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

Apple Silicon M1 MacBook Pro Earns 7508 Multi-Core Score in Cinebench Benchmark

Rigby said:
The 3 GHz Tiger Lake has slightly higher single-core performance (~1570). The multi-core performance is slower than the M1 at ~6200, but that's with only half the CPU cores. The M1 performance is impressive, but it doesn't really "blow away" Intel at least in this benchmark.

I was under the impression that the M1 Macbook Pro has active cooling, but I may be wrong.
Click to expand...
That’s at 28W. At 15W, Tiger Lake comes in at ~1400 single thread, and that’s still 5W higher than M1’s power envelope. The multithreaded benchmark is more difficult to interpret due to SMT vs MP, but given that both have 8 threads, Tiger Lake’s 6200 versus M1’s 7500 score when Tiger Lake “wins” in single thread indicates it loses by 20% due to extensive throttling. Given Tiger Lake’s 2.8x higher allowed TDP in this comparison that’s a thorough blowout by the M1 in my book. Now, this is currently Intel’s fastest single-thread offering, mind you. The vast majority of ChIPs that Intel currently sells in numbers is slower ST than even a 15W Tiger Lake. Given Intel’s 10 nm problems, Tiger Lake represents only a tiny fraction of both their current portfolio and available inventory, which is why Apple is probably relieved they don’t need Tiger Lake anymore, nor have to competitively source against e.g. Dell right now, in order to make some kickass MacBooks. Plus, they will be free to do whatever they please, design-wise, if they can stick their high-performance mobile chips behind any display without ventilation, and run any of their OSs on them they want... Exciting times!

PS: This is not to say that I am not elated about Zen 3, but that’s for another day.
 
  • Like
Reactions: G5isAlive, NetMage, Krypton Deer and 3 others
januarydrive7 said:
If benchmarks are to be believed, then AMD is far, far behind apple in multicore performance on 15-25W laptop chips. Apple could have gone with less powerful/efficient chips such as R7 4800U, but they chose to do one better.
Click to expand...
R7 4800U gets 10156 in Cinebench R23, as opposed to M1's 7566. Granted that's on 25W, on 15W it would still be over 8000 (according to 15/25W score in Cinebench R20). So how is AMD far far behind Apple in multicore performance in 15-25 laptop chips?
 
  • Disagree
  • Like
Reactions: echodriver and NetMage
Robospungo said:
That’s nice and all. But I’m not blown away like I was after Apple’s presentation and seeing the geekbench results. I’ve gone from “Apple just disrupted the entire industry” to “meh, I guess it’s a good first try.”
Click to expand...
Apple has specifically stated that the M1 is the first step in a FAMILY of SoC's designed for the Mac and that it's targeted at the mainstream of the Mac market: MBA, base model 13" MBP, and base model Mac Mini. Let's wait and see what they have in the pipeline.
 
  • Like
Reactions: NetMage and SuperCachetes
EugW said:
Cinebench is not better than Geekbench. Geekbench is a multi-faceted test that tracks similarly to SPEC. Cinebench only does one single task, which is to measure how well a machine would perform in Cinema4D, so its results are less generalizable than Geekbench’s. The main advantage of Cinebench is that it takes a longer time to run, so it’s easier to see throttling effects.

tl;dr:

Geekbench is a better general all around benchmark than Cinebench, but Cinebench runs longer so it can factor in CPU throttling.
Click to expand...
Geekbench is only good for determining how fast a system can run Geekbench. At least Cinebench aligns with a real world application, not so for Geekbench. Geekbench is almost worthless.
 
  • Like
  • Angry
  • Disagree
Reactions: echodriver, NetMage, Moyapilot and 1 other person
Morgenrot said:
That’s at 28W. At 15W, Tiger Lake comes in at ~1400 single thread, and that’s still 5W higher than M1’s power envelope. The multithreaded benchmark is more difficult to interpret due to SMT vs MP, but given that both have 8 threads, Tiger Lake’s 6200 versus M1’s 7500 score when Tiger Lake “wins” in single thread indicates it loses by 20% due to extensive throttling. Given Tiger Lake’s 2.8x higher allowed TDP in this comparison that’s a thorough blowout by the M1 in my book. Now, this is currently Intel’s fastest single-thread offering, mind you. The vast majority of ChIPs that Intel currently sells in numbers is slower ST than even a 15W Tiger Lake. Given Intel’s 10 nm problems, Tiger Lake represents only a tiny fraction of both their current portfolio and available inventory, which is why Apple is probably relieved they don’t need Tiger Lake anymore, nor have to competitively source against e.g. Dell right now, in order to make some kickass MacBooks. Plus, they will be free to do whatever they please, design-wise, if they can stick their high-performance mobile chips behind any display without ventilation, and run any of their OSs on them they want... Exciting times!

PS: This is not to say that I am not elated about Zen 3, but that’s for another day.
Click to expand...
I think it’s a mistake to assume M1 is equivalent to 10 W in Intel TDP terms.
 
  • Like
Reactions: iPadified
matt_k said:
Yes, when you compare it with what intel has to offer, it's great. But Intel is currently far, far behind AMD in multicore performance on 15-25W laptop chips. So if just being better than intel was the goal, Apple could have gone with R7 4800U, which would mean lot less hassle with porting software and giving up Windows compatibility. For some people this transition is going to be very painful, so I'm really hoping for something that doesn't just beat Intel (that is a dead horse already), but also AMD, with a significant margin to make it worth the hassle.
Click to expand...
No thanks, I don't need Windows nor AMD as most people. You can take a cloud windows pc if you really need it.
AMD is doing nicely, but just recently, too big of a gamble for Apple to go for them. Pretty sure they can do much better then them.
 
djjeff said:
Geekbench is only good for determining how fast a system can run Geekbench. At least Cinebench aligns with a real world application, not so for Geekbench. Geekbench is almost worthless.
Click to expand...
You could say the same thing for spec, yet many regard it as a meaningful test.

You are behind the times. What you say about Geekbench may have been true for Geekbench 2 or Geekbench 3, but Geekbench 5 has a fairly reasonable reputation.

As mentioned, Geekbench 5 tracks spec pretty well.
 
  • Like
Reactions: NetMage and weaztek
macduke said:
Side note: Did Apple actually say M was for Mac in the video? If not, maybe M actually stands for Mobile? The MacBook Pro 16" could have an M1X and maybe the desktops could have D1 on the iMac and D1X on the iMac Pro and Mac Pro? I know the Mac Mini got the M1 but I think it has always used mostly mobile parts.



And it's also burning up our laps. But yeah, don't do it. I'm going to upgrade to the second or third generation of the high performance chips. By then there will be a lot more third party software support, fewer bugs, and even nuttier performance with M2X or M3X running most everything native.
Click to expand...
Given the comments I've read in the forums I think M actually stands for magical.
 
  • Like
Reactions: akidd
EugW said:
You could say the same thing for spec, yet many CPU architects regard it as a meaningful test.

You are behind the times. What you say about Geekbench may have been true for Geekbench 2 or Geekbench 3, but Geekbench 5 has a fairly reasonable reputation.
Click to expand...
I will say the same thing about spec.
 
  • Like
Reactions: raybo
vinegarshots said:
Just to be clear, for anyone who has used Cinebench in the past— this is the new version, so the scores have different values from before.

My 10900k in my PC just benched 14,217 multicore. So... M1 is not the PC crusher yet, in Cinebench at least.
Click to expand...

A $600 CPU requiring 125 WATTS should outpace the M1 by quite a bit more than that, shouldn't it?
 
  • Like
Reactions: 2Stepfan, freedomlinux, NetMage and 12 others
Robospungo said:
People were comparing the M1 to desktop chips last week when geekbench results started to leak. It’s clear now that the M1 can’t hang with desktop processors and Apple has a LONG way to go to be the market leader in performance.

Currently: AMD > Intel > Apple
Click to expand...
But gee golly wize, it's not a desktop chip, it's a low-end laptop chip. But let's face it: You're only here to piss on everyone's parade, so ...
 
  • Like
Reactions: zapm, 2Stepfan, G5isAlive and 4 others
smulji said:
Apple has specifically stated that the M1 is the first step in a FAMILY of SoC's designed for the Mac and that it's targeted at the mainstream of the Mac market: MBA, base model 13" MBP, and base model Mac Mini. Let's wait and see what they have in the pipeline.
Click to expand...
I think this is something that is forgotten in all of the discussion about ARM versus Intel. The M1 is an SoC design whereas Intel processors are not.
 
EPO75 said:
No thanks, I don't need Windows nor AMD as most people. You can take a cloud windows pc if you really need it.
AMD is doing nicely, but juwt recently, too big of a gamble for Apple to go for them. Pretty sure they can do much better then them.
Click to expand...
Well, that's great for you, but many people do need windows. And not just Windows, I have VMs of old macOS versions that I need to test software with. That's also not possible to run on M1 anymore. For some people (including me) this is going to be a very painful transition, and if the CPU is only marginally better than what AMD has to offer (forget intel), that's a difficult pill to swallow. All this said, M1 is likely not meant to complete with R7 4800U, so fingers crossed for M1X.
 
  • Like
Reactions: Nugget
Morgenrot said:
That’s at 28W. At 15W, Tiger Lake comes in at ~1400 single thread, and that’s still 5W higher than M1’s power envelope.
Click to expand...
Do we know for certain what the power envelope is in the Macbook Pro?
Morgenrot said:
The multithreaded benchmark is more difficult to interpret due to SMT vs MP, but given that both have 8 threads, Tiger Lake’s 6200 versus M1’s 7500 score when Tiger Lake “wins” in single thread indicates it loses by 20% due to extensive throttling.
Click to expand...
You can't directly compare 4 cores + hyperthreading to 8 discrete cores. I suspect the 6- and 8-core versions of Tiger Lake will "blow away" the M1 in multi-threaded performance (not that MT is particularly important for an ultraportable).
Morgenrot said:
The vast majority of ChIPs that Intel currently sells in numbers is slower ST than even a 15W Tiger Lake.
Click to expand...
I think it's only fair to compare Apple latest CPU to the latest CPUs of the competitors, rather than to older generations.
Morgenrot said:
Given Intel’s 10 nm problems, Tiger Lake represents only a tiny fraction of both their current portfolio and available inventory
Click to expand...
Tiger Lake laptops are readily available now.
 
tgwaste said:
Maybe im reading it wrong but 7508 seems to be WAY down on the list making that a really bad score.
i5s score higher (according to the site listing other scores).
Click to expand...

On Cinebench23 single core M1s are 40% faster than any I5.
 
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back