M1 destroys Intel 10th gen + Nvidia RTX 3080 on Premiere Pro

[iBug2]

To be honest, I wasn't expecting this, considering Premiere is heavily Cuda optimized, but even the beta version of Premiere running faster on M1 than on Intel 10th gen + 3080.

Head to Head: Apple Final Cut Pro vs Adobe Premiere Pro

We pit Adobe's Premiere Pro against Apple's Final Cut Pro to see which video editor takes the performance crown. Same timeline, same footage, same effects... very different results.

www.dpreview.com

 

[cmaier]

But how does it do on random chess benchmarks?

 

[dogslobber]

But Intel has a bunch of benchmarks that show Intel wipes the floor with the M1 in those particular ones. Intel Marketing says those are the only ones that matter. So unless M1 beats Intel in those then it'll always be flawed.

 

[leman]

It probably depends on how much processing is going on, I’d expect the Nvidia GPU to get ahead on workloads that have many processing effects applied. But for simpler stuff it’s not surprising that M1 is faster - memory transfer between the CPU and the GPU is going to be a massive bottleneck of that dGPU. Not much point in having all that compute power if you can’t feed it fast enough.

 

[mi7chy]

They avoid Davinci Resolve with dGPU because it's faster than both Premiere Pro and FCP.

 

[quarkysg]

I would have expected the Razer to come out ahead for the Premier Pro shootout tho. On paper, the Razer is supposedly better in every way compared to the iMac, espcially in the GPU department. Goes to show that all those power are really hobbled by the PCIe bus with the back and fro copying of data.

The M1 iMac is able to push data to the GPU twice as fast as the Razer, and this delta will get wider with faster RAMs and wider memory data buses for upcoming Apple Silicon Macs. I would think the 3080 is idling most of the time in the Razer. For that matter, the M1 GPU would also be idling most of the time, I would think, but it is still fed more data per unit of time compared to the 3080.

 

[LinkRS]

The mobile RTX 3080 does not have 16 GBs of VRAM (source https://www.razer.com/gaming-laptops/Razer-Blade-15/RZ09-0409CEC3-R3U1). Plus, you are comparing a "desktop" against a laptop, desktops usually outperform laptops. It is not surprising in that benchmark the M1 outperforming the Intel CPU. You are comparing a 3.2 GHz processor against a 2.3 GHz, unless that Razer is cooled really well, that CPU will not run anywhere near its turbo speed for long. Even then, in mobile guise, Intel CPUs spend a lot of time throttling down, which the M1 will not have to do in the iMac chassis. This is a test that puts the Apple system in a much better situation for performing well than that laptop.

UPDATE: Correction, they do apparently have an SKU of the mobile RTX 3080 with 16 GBs of VRAM https://www.nvidia.com/en-us/geforce/gaming-laptops/ My apologies to the original poster


Rich S.

 

[Kung gu]

They avoid Davinci Resolve with dGPU because it's faster than both Premiere Pro and FCP.

You are wrong.
Source:

MacBook Pro M1 specs:
M1
8GB RAM 256SSD

PC Specs:
-CPU: AMD Ryzen 7 3700x cooled by a Noctua U14S
-GPU: NVidia GeForce GTX 1060 6GB
-Motherboard: GIGABYTE AORUS PRO B550 ATX AM4
-RAM: 16GB 3200Mhz CL16

The PC has a GTX 1060 and M1 is around GTX 1050.

The PC took 2:34 minutes and M1 took 1:34 minutes.

The M1 is faster than Ryzen with dGPU in exporting in Davinci Resolve.

You just want to down play the M1 and it shows.

I mean a RTX 3080 or a Ryzen 3700X getting beaten by a SoC thats used in an iPad in video editing is sad but impressive for Apple.

Bring on the next gen Apple Sillicon Apple and shut these haters up or will they find more excuses by using useless
fish benchmarks.

 

[quarkysg]

The mobile RTX 3080 does not have 16 GBs of VRAM (source https://www.razer.com/gaming-laptops/Razer-Blade-15/RZ09-0409CEC3-R3U1). Plus, you are comparing a "desktop" against a laptop, desktops usually outperform laptops. It is not surprising in that benchmark the M1 outperforming the Intel CPU. You are comparing a 3.2 GHz processor against a 2.3 GHz, unless that Razer is cooled really well, that CPU will not run anywhere near its turbo speed for long. Even then, in mobile guise, Intel CPUs spend a lot of time throttling down, which the M1 will not have to do in the iMac chassis. This is a test that puts the Apple system in a much better situation for performing well than that laptop.

Rich S.

The M1 MacBook Pro 13" has identical performance compared to the M1 iMac if I'm not wrong.

 

[Kung gu]

The mobile RTX 3080 does not have 16 GBs of VRAM (source https://www.razer.com/gaming-laptops/Razer-Blade-15/RZ09-0409CEC3-R3U1). Plus, you are comparing a "desktop" against a laptop, desktops usually outperform laptops. It is not surprising in that benchmark the M1 outperforming the Intel CPU. You are comparing a 3.2 GHz processor against a 2.3 GHz, unless that Razer is cooled really well, that CPU will not run anywhere near its turbo speed for long. Even then, in mobile guise, Intel CPUs spend a lot of time throttling down, which the M1 will not have to do in the iMac chassis. This is a test that puts the Apple system in a much better situation for performing well than that laptop.

Rich S.

This is complete crap. The M1 is used in MBP M1 and benchmarks the same as the iMac 24" and you WILL find the M1
MacBook Pro perform the same even after the running the benchmark for 10 minutes.

The iMac 24" M1 gets the same cinebench score as the M1 MBP.
Stop avoiding facts, there are plenty of YT videos that the M1 iMac 24", Mac mini M1 and MBP 13" M1 perform the same and this is evident in the 10 minute cinebench benchmarks.

The M1 is clocked at 3.2ghz in all M1 machines and it only throttles down a LOT in M1 MBA and M1 iPad Pro.

Also the Razer blade 15 advanced is thicker than the iMac 24".

The Razer blade is 15.8mm and the iMac is 11.5mm thick.
source: https://www2.razer.com/au-en/gaming-systems/razer-blade

 

[Kung gu]

You what's funny the the Razer blade is thicker than the iMac.
The Razer blade is 15.8mm and the iMac is 11.5mm thick.

 

[cmaier]

The mobile RTX 3080 does not have 16 GBs of VRAM (source https://www.razer.com/gaming-laptops/Razer-Blade-15/RZ09-0409CEC3-R3U1). Plus, you are comparing a "desktop" against a laptop, desktops usually outperform laptops. It is not surprising in that benchmark the M1 outperforming the Intel CPU. You are comparing a 3.2 GHz processor against a 2.3 GHz, unless that Razer is cooled really well, that CPU will not run anywhere near its turbo speed for long. Even then, in mobile guise, Intel CPUs spend a lot of time throttling down, which the M1 will not have to do in the iMac chassis. This is a test that puts the Apple system in a much better situation for performing well than that laptop.

Rich S.

The M1 also does not have 16GB of vram.

All laptops with M1 have exactly the same performance as desktops with M1, so how does the fact that this particular test was a desktop form factor matter?

 

[LinkRS]

This is complete crap. The M1 is used in MBP M1 and benchmarks the same as the iMac 24" and you WILL find the M1
MacBook Pro perform the same even after the running the benchmark for 10 minutes.

The iMac 24" M1 gets the same cinebench score as the M1 MBP.

You just helped my point. The M1 is a low-power (not low-performance) SOC. It runs much cooler, and with the exception of the MacBook Air, the M1 doesn't tend to be thermally throttled. It runs even better in the Mac Mini, and presumably the iMac due to even more room to move heat away from the system. In order to really test the M1 against a "RTX 3080" and Intel, you need to throw a desktop into the mix. I also stand corrected, apparently there is a version of the mobile RTX 3080 equipped with 16 GBs of VRAM, but that aside, the mobile 3080 is not the same performance level as the desktop 3080, despite the larger framebuffer.

Don't take me wrong, the M1 is an amazing SOC. What Apple has done is nothing short of extraordinary. But people need to stop puffing up their chests and treating it like Chuck Norris (https://winkgo.com/chuck-norris-memes/). Let's wait for the professional versions of Apple Silicon comes out (presumably) this year. I don't think we will need biased benchmarks to prove a point.

 

[LinkRS]

The M1 also does not have 16GB of vram.

All laptops with M1 have exactly the same performance as desktops with M1, so how does the fact that this particular test was a desktop form factor matter?

Apparently the current M1 seems to have the same performance specs despite the chassis (except for the MacBook Air which is passively cooled). This is NOT true for Intel systems and nVidia GPUs. The desktop counterparts are always faster. Desktops have more power and better cooling allowing clocks to stay higher and performance to be better. The referenced RTX 3080 "only" has 6144 CUDA cores, whereas the desktop version has 8704. The Core i7 10875H runs at 2.3 GHz base where as the "desktop equivalent" Core i7 10700 runs at 2.9 GHz (https://ark.intel.com/content/www/u...10700-processor-16m-cache-up-to-4-80-ghz.html).

 

[alien3dx]

hehe, just video . i need cyberpunk run on m1 ,

 

[Kung gu]

Apparently the current M1 seems to have the same performance specs despite the chassis (except for the MacBook Air which is passively cooled). This is NOT true for Intel systems and nVidia GPUs. The desktop counterparts are always faster. Desktops have more power and better cooling allowing clocks to stay higher and performance to be better. The referenced RTX 3080 "only" has 6144 CUDA cores, whereas the desktop version has 8704. The Core i7 10875H runs at 2.3 GHz base where as the "desktop equivalent" Core i7 10700 runs at 2.9 GHz (https://ark.intel.com/content/www/u...10700-processor-16m-cache-up-to-4-80-ghz.html).

since the M1 MBP 13" will have the same performance as the iMac 24", it shows that if this was conducted using a MBP 13" M1 the result will be same.

Thus showing that the M1 does indeed beat a RTX 3080 in video editing. Keep in mind the M1 GPU consumes around 10Watts of power however the RTX 3080 mobile has a MUCH MUCH higher power consumption.

Yes desktop had more power and I am sure the Apple Sillicon Mac Pro will have the same SoC as the 16" Apple Sillicon MBP when they do comes out.

I think the point you are missing to see is that the M1 is also an iPad Pro chip and when you look it from that perspective its hosestly embarrassing for Nvidia that the M1 does beat the RTX 3080 mobile(yes I know its the mobile version but the mobile version rtx 3080 is equal to a desktop 3070). The point is the RTX 3080 mobile should be faster
than M1 but the fact it is not is just sad considering the watts the mobile 3080 pulls and uses.

 

[Kung gu]

hehe, just video . i need cyberpunk run on m1 ,

it can run if CD project red make it for M1. The M1 is more powerful than PS4. I expect 720P. at low settings at 30 fps tho.

The 32Core GPU is what I am looking forward too.

 

[Appletoni]

But how does it do on random chess benchmarks?

Right, that’s the most important question?

 

[mi7chy]

You are wrong.
Source:

MacBook Pro M1 specs:
M1
8GB RAM 256SSD

PC Specs:
-CPU: AMD Ryzen 7 3700x cooled by a Noctua U14S
-GPU: NVidia GeForce GTX 1060 6GB
-Motherboard: GIGABYTE AORUS PRO B550 ATX AM4
-RAM: 16GB 3200Mhz CL16

The PC has a GTX 1060 and M1 is around GTX 1050.

The PC took 2:34 minutes and M1 took 1:34 minutes.

The M1 is faster than Ryzen with dGPU in exporting in Davinci Resolve.

You just want to down play the M1 and it shows.

I mean a RTX 3080 or a Ryzen 3700X getting beaten by a SoC thats used in an iPad in video editing is sad but impressive for Apple.

Bring on the next gen Apple Sillicon Apple and shut these haters up or will they find more excuses by using useless
fish benchmarks.

Look at the title of the thread. It's comparison with 3080 not 1060. Nice bait and switch.

 

[Kung gu]

Look at the title of the thread. It's comparison with 3080 not 1060. Nice bait and switch.

DUDE I MENTIONED THAT IN MY COMMENT. You annoy the **** out of me.

You made a point that the Davinci Resolve will be better with a dGPU and I showed you a PC with a dGPU that was using Davinci Resolve as a benchmark and the M1 was faster in Davinci even THOUGH the PC was using Davinci with a dGPU.

Now who's baiting and switching who when YOU brought up DAVINCI. As far I am concerned the OP did not mention Davinci.
You brought up Davinci, I showed a video where a PC using a more powerful GPU(GTX 1060) than the M1 gets beaten in Davinci.

Last time I checked the GTX 1060 is faster than the M1 GPU but the M1 still beat the Ryzen 7 GTX 1060 PC in video editing.

 

[Bug-Creator]

Plus, you are comparing a "desktop" against a laptop, desktops usually outperform laptops.

That laptop is probraly the closest PC to the (thermal) footprint of the iMac you could find that isn't a basline NUC (and hell that comparison would have sucked....) so it's fair game.

Noone knows how a proper desktop Apple-Silicon Mac would bench, cos Apple doesn't make those yet (and it will be a while).

 

[thunng8]

They avoid Davinci Resolve with dGPU because it's faster than both Premiere Pro and FCP.

Where is your evidence?

 

[thunng8]

That laptop is probraly the closest PC to the (thermal) footprint of the iMac you could find that isn't a basline NUC (and hell that comparison would have sucked....) so it's fair game.

Noone knows how a proper desktop Apple-Silicon Mac would bench, cos Apple doesn't make those yet (and it will be a while).

That laptop has a much higher thermal envelope than the iMac. The Rtx 3080 mobile chip by itself uses much more power than the entire iMac.

 

[eizen]

Look at the title of the thread. It's comparison with 3080 not 1060. Nice bait and switch.

FYI - That PC isn't using the studio version of resolve. That means it's not using the GPU for hardware acceleration.
The free version of Resolve on OSX does use hardware acceleration.

 

[dmccloud]

But Intel has a bunch of benchmarks that show Intel wipes the floor with the M1 in those particular ones. Intel Marketing says those are the only ones that matter. So unless M1 beats Intel in those then it'll always be flawed.

Intel also loves to lie and mislead people in their pressers. When they released the 11th gen mobile processors last year, Intel would only compare the performance of their rebranded iGPUs against the Radeon Vega 6 (an older iGPU) and the nVidia 250m, which is an older, workstation-class dGPU with less performance than the GTX 1050 it was based upon. I remember watching their event and laughing because their comparisons were so blatantly cherry-picked to prove Intel's "superiority". When it comes to assessing the performance of anything from Intel, I'll rely on testing by anyone outside Sunnyvale before I even look at Intel's drivel as anything more than a joke.