single-threaded performance? Here you go:
AMD’s New EPYC 7F52 Reviewed: The F is for ᴴᴵᴳᴴ Frequency
AnandTech Forums: Technology, Hardware, Software, and Deals
Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
www.anandtech.com
Note that the entire iPhone draws just around 5 watts during these tests
What do you mean with ultra high end exactly? Apple's architecture currently has the potential of reaching highest per-core peak performance on the market, and they can reach very good sustained per-core performance with ultra-low power consumption. Which means this architecture can in principle serve both the needs of burst workloads (few high-priority threads) and sustained workloads (many long-running threads).
The real question here is scalability. In order to provide competitive sustained performance to the likes of EPYC or Xeon, they will need more cores. Assuming they have the technology (which is far from being certain), packing together 32 A13-equivalent cores running, say, at 2.2+ ghz, they will outpace anything Intel or AMD can put out, at a much lower power consumption level. At any rate, even if Apple is working on a "big" CPU like this (which again is not clear), we won't see it until 2022 at earliest.
So what are the chances of this featuring in a newer 13 inch MBPro?
I’d quite like to get that to ensure intel compatibility for as long as possible during the Apple CPU transition.
I’d quite like to get that to ensure intel compatibility for as long as possible during the Apple CPU transition.