We will likely see them at a regular cadence, though perhaps not every machine every year. Apple may have its roadmap affected by other issues besides the CPU - e.g. miniLED supplies, radios, etc.
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.
Apple Silicon M1 Emulating x86 is Still Faster Than Every Other Mac in Single Core Benchmark
- Thread starter MacRumors
- Start date
- Sort by reaction score
I think it will have something to do with the lack of the 32GB RAM that so few people actually need.
Here's an example of why:
Reference counted memory management is about 5x faster on Apple Silicon (with native code) and 2x faster (while emulating x86). This is an operation modern software does hundreds of thousands of times per second, sometimes millions of times per second. There is no excuse for Intel chips being so slow at this and many other common operations.
Apple clearly has better engineers than Intel.
That is super zippy for entry level machines.
Those scores don’t mean I can load up one of my 50-60 layer After Effects projects though. Even a dedicated ultra fast desktop with an RTX card struggles with this load. Lots of RAM and a discrete GPU are the only choice until SOCs take a much more gigantic leap forward.
Those scores don’t mean I can load up one of my 50-60 layer After Effects projects though. Even a dedicated ultra fast desktop with an RTX card struggles with this load. Lots of RAM and a discrete GPU are the only choice until SOCs take a much more gigantic leap forward.
Based on what I have read in various "deep dives" of the M1, those differences are part of the reason why the M1 - even in emulation - is scoring higher than the Intel CPUs.
...maybe there is a crushed Terminator head and arm deep in an underground Apple lab somewhere....
I believe s/he is referring to the mobile A-series, which are stacked. The M-series are not. As you know, it's all about space and thermal envelops.
He's earned the right to that attitude. You're new here so I won't go off the rails but rest assured, there's more than just iDiots hanging out here.
It’s phenomenal. I’m thinking about the mini but father-in-law needs a new Apple Laptop. I might generously let him be the guinea pig 😆. Let’s see what the power users make of it but these figures are stunning!
Massive reorder buffer: UltraSparc V had that. I know, because I was the original designer of the reorder unit on that chip.
On die memory: there is no on die memory. It’s in the package, but not on the die. This is easy to see from the actual die photographs that have appeared on Ars (I addressed this claim in another thread and posted the picture). There are a number of LPDDR4X channels with off-chip drivers, so you can even see how the die connects to off-die RAM. Here’s the photo: https://images.anandtech.com/doci/16226/M1.png
”width”: what width are you referring to? There is nothing unusual about the execution width. It’s, in fact, identical to that used in, say Athlon-64 and Opteron. (I know, because I owned the integer execution unit for the first of those designs)
Dedicated units: most chips are now designed as SoC with on-chip encryption units, etc. AMD transitioned to that design methodology with Bulldozer. I know, because I left AMD right around when that started happening.
The CPU portion of the chip is very similar to every other CPU I ever designed. The SoC methodology is now a very common methodology.
What’s different here is competence, not some radical difference between M1 and x86 chips.
Does anyone have an educated guess as to when we will see AS iMacs?
Well I was waiting for this. Why on earth would you buy a 2020 Intel Macbook Air or Pro now?
Right NOw the only benefits are boot camp and egpu support or more than one external display support
Sounds like they have been learning a lot all these years building iPhones and iPads. Especially the big iPad Pro where the distinction between laptop and dablet is getting increasingly fuzzy. The competition generally fails here - companies like Samsung don't build computers, just phones and tablets. MS is trying, but they don't have the experience with personal computing as a hardware+software combo other than the Xbox.
Apple has been building the know-how for this since the 80s - full control over hardware and software, high efficiency low power computing focus, using the consumer as a benchmark in portables.
1st quarter 2021
Future Apple Silicon Macs Rumored to Include Redesigned 14-Inch and 16-Inch MacBook Pro, 24-Inch iMac, and Smaller Mac Pro
Apple's transition away from Intel processors in Macs is officially underway with this week's introduction of new 13-inch MacBook Pro,...
www.macrumors.com
Late Q2 or early Q3 next year. Probably when M1x (or whatever they call it) is ready, for MBP 16”, high end MBP 13”, and iMacs. (Supporting more RAM, higher I/O bandwidth for more PCI lanes / TB3 ports, etc)
The 16” MBP is rumored for that time frame, and I assume that’s the hold-up for iMac as well.
Though there are also rumors of early next year.
No, we aren't sure, you are. And you're wrong by the way. No one said Rosetta 2 can't run on multiple cores. The article just emphasises the single core perf because that already exceeds all the Intels so it results in a funnier article. Reading comprehension, people.
There is a multi-core score on the benchmark site. It just happens to be beaten by Intel Macs, because apparently multi-core emulation doesn't nearly scale linearly as in a native run, where multi-core score for a proper multi-threaded executable with tasks that are easy to run in parallel is almost exactly single score times core number. Here the single core perf is 1313, so you'd expect a multi-core score of around 10504, yet in the real world it only scores 5888. So multi-core scalability isn't great, but it does exist.
Apple will work hard either to improve on that, or push ARM ports for the most important apps. Or both.
It doesn't have a fan. Which means it will throttle under heavy loads.
Geekbench is supposed to represent "real-world tasks and applications" and in the real world, most people do not peg their CPU under high load for long periods of time. We click a button, it does some work and gets hot, then it gets a chance to cool down before we click another button.
If your workload involves lengthy workloads, the MacBook Pro will be significantly faster.
The Pro also has a larger/heavier battery. They both should comfortably last a full day under "typical" use, but only the Pro will last all day under heavy use.
Also keep in mind Apple has three MacBook Pro models (low end 13", high end 13", high end 16") and only the first one is available with Apple Silicon. Faster MacBook Pros are coming.
Or not wanting to be an early adopter. I wouldn't go for ASi neither for myself nor my girlfriend right now - for me they are still too small / weak (I need 15"+ screen and 32+GB RAM) and neither of us can afford risking infancy period issues with the hardware, software and third party apps (for example she has to use the entire Office suite, including Outlook and Teams).
You completely misunderstood the question and my response.
The premise was that Rosetta 2 might be converting single-threaded code into multi-threaded code.
I said it is not doing that.
That is very different than whether Rosetta 2 can run multi-threaded code - of course it can. But only if the app was already written to handle multi-threaded code.
“Reading comprehension?”
Last edited:
This is true. On the other hand, Intel MacBooks are so heat constrained due to the abysmal fan design that even constant 60-70% of their peak performance would still get you better real world performance...
The lack of throttling in Geekbench benchmarks doesn't tell us the thermal limitations though, because it's only takes a couple of minutes to run. We will need to see some stress tests that simulate long-running intensive tasks. Cinebench, Handbrake or other video encoding tasks & some gaming benchmarks like Unigen Heaven will be interesting to see. I'm hoping that the MacBook Pro & Mini will be at least 30% faster for such tasks.