Mac Studio performance compared to 40xx GPUs

Get answers to your questions about color grading, editing and finishing with DaVinci Resolve.
  • Author
  • Message
Offline
User avatar

joe12south

  • Posts: 901
  • Joined: Tue Jan 16, 2018 4:14 pm
  • Location: Nashville, TN
  • Real Name: Joseph Moore

Mac Studio performance compared to 40xx GPUs

PostTue Jun 13, 2023 2:28 pm

I greatly prefer to use MacOS, but for performance reasons I use a custom-built Windows PC as my Resolve workstation.

Currently, that workstation is an i9-13900K with 128 GB RAM and an RTX 4080 GPU. I'm reasonably happy with the performance, including AI features and NR.

I'm curious if anyone has experience comparing RTX 40xx performance to M2 Ultra performance? I know it will do just fine with editing and basic grading, but concerned how it will fair with the heavier stuff. (I lean on NR, Magic Mask, and the new AI audio features quite a bit.)
Dedicated curmudgeon. Part-time artiste.
Offline

mpetech

  • Posts: 890
  • Joined: Wed Sep 04, 2013 9:52 pm
  • Real Name: Dom Silverio

Re: Mac Studio performance compared to 40xx GPUs

PostTue Jun 13, 2023 3:20 pm

Too early to tell. Most publications have the M2 Ultra's performance around the RTX 4070. The CPU performance is below Intel's 13900. Remember also, that you are stuck with that GPU for the life of that computer.

That said, you get an unmatched boost from ProRes encode and decode since Afterburner is built in.
Offline
User avatar

Uli Plank

  • Posts: 25430
  • Joined: Fri Feb 08, 2013 2:48 am
  • Location: Germany and Indonesia

Re: Mac Studio performance compared to 40xx GPUs

PostWed Jun 14, 2023 2:48 am

If you have the space and don't mind the energy bill, a PC desktop with a strong CPU and the 4090 will not be beaten by a Mac. But that little box is quite capable, nevertheless.
My disaster protection: export a .drp file to a physically separated storage regularly.
www.digitalproduction.com

Studio 19.1.3
MacOS 13.7.4, 2017 iMac, 32 GB, Radeon Pro 580 + eGPU
MacBook M1 Pro, 16 GPU cores, 32 GB RAM, MacOS 14.7.2
SE, USM G3
Offline
User avatar

joe12south

  • Posts: 901
  • Joined: Tue Jan 16, 2018 4:14 pm
  • Location: Nashville, TN
  • Real Name: Joseph Moore

Re: Mac Studio performance compared to 40xx GPUs

PostWed Jun 14, 2023 3:07 am

If you have the space and don't mind the energy bill, a PC desktop with a strong CPU and the 4090 will not be beaten by a Mac.

At this point, I'm just tired of feeling beaten by Windows. ;)
Dedicated curmudgeon. Part-time artiste.
Offline
User avatar

Uli Plank

  • Posts: 25430
  • Joined: Fri Feb 08, 2013 2:48 am
  • Location: Germany and Indonesia

Re: Mac Studio performance compared to 40xx GPUs

PostWed Jun 14, 2023 3:40 am

I like that one ;-)
My disaster protection: export a .drp file to a physically separated storage regularly.
www.digitalproduction.com

Studio 19.1.3
MacOS 13.7.4, 2017 iMac, 32 GB, Radeon Pro 580 + eGPU
MacBook M1 Pro, 16 GPU cores, 32 GB RAM, MacOS 14.7.2
SE, USM G3
Offline

Sam Steti

  • Posts: 3141
  • Joined: Tue Jun 17, 2014 7:29 am
  • Location: France

Re: Mac Studio performance compared to 40xx GPUs

PostWed Jun 14, 2023 11:35 am

joe12south wrote:At this point, I'm just tired of feeling beaten by Windows. ;)
Which you can admit go a bit off the tracks of your original post based on performance only ;)
*MacMini M1 16 Go - Sonoma - Ext nvme SSDs on TB3 - 14 To HD in 2 x 4 disks USB3 towers
*Legacy MacPro 8core Xeons, 32 Go ram, 2 x gtx 980 ti, 3SSDs including RAID
*Resolve Studio everywhere, Fusion Studio too
*https://www.buymeacoffee.com/videorhin
Offline
User avatar

joe12south

  • Posts: 901
  • Joined: Tue Jan 16, 2018 4:14 pm
  • Location: Nashville, TN
  • Real Name: Joseph Moore

Re: Mac Studio performance compared to 40xx GPUs

PostWed Jun 14, 2023 2:57 pm

Sam Steti wrote:Which you can admit go a bit off the tracks of your original post based on performance only ;)

Well, no it's exactly to the point. If all I cared about was performance, I'd stick with Windows. I'd prefer "good enough" performance in a less user-hostile environment. So, the only question for me to figure out is if the M2 Ultra is "good enough" to squeeze out what I need.

PS. If you like Windows, great. No need to try to convince me I should.
Dedicated curmudgeon. Part-time artiste.
Offline

Sam Steti

  • Posts: 3141
  • Joined: Tue Jun 17, 2014 7:29 am
  • Location: France

Re: Mac Studio performance compared to 40xx GPUs

PostWed Jun 14, 2023 3:05 pm

:lol: I'm on OSX since... OS9 actually...
So no, don't rely on me to praise Windows I've been avoiding for ages ;)
*MacMini M1 16 Go - Sonoma - Ext nvme SSDs on TB3 - 14 To HD in 2 x 4 disks USB3 towers
*Legacy MacPro 8core Xeons, 32 Go ram, 2 x gtx 980 ti, 3SSDs including RAID
*Resolve Studio everywhere, Fusion Studio too
*https://www.buymeacoffee.com/videorhin
Offline
User avatar

Rakesh Malik

  • Posts: 3378
  • Joined: Fri Dec 07, 2012 1:01 am
  • Location: Vancouver, BC

Re: Mac Studio performance compared to 40xx GPUs

PostWed Jun 14, 2023 5:41 pm

joe12south wrote:
Sam Steti wrote:Which you can admit go a bit off the tracks of your original post based on performance only ;)

Well, no it's exactly to the point. If all I cared about was performance, I'd stick with Windows. I'd prefer "good enough" performance in a less user-hostile environment. So, the only question for me to figure out is if the M2 Ultra is "good enough" to squeeze out what I need.


That's a tough call, because it depends on what "good enough" is for you, which isn't the same for everyone. For editors a mac is fine.

The reality is that Apple will never be able to keep up with x86 ecosystem in unfettered performance, and it's pretty much telling you that by comparing to very outdated Intel CPUs and ignoring the AMD CPUs that are cleaning Intel's clocks. And never mind the GPU... it's not in the ballpark, but for video makes up for it with hardware ProRes, which is an inexpensive way to boost performance in a power miserly way at the expense of flexibility.

For color grading, it's probably enough. For me not any more, since I'm doing more and more VFX these days, and the more I develop my skills in Houdini the bigger the machine I need.

Just an example -- you know your needs better than anyone else.

I'm pretty much OS neutral for the most part; once you fire up the software that's the UI. Houdini is Houdini, Resolve is Resolve, Scratch is Scratch.
Rakesh Malik
Cinematographer, VFX Artist, photographer, adventurer, martial artist
http://WinterLightStudios.ca
System:
Asus ProArt 16/64GB/12 core Zen5/nVidia RTX 4070 8GB
Nuke/Houdini/Resolve
Offline
User avatar

Uli Plank

  • Posts: 25430
  • Joined: Fri Feb 08, 2013 2:48 am
  • Location: Germany and Indonesia

Re: Mac Studio performance compared to 40xx GPUs

PostThu Jun 15, 2023 2:01 am

That sums it up pretty well. A current Mac is fine for editing and regular color grading, including secondary. Plus, it’s silent and contributes less to climatic change.
But as soon as you get into heavy VFX, whatever software you use, you can’t throw too much raw computing power and RAM at it.
My disaster protection: export a .drp file to a physically separated storage regularly.
www.digitalproduction.com

Studio 19.1.3
MacOS 13.7.4, 2017 iMac, 32 GB, Radeon Pro 580 + eGPU
MacBook M1 Pro, 16 GPU cores, 32 GB RAM, MacOS 14.7.2
SE, USM G3
Offline

Nick2021

  • Posts: 907
  • Joined: Thu May 13, 2021 3:19 am
  • Real Name: Nick Zentena

Re: Mac Studio performance compared to 40xx GPUs

PostThu Jun 15, 2023 3:23 am

If you're worried about climate change you want something you can update or repair easily over time. There is a reason right to repair laws have spread.

Let's also not forget the biggest energy user for many is the multiple large monitors. My desktop at idle when it is waiting for me to do something uses very little power. It only ramps up when it needs to.

Of course total power usage is over time. A slow machine that needs to work longer can actually end up using more total power.
Offline
User avatar

Uli Plank

  • Posts: 25430
  • Joined: Fri Feb 08, 2013 2:48 am
  • Location: Germany and Indonesia

Re: Mac Studio performance compared to 40xx GPUs

PostThu Jun 15, 2023 4:17 am

Important points for sure. BTW, we may need a discussion about HDR too.
In case you didn't yet read about it: at least for batteries there will be strict laws in Europe soon, which even Apple needs to respect. No more glued in batteries or the like.
My disaster protection: export a .drp file to a physically separated storage regularly.
www.digitalproduction.com

Studio 19.1.3
MacOS 13.7.4, 2017 iMac, 32 GB, Radeon Pro 580 + eGPU
MacBook M1 Pro, 16 GPU cores, 32 GB RAM, MacOS 14.7.2
SE, USM G3
Offline
User avatar

Rakesh Malik

  • Posts: 3378
  • Joined: Fri Dec 07, 2012 1:01 am
  • Location: Vancouver, BC

Re: Mac Studio performance compared to 40xx GPUs

PostThu Jun 15, 2023 7:41 am

True, HDR is a pretty obnoxious can of worms still.
Rakesh Malik
Cinematographer, VFX Artist, photographer, adventurer, martial artist
http://WinterLightStudios.ca
System:
Asus ProArt 16/64GB/12 core Zen5/nVidia RTX 4070 8GB
Nuke/Houdini/Resolve
Offline
User avatar

Laniakea

  • Posts: 83
  • Joined: Thu Aug 12, 2021 6:58 pm
  • Real Name: Jay Deacon

Re: Mac Studio performance compared to 40xx GPUs

PostThu Jun 15, 2023 10:46 am

joe12south wrote:
If you have the space and don't mind the energy bill, a PC desktop with a strong CPU and the 4090 will not be beaten by a Mac.

At this point, I'm just tired of feeling beaten by Windows. ;)


Though many here have spoken about the ‘greatness’ of using Windows over MacOS, I never see Blackmagic demo Resolve on the Windows platform?!
Resolve Studio 20 beta
MacOS 14.7.2
2 x Apple Studio Monitor
MacStudio M1 Ultra 128GB/2TB SSD 48c
BM Colour Mini Panel
2 x G-Drive Professional 7.68TB SSD (TB3)
GTEC-8TB RAID 0 (TB2)
WD 6TB HDD (USB 3)
Focusrite Clarett+ 8Pre USB-C AI
3 x Yamaha HS7 (3.1)
Offline
User avatar

joe12south

  • Posts: 901
  • Joined: Tue Jan 16, 2018 4:14 pm
  • Location: Nashville, TN
  • Real Name: Joseph Moore

Re: Mac Studio performance compared to 40xx GPUs

PostSat Jun 24, 2023 9:55 pm

Rakesh Malik wrote:That's a tough call, because it depends on what "good enough" is for you, which isn't the same for everyone. For editors a mac is fine.

On a 4K timeline:
- Real-time performance from NR.
- Minimal pre-rendering for AI nodes like Magic Mask.
- Minimal timeline lag with the AI dialog and compressor effects.

In short, basically the same performance as I get out of a single 4080. I think I'll need to take a project to the Apple store and test for myself.
Dedicated curmudgeon. Part-time artiste.
Offline

Alex Silva

  • Posts: 457
  • Joined: Thu Jul 10, 2014 8:12 am

Re: Mac Studio performance compared to 40xx GPUs

PostSat Jun 24, 2023 11:32 pm

Just a point regarding energy consumption.
Note that the time spent to accomplish a job matters for energy consumption.

If a 800w system does the job in 10 minutes and a 200w system does the job in 60 minutes. The 800w system will have spent less energy overall: proportionally 8 vs 12 and that not counting other ancillary equipment that needs to be turned on concurrently.
Offline
User avatar

Uli Plank

  • Posts: 25430
  • Joined: Fri Feb 08, 2013 2:48 am
  • Location: Germany and Indonesia

Re: Mac Studio performance compared to 40xx GPUs

PostSun Jun 25, 2023 12:11 am

Absolutely. But things get more complex when you consider that most systems are not rendering all the time. Both a PC and a Mac will reduce power consumption when you are editing, grading and thinking.
This can and should all be measured for a decision, if energy consumption matters to you.
My disaster protection: export a .drp file to a physically separated storage regularly.
www.digitalproduction.com

Studio 19.1.3
MacOS 13.7.4, 2017 iMac, 32 GB, Radeon Pro 580 + eGPU
MacBook M1 Pro, 16 GPU cores, 32 GB RAM, MacOS 14.7.2
SE, USM G3
Offline
User avatar

Laniakea

  • Posts: 83
  • Joined: Thu Aug 12, 2021 6:58 pm
  • Real Name: Jay Deacon

Re: Mac Studio performance compared to 40xx GPUs

PostWed Aug 30, 2023 11:31 am

With regard to video post - I'd trade a little bit of performance to use MacOS instead of Window any day! I have a strong feeling BM and most other companies that produce similar products, favour Mac over Windows. This may upset certain people, but it appears to be the same between iOS and Android. I use a lot of NR, often on a 4k timeline, and Im able to work just fine without proxies. Maybe adjust your workflow slightly and you'll be back in Macland!
Resolve Studio 20 beta
MacOS 14.7.2
2 x Apple Studio Monitor
MacStudio M1 Ultra 128GB/2TB SSD 48c
BM Colour Mini Panel
2 x G-Drive Professional 7.68TB SSD (TB3)
GTEC-8TB RAID 0 (TB2)
WD 6TB HDD (USB 3)
Focusrite Clarett+ 8Pre USB-C AI
3 x Yamaha HS7 (3.1)
Offline
User avatar

Uli Plank

  • Posts: 25430
  • Joined: Fri Feb 08, 2013 2:48 am
  • Location: Germany and Indonesia

Re: Mac Studio performance compared to 40xx GPUs

PostWed Aug 30, 2023 12:08 pm

I've tested my benchmark project on a maxed out Studio M2 Ultra and I have to say I'm impressed:
9 Watts when idle and 104 Watts max, but most of the time around 90. All 76 GPU cores maxed out, using 80-90 GB RAM. And very low noise, the fans running at about 30% in a room with 27 Celsius.

Make sure to buy any config with enough RAM!
38 GPU cores will want more than 32 GB (like 40), and 16 will ask for about 20.
My disaster protection: export a .drp file to a physically separated storage regularly.
www.digitalproduction.com

Studio 19.1.3
MacOS 13.7.4, 2017 iMac, 32 GB, Radeon Pro 580 + eGPU
MacBook M1 Pro, 16 GPU cores, 32 GB RAM, MacOS 14.7.2
SE, USM G3
Offline
User avatar

Laniakea

  • Posts: 83
  • Joined: Thu Aug 12, 2021 6:58 pm
  • Real Name: Jay Deacon

Re: Mac Studio performance compared to 40xx GPUs

PostWed Aug 30, 2023 12:12 pm

Uli Plank wrote:I've tested my benchmark project on a maxed out Studio M2 Ultra and I have to say I'm impressed:
9 Watts when idle and 104 Watts max, but most of the time around 90. All 76 GPU cores maxed out, using 80-90 GB RAM. And very low noise, the fans running at about 30% in a room with 27 Celsius.

Make sure to buy any config with enough RAM!
38 GPU cores will want more than 32 GB (like 40), and 16 will ask for about 20.


Have you got one on order Uli? ;-)
Resolve Studio 20 beta
MacOS 14.7.2
2 x Apple Studio Monitor
MacStudio M1 Ultra 128GB/2TB SSD 48c
BM Colour Mini Panel
2 x G-Drive Professional 7.68TB SSD (TB3)
GTEC-8TB RAID 0 (TB2)
WD 6TB HDD (USB 3)
Focusrite Clarett+ 8Pre USB-C AI
3 x Yamaha HS7 (3.1)
Offline
User avatar

Uli Plank

  • Posts: 25430
  • Joined: Fri Feb 08, 2013 2:48 am
  • Location: Germany and Indonesia

Re: Mac Studio performance compared to 40xx GPUs

PostWed Aug 30, 2023 12:38 pm

No, we got one for testing for two months, together with a colleague who is into compositing.
We both write for Digital Production (Munich).

If you want to test yours with our benchmark, drop me a PM.
My disaster protection: export a .drp file to a physically separated storage regularly.
www.digitalproduction.com

Studio 19.1.3
MacOS 13.7.4, 2017 iMac, 32 GB, Radeon Pro 580 + eGPU
MacBook M1 Pro, 16 GPU cores, 32 GB RAM, MacOS 14.7.2
SE, USM G3
Offline
User avatar

Uli Plank

  • Posts: 25430
  • Joined: Fri Feb 08, 2013 2:48 am
  • Location: Germany and Indonesia

Re: Mac Studio performance compared to 40xx GPUs

PostSun Oct 01, 2023 7:34 pm

So, I can now tell you, thanks to a nice colleague in the German forum for DR, that a Ryzen 9 3950X with 128GB RAM and a RTX 4090 with 24GB VRAM is rendering our benchmark project in UHD in 6 minutes and 11 seconds into H.265 (set to Master in DR).

Our maxed out Mac Studio M2 needed 7 minutes 42 seconds.
My disaster protection: export a .drp file to a physically separated storage regularly.
www.digitalproduction.com

Studio 19.1.3
MacOS 13.7.4, 2017 iMac, 32 GB, Radeon Pro 580 + eGPU
MacBook M1 Pro, 16 GPU cores, 32 GB RAM, MacOS 14.7.2
SE, USM G3
Offline

John Spirou

  • Posts: 125
  • Joined: Sun Sep 17, 2017 3:51 pm

Mac Studio performance compared to 40xx GPUs

PostSun Oct 01, 2023 8:49 pm

Can mac ultra play 4k native timeline in real time with heavy noise reduction and colour grading?

Can we have a link with the demo benchmark project?
Offline
User avatar

Paul Draper

  • Posts: 202
  • Joined: Thu Apr 27, 2017 8:33 pm

Re: Mac Studio performance compared to 40xx GPUs

PostSun Oct 01, 2023 10:30 pm

Uli Plank wrote:If you have the space and don't mind the energy bill, a PC desktop with a strong CPU and the 4090 will not be beaten by a Mac. But that little box is quite capable, nevertheless.
Agreed, we're never going to get the necessary GPU performance for this kind of NLE work out of Apple's over-hyped & overpriced boxes right now, maybe check back in another five years? For now, the Mac Studios certainly have their pro-workflow uses (like a recording studio for example, where they can suit very well indeed & where the GPU is largely irrelevant).

A couple of Nividia cards with a good, silent chassis & PSU on Win11 or Linux. Not that hard or overpriced really & likely worthwhile if this workflow is important to making a living vs. winging about 'Apples vs PCs' etc. Also check out Puget reviews of real systems in action for real stats vs. Apple's local BS.

https://www.pugetsystems.com/all-articles/
MacPro 7,1 16 core, 192GB, MacOS 15.1, Vega II Duo. Antelope Pure 2 & Orion 32+, UA Apollo x8. RAID-4 Thunderbay 6, RAID-0 Sonnet M.2 4x4. MiniMon 4k, Dell U3415W & BenQ SW2700PT. Nuendo, Wavelab, Resolve Studio 19
Offline
User avatar

Marc Wielage

  • Posts: 13241
  • Joined: Fri Oct 18, 2013 2:46 am
  • Location: Palm Springs, California

Re: Mac Studio performance compared to 40xx GPUs

PostSun Oct 01, 2023 10:50 pm

John Spirou wrote:Can mac ultra play 4k native timeline in real time with heavy noise reduction and colour grading? Can we have a link with the demo benchmark project?

I can tell you that the 2019 Mac Pro we have can. We have a couple of 4K projects (on Red R3D raw and one Alexa 3.5K Raw) that play back fine in real time, both with portions using Dehancer as a look, both with 36 nodes on every shot. The only time it slows down is on transitions with multiple tracks of 4K or 5K material, and those I had to cache.
Certified DaVinci Resolve Color Trainer • AdvancedColorTraining.com
Offline

ricardo marty

  • Posts: 1761
  • Joined: Wed Apr 26, 2017 4:03 am

Re: Mac Studio performance compared to 40xx GPUs

PostSun Oct 01, 2023 11:55 pm

Its no secret That Grant Petty is a mac head, I dont think hes making a statement. He started on that system and loves it. BMD first products are all mac. But he is a buisness man and his company is also about making money.
DV Studio 20.| Asus ProArt PD5, 2.5 GHz i7 16-Core 64GB nvidia 3070 572.83| Apple M4 10-Core 24GB Unified RAM 512GB-Core Neural Engine | Lenovo 7x Snapdragon 78-100 12 core elite 32gb ram adreno gpu, Win 11.| Sony A6700. PreSonus AudioBox |Speed Editor
Offline
User avatar

Uli Plank

  • Posts: 25430
  • Joined: Fri Feb 08, 2013 2:48 am
  • Location: Germany and Indonesia

Re: Mac Studio performance compared to 40xx GPUs

PostMon Oct 02, 2023 6:13 am

John Spirou wrote:Can we have a link with the demo benchmark project?

It's using demo material from sources like Arri, Sony and others, so I don't want to make it public. But if you send me a PM with any email address of yours, I'll let you into that Google Drive folder.

It can be rendered in 8K, including some Superscale upscaling of 6K sources (all others are 8K). Rendering in 4K/UHD should happen without that set, of course.
My disaster protection: export a .drp file to a physically separated storage regularly.
www.digitalproduction.com

Studio 19.1.3
MacOS 13.7.4, 2017 iMac, 32 GB, Radeon Pro 580 + eGPU
MacBook M1 Pro, 16 GPU cores, 32 GB RAM, MacOS 14.7.2
SE, USM G3
Offline
User avatar

Laniakea

  • Posts: 83
  • Joined: Thu Aug 12, 2021 6:58 pm
  • Real Name: Jay Deacon

Re: Mac Studio performance compared to 40xx GPUs

PostThu Oct 05, 2023 6:14 pm

Uli Plank wrote:If you have the space and don't mind the energy bill, a PC desktop with a strong CPU and the 4090 will not be beaten by a Mac. But that little box is quite capable, nevertheless.



While a PC with a strong CPU and a 4090 graphics card undoubtedly offers remarkable performance, it's crucial to consider the reliability and ecosystem factors. Many professionals in the film and post-production industry prefer MacOS over Windows due to its stability and streamlined user experience. This preference is magnified when using DaVinci Resolve, which was developed with a clear Mac-first approach in mind, as indicated by Blackmagic Design's director Grant Petty's well-known affinity for Mac systems.

The software is optimized for MacOS, providing a smoother, more stable experience, which is especially important in a professional setting where downtime is costly. Additionally, Blackmagic's hardware offerings, such as the Speed Editor or UltraStudio Monitor 3G, are often better integrated with Mac systems, adding another layer of efficiency and reliability.

So while the raw power of a high-end PC can't be ignored, the combination of MacOS optimization and Blackmagic's Mac-focused approach can make a Mac system, especially with the M1 Pro chip, a compelling choice for professionals in film editing and post-production. The decision isn't just about hardware specs; it's about the entire ecosystem and how well all the components work together.
Resolve Studio 20 beta
MacOS 14.7.2
2 x Apple Studio Monitor
MacStudio M1 Ultra 128GB/2TB SSD 48c
BM Colour Mini Panel
2 x G-Drive Professional 7.68TB SSD (TB3)
GTEC-8TB RAID 0 (TB2)
WD 6TB HDD (USB 3)
Focusrite Clarett+ 8Pre USB-C AI
3 x Yamaha HS7 (3.1)
Offline
User avatar

Uli Plank

  • Posts: 25430
  • Joined: Fri Feb 08, 2013 2:48 am
  • Location: Germany and Indonesia

Re: Mac Studio performance compared to 40xx GPUs

PostFri Oct 06, 2023 8:43 am

While I didn't get many responses here, a few users in the German forum did the same benchmark on PCs.
The best result was 15 minutes and 57 seconds for the 8K version on a machine with a Ryzen 9 3950X, 128GB RAM, and a RTX 4090 24GB.
The Mac Studio Ultra needed 35 minutes and 18 seconds.
But then, the Mac was running a tad over 50 Watts, while the PC needs more than 10 times as much.

If you want to compare your hardware, drop me a PM and any email address, so I can give you access.
I have a version in UHD too for those who think their machine will crash with 8K. From our other testers, nothing came close to the 4090 machine and most other configurations didn't get through the 8K version.

The time needed for the same project, but without any Superscale Enhanced (all sources are over 4K), was 7:42 on the Mac and 6:11 on the PC. It seems that the lack of hardware encoders is playing a bigger role here.
My disaster protection: export a .drp file to a physically separated storage regularly.
www.digitalproduction.com

Studio 19.1.3
MacOS 13.7.4, 2017 iMac, 32 GB, Radeon Pro 580 + eGPU
MacBook M1 Pro, 16 GPU cores, 32 GB RAM, MacOS 14.7.2
SE, USM G3
Offline

VMFXBV

  • Posts: 803
  • Joined: Wed Aug 24, 2022 8:41 pm
  • Real Name: Andrew I. Veli

Re: Mac Studio performance compared to 40xx GPUs

PostFri Oct 06, 2023 10:18 am

Uli Plank wrote:But then, the Mac was running a tad over 50 Watts, while the PC needs more than 10 times as much.



Don't wanna be that guy but even their own spec page say the Ultra consumes 215W when loaded. And even that seems a bit exagerrated on the low side, Apple style. ARM doesn't scale well beyond its sweet spot.
AMD Ryzen 5800X3D
AMD Radeon 7900XTX
Ursa Mini 4.6K
Pocket 4K
Offline
User avatar

Uli Plank

  • Posts: 25430
  • Joined: Fri Feb 08, 2013 2:48 am
  • Location: Germany and Indonesia

Re: Mac Studio performance compared to 40xx GPUs

PostFri Oct 06, 2023 11:20 am

Well, I used an external meter. I had tested it with household items for precision, like Tungsten lamps (yes, there are still a few) and a heater. I was surprised too. But I'd be glad if someone with that machine double-checks. BTW, a bit over 50 was the average, it went up to 102 in short bursts.

When idle it went down to 9.
My disaster protection: export a .drp file to a physically separated storage regularly.
www.digitalproduction.com

Studio 19.1.3
MacOS 13.7.4, 2017 iMac, 32 GB, Radeon Pro 580 + eGPU
MacBook M1 Pro, 16 GPU cores, 32 GB RAM, MacOS 14.7.2
SE, USM G3
Offline
User avatar

Yasser Saeed

  • Posts: 736
  • Joined: Tue Feb 16, 2016 3:46 pm
  • Location: Egypt

Re: Mac Studio performance compared to 40xx GPUs

PostFri Oct 06, 2023 12:29 pm

Laniakea wrote:Many professionals in the film and post-production industry prefer MacOS over Windows due to its stability

Sorry but I totally disagree as this statement used to be true in the far past. I know many friends who own in their studios both Mac and Windows workstations and they often confess to me that the reputation of Mac stability advantage over Windows is not true at all as their Mac computers crash as much as Windows and sometimes even more.

Keep in mind they are using reputed brands of Windows workstations that costed them as much as thier Macs and this is the fairest way to compare Windows to Mac. It is not fair to compare a Mac computer that costs 6K to unbranded and custom build PC that costes 2K! Same concept when comparing smartphones .. the fair way is to compare iPhone to Pixel when it comes to performance and reliability .. not iPhone to Galaxy.

Laniakea wrote:his preference is magnified when using DaVinci Resolve, which was developed with a clear Mac-first approach in mind, as indicated by Blackmagic Design's director Grant Petty's well-known affinity for Mac systems. The software is optimized for MacOS, providing a smoother, more stable experience, which is especially important in a professional setting where downtime is costly.
This is also old news and now days I see BMD cares about Windows as much as Mac systems.

Laniakea wrote:The decision isn't just about hardware specs; it's about the entire ecosystem and how well all the components work together.

Agree, that is why it is important to consider Windows workstations from reputed brands such as HP, DELL and BOXX who design most of hardware, fully test it making sure all components work together at maximum compatibility and reliability.
HP Z2 Mini G9 Workstation
OS: Windows 11 Pro
Display: HP Z27 QHD DreamColor
CPU: i9-13900K 24 core
GPU: RTX 4000 SFF 20GB Ada Gen
RAM: 96GB DDR5-5200
NVMe SSD: OS 4TB, Data 4TB
NLE: DaVinci Resolve Studio 20 Beta
Offline

SkierEvans

  • Posts: 1277
  • Joined: Wed Jan 24, 2018 9:59 pm
  • Location: Ottawa, Ontario
  • Real Name: Ron Evans

Re: Mac Studio performance compared to 40xx GPUs

PostFri Oct 06, 2023 1:04 pm

Yasser Saeed wrote: This is also old news and now days I see BMD cares about Windows as much as Mac systems.


Agree. I am sure there are more BM hardware cards in PC's than Mac's. BM is a hardware company after all. They work with a lot of NLE's that are PC only like EDIUS , Vegas etc. I am an example. I bought BM cards to work with EDIUS a long time before I used Resolve and only last year got a Studio Max. My first Apple product ever.

My take now having both. The Studio Max is silent and will decode h265/h264 realtime for editing. As for further editing I often move to my PC that for render is almost as fast even though it does not have the hardware decode. For 8bit source the PC is faster. The PC also has the advantage in available software. I do most of my encoding with TMPGenc which is PC only and author for DVD and Bluray. Use Nero to burn multiple discs at a time or even multiple USB drives. Does the PC cause issues at times. Yes, mainly by Windows updating or partial update without warning even if it has been told not to update !! Just like the Mac one may find that user software versions expect a particular version of the OS to work correctly.
Threadripper 1920, Gigabyte X399 DESIGNARE EX, 32G RAM, Gigabyte 4070Ti 12G, ASUS PB328Q, IP4K, WIN10 Pro 22H2, Speed Editor

Resolve Studio 19, EDIUS 9WG,EDIUS X WG, Vegas 18

Studio Max M1 24 core GPU, 32G, 1T drive. iPad Pro 12.9` M2 16G, 1T
Offline

VMFXBV

  • Posts: 803
  • Joined: Wed Aug 24, 2022 8:41 pm
  • Real Name: Andrew I. Veli

Re: Mac Studio performance compared to 40xx GPUs

PostFri Oct 06, 2023 1:38 pm

SkierEvans wrote:
Agree. I am sure there are more BM hardware cards in PC's than Mac's.


I really doubt any professional really uses MACs for any serious work beside simple editing and color grading. CGI needs cores and copious amounts of RAM (think 512GB+ for Houdini sims) as some scenes cannot be contained on today's GPU VRAM, even on those that share system RAM.
AMD Ryzen 5800X3D
AMD Radeon 7900XTX
Ursa Mini 4.6K
Pocket 4K
Offline

steve oakley

  • Posts: 723
  • Joined: Fri Aug 16, 2013 6:07 pm

Re: Mac Studio performance compared to 40xx GPUs

PostFri Oct 06, 2023 4:26 pm

I'm editing an 8K project on a M1 Ultra 128. it's a TL thats running 8-16 layers of 1080/2K or 4K + PNG's , its actually 3 screens of 1080 plus space to accommodate physical layout. in many sections it will just play if I have just a simple 2 node grade, maybe 1 or 2 resolve FX on several layers ( blur, grain). everything has position, crop & scaling on it. camera clips are h264 10bit 422 400mbit from C300-2, h265 4K 10bit h265 400 bit from XH2s ( table top motion control rig with laowa lens ), plus layers of graffic elements rendered in ProRes 2K-8K + apply mode. It's a LOT of stuff going on and it's ok. the fact I can play sections in early stages w/o rendering is pretty amazing to me because I know this would of crushed my older intel Macs i9 with AMD GPU's. Project would have been unworkable. Even when I do render the TL ( auto cache mode ) it's pretty painless, especially once intermediate caches are present. I can make changes and play through at 12fps or better.

Only Deahancer makes for boggy sections, but I'm also using 2-4 instances on a given frame :) the 7.0 update has helped performance a bit.

if you are just editing 4K with NR and some other stuff, no worries whatsoever.

most of the time the system is using 20-50w. under max render load its about 150w :)

storage is dedicated TB3 SSD for project, dedicated TB3 SSD for cache on separate bus. both hit 2500-2700mb/sec max bus speed. This is probably the big limiting factor right now.

also rendering out bits in Fusion when it makes sense. all good, in fact much better. probably so far most stable release yet.

R is pegging out the GPU at 90-98% :) and thanks to the integrated RAM never runs out of VRAM :) when you get to 8K you'll like this.

troublesome GoPro or DJI drone footage that always choked intel systems edits like I frame now :)

all said, the M1/M2 ultra is a beast of a machine that will get what you need done fast, fluid, stable, low power. a machine that uses 10X the power to be 1 min faster on a 6 min render just doesn't matter, in fact the energy use is a big detraction, never mind size, heat, noise. At some point you realize the hardware we have now is so fast and capable you just don't need more. at least for a while. maybe if 8K becomes common you'll be wanting more, but for most 4K single screen work even if its a. bit complex, there is more than enough HP for *most* users *most* of the time. extreme case users feel free to make your extreme use case, but getting into a machine if your needs is gross overkill for most folks from all perspectives.

S
S
mac Studio Ultra 128gb OS 14.5+
TB3 ext 1TB SSD Cache, 1TB SSD Sys
4K 27Dell
Presonus Quantum 2626 TB3 Word clocked 96K
MiniPanel via enet
Behinger Xtouch USB via USB3 Hub->PCIe USB3X4 card
Wacom XL Pro
Offline

VMFXBV

  • Posts: 803
  • Joined: Wed Aug 24, 2022 8:41 pm
  • Real Name: Andrew I. Veli

Re: Mac Studio performance compared to 40xx GPUs

PostFri Oct 06, 2023 9:42 pm

steve oakley wrote:
a machine that uses 10X the power to be 1 min faster on a 6 min render just doesn't matter, in fact the energy use is a big detraction, never mind size, heat, noise.



It really doesn't use 10X power...More like 2-3 times on average but then again its much faster in other aspects as well.
AMD Ryzen 5800X3D
AMD Radeon 7900XTX
Ursa Mini 4.6K
Pocket 4K
Offline

steve oakley

  • Posts: 723
  • Joined: Fri Aug 16, 2013 6:07 pm

Re: Mac Studio performance compared to 40xx GPUs

PostFri Oct 06, 2023 11:36 pm

If you are playing 24/25/30 FPS w/o dropping frames with multiple layers of video in 4K with filters, grades.... how much faster does it need to be ? 10% faster doesn't really make a difference for most people most of the time. Thats not "much faster", thats marginally faster. there is a point of diminishing returns for $, power usage. edge cases don't count. average user needs do and these machines are very much more than fast enough, and faster still just doesn't matter anymore, until 8K. Even that will just be next generation of machines upgrade and you are good.

S
S
mac Studio Ultra 128gb OS 14.5+
TB3 ext 1TB SSD Cache, 1TB SSD Sys
4K 27Dell
Presonus Quantum 2626 TB3 Word clocked 96K
MiniPanel via enet
Behinger Xtouch USB via USB3 Hub->PCIe USB3X4 card
Wacom XL Pro
Offline

Nick2021

  • Posts: 907
  • Joined: Thu May 13, 2021 3:19 am
  • Real Name: Nick Zentena

Re: Mac Studio performance compared to 40xx GPUs

PostSat Oct 07, 2023 3:46 am

steve oakley wrote:until 8K. Even that will just be next generation of machines upgrade and you are good.

S


And there goes any savings you've managed. Tossing out a machine is far less efficient then just upgrading a GPU or other part.

The whole reason for the right to repair movement is the disposable culture is the worst thing possible for the environment.
Offline

Michael Haggard

  • Posts: 8
  • Joined: Tue Sep 19, 2023 3:38 pm
  • Real Name: Michael Haggrd

Re: Mac Studio performance compared to 40xx GPUs

PostSat Oct 07, 2023 4:57 am

No benchmarks, but I have a Mac Studio M2 ultra 128GB standard cores and it doesn’t perform half as well as I would have excepted.

I’m editing 4K footage on a 25fps HD timeline and a single node of Resolve’s noise reduction on ‘better’ setting drops me to 5-10fps on playback. Without noise reduction, a few masks with blur nodes will do the same.

If I bypass effects, I can playback the footage without dropouts. Cloning on fusion is another heavy operation that will drop me below 25fps.

Not the workhorse I was expecting. It also doesn’t play well with non-Apple external displays.
Offline

steve oakley

  • Posts: 723
  • Joined: Fri Aug 16, 2013 6:07 pm

Re: Mac Studio performance compared to 40xx GPUs

PostSat Oct 07, 2023 6:12 am

Nick2021 wrote:
steve oakley wrote:until 8K. Even that will just be next generation of machines upgrade and you are good.

S


And there goes any savings you've managed. Tossing out a machine is far less efficient then just upgrading a GPU or other part.

The whole reason for the right to repair movement is the disposable culture is the worst thing possible for the environment.


Huh ? I can edit 8K right now - read my first post please. I'm doing it. these machines just won't do as many layers. I guess if you don't write an encylopedia then any summarized statement will be 100% misconstrued, taken of out context and twisted around to mean what you didn't actually mean :roll: .

I don't view the studio machines as "disposable" or only good for 2-3 years. They have a good 5+ year life as primary machines and can then slip into secondary uses. certainly can run as a DAW indefinitely at this point. I've had 60+ tracks running just fine... unless you want to claim you need to run 250 tracks of 192K 32bit in which case you can VST link two or three machines together with Cubase :o :D

I made no claims about saving any money on cost of hardware one way or the other, so no idea where that came from.

by the time I'm ready for another machine, a GPU upgrade would be silly because even if it was upgradeable, then rest of the machine would be far to slow to really support it and make the most of it. I've been down this path upgrading Mac towers until the bitter end. I probably had 3 generations of nV cards in the machine, but the slower bus and ram is what really crippled it in its later years even if the CPU spec was still decent. so the belief you can upgrade all the parts one at a time here and there is a fallacy when critically analyzed to show its of limited value at best depending on exact circumstances.

S
S
mac Studio Ultra 128gb OS 14.5+
TB3 ext 1TB SSD Cache, 1TB SSD Sys
4K 27Dell
Presonus Quantum 2626 TB3 Word clocked 96K
MiniPanel via enet
Behinger Xtouch USB via USB3 Hub->PCIe USB3X4 card
Wacom XL Pro
Offline

SkierEvans

  • Posts: 1277
  • Joined: Wed Jan 24, 2018 9:59 pm
  • Location: Ottawa, Ontario
  • Real Name: Ron Evans

Re: Mac Studio performance compared to 40xx GPUs

PostSat Oct 07, 2023 1:23 pm

For information to this thread. About a year ago I wanted to upgrade my PC , then with a 1080Ti, to a 3090. Of course if you all remember there were no GPU's on the market then other than buying a ready made PC. However the list price at the time was about $2600 Canadian. The Studio Max was about $2400 Canadian. As to upgrade or repair myself neither of these fit that action ! So I bought the Studio Max which I got in 2 weeks. My first Apple product ( since got an iPad M2 12.9" ). Intent was purely to run Resolve easily for my GH6 and GH5S files in multicam that were too much for my PC. To this day I only have Resolve, Topaz Ai, Chrome and Affinity Photo 2 installed. It's main value for me is in the hardware decoders/encoders and the fact it is silent. I have since upgraded my PC with a RTX 4070Ti and useful now to compare performance. Of course the PC decode has not improved, however encode has very much. It is still slower overall for a 2 hour video of course because of all the need for decode however watching how the performance changes with the different source as it encodes is interesting. The PC is about 15% slower dealing with the GH6 files but 25% faster with Fusion titles or the other camera files. Both encodes are considerably faster than realtime. Minutes difference in fact between them. I have taken to editing on the Studio Max but encoding on the PC as the rest of the processing for DVD or BLuray etc is on the PC anyway.

I may as well add why I got the iPad. Initially to use as a clean feed monitor from the Studio Max ( fail ! ) and also to use with my TV to check HDR colour etc with Airplay ( success , was one of the other reasons to get iPad as well was Airplay other sources) I cannot justify a calibrated HDR monitor for my hobby or have the space for it. My HDR TV is a better solution as for HDR I am the client. I have my source on external SSD anyway so move to iPad rather the PC ( same files and project of course ) first run actually looking at the edit on the iPad screen having changed colour management output from rec 709 to HDR. Then cross check using Airplay to the TV as clean feed from iPad. This approach has worked well.
Threadripper 1920, Gigabyte X399 DESIGNARE EX, 32G RAM, Gigabyte 4070Ti 12G, ASUS PB328Q, IP4K, WIN10 Pro 22H2, Speed Editor

Resolve Studio 19, EDIUS 9WG,EDIUS X WG, Vegas 18

Studio Max M1 24 core GPU, 32G, 1T drive. iPad Pro 12.9` M2 16G, 1T
Offline

John Spirou

  • Posts: 125
  • Joined: Sun Sep 17, 2017 3:51 pm

Re: Mac Studio performance compared to 40xx GPUs

PostTue Oct 17, 2023 5:44 pm

Michael Haggard wrote:No benchmarks, but I have a Mac Studio M2 ultra 128GB standard cores and it doesn’t perform half as well as I would have excepted.

I’m editing 4K footage on a 25fps HD timeline and a single node of Resolve’s noise reduction on ‘better’ setting drops me to 5-10fps on playback. Without noise reduction, a few masks with blur nodes will do the same.

If I bypass effects, I can playback the footage without dropouts. Cloning on fusion is another heavy operation that will drop me below 25fps.

Not the workhorse I was expecting. It also doesn’t play well with non-Apple external displays.
I have seen reviews that shows M2 Ultra studio that plays in real-time 4k with noise reduction… what kind of footage do you use?
Offline
User avatar

Uli Plank

  • Posts: 25430
  • Joined: Fri Feb 08, 2013 2:48 am
  • Location: Germany and Indonesia

Re: Mac Studio performance compared to 40xx GPUs

PostWed Oct 18, 2023 2:59 am

Well, I don't have that maxed out Ultra M2 anymore, which was a loan for reviewing. But my humble laptop (see sig) is doing NR in a UHD timeline at 20 fps for temporal NR only and around 10 fps with some spatial added. This is HEVC from a drone in UHD. No reduced TL proxy resolution.

So, I don't doubt that a Mac Studio M2 would do realtime.
My disaster protection: export a .drp file to a physically separated storage regularly.
www.digitalproduction.com

Studio 19.1.3
MacOS 13.7.4, 2017 iMac, 32 GB, Radeon Pro 580 + eGPU
MacBook M1 Pro, 16 GPU cores, 32 GB RAM, MacOS 14.7.2
SE, USM G3
Offline
User avatar

joema4

  • Posts: 435
  • Joined: Wed Feb 03, 2021 3:26 pm
  • Real Name: Joe Marler

Re: Mac Studio performance compared to 40xx GPUs

PostWed Oct 18, 2023 3:42 pm

Yasser Saeed wrote:...I know many friends who own in their studios both Mac and Windows workstations and they often confess to me that the reputation of Mac stability advantage over Windows is not true at all as their Mac computers crash as much as Windows and sometimes even more...Keep in mind they are using reputed brands of Windows workstations that costed them as much as thier Macs and this is the fairest way to compare Windows to Mac. It is not fair to compare a Mac computer that costs 6K to unbranded and custom build PC that costes 2K....that is why it is important to consider Windows workstations from reputed brands such as HP, DELL and BOXX who design most of hardware, fully test it making sure all components work together at maximum compatibility and reliability...


I totally agree. I currently have six Macs and have lots of experience at both MacOS and Windows system-level debugging, and I formerly did digital hardware design. I like many aspects of MacOS, but IMO my M1 Ultra Mac Studio is no more reliable than a professionally built Windows workstation.

This is likely due to several things:

(1) MacOS is not tested with the same rigor as Windows. You can't run SQL Server or other high-end multithreaded apps on a MacOS server, because that doesn't exist. That kind of real-world stress uncovers various difficult transient bugs, such as multithreaded race conditions. Those fixes are rolled into all Windows platforms.

(2) The Windows debugging and support tools for an *operational* scenario are vastly better than MacOS. Compare Windows PerfMon to MacOS Activity Monitor. PerfMon can monitor hundreds of parameters, show graphs, tables, record those to a file, do all that on a remote machine, etc. E.g, if you want to monitor disk queue depth or interrupts per second, PerfMon can do that. By comparison Activity Monitor is a toy.

MacOS formerly had the command-line DTrace tools for monitoring but those have been steadily deprecated, require disabling System Integrity Protection, and even then many don't work.

NT architect Dave Cutler came from Digital Equipment Corp, so had great sensitivity to real-world operational conditions. He mandated that high-quality monitoring and support tools be part Windows NT from the very beginning. Those are still there and more extensive than ever.

By contrast Apple's focus has traditionally been on application developers. Apple takes good care of those application developers, and the XCode Instruments tracing tool is excellent. However it is designed only for app developers, and is not suited for troubleshooting an operational, deployed system, especially a remote one.

(3) One cause of reliability issues with Windows comes from self-built machines. Those can work well, but they obviously don't undergo professional integration testing. When Dell or HP makes a professional workstation, they don't just slap a motherboard in a case. They have trained engineers using high-end instrumentation to debug transient or narrow problems. Those engineers find issues, then interact with their counterparts within their company or at other companies. I spent years sitting in front of a Tektronix logic analyzer doing this: https://joema.smugmug.com/Computers/Tek ... -52kXRLG/A

Even given a studious approach, a single person can only do certain tests on a single machine. A major manufacturer can do long-duration automated, instrumented tests on hundreds of machines simultaneously. Apple does those tests, Microsoft does, the major PC workstation manufacturers do, but a single person just can't do that.

That may be why it's much more common on this forum to see posts from people with hardware and reliability problems on a self-built Windows PC. It's great this forum is so welcoming and helpful on hardware and Windows system-layer issues, but those people really should be getting system-layer help from a place more targeted to the problem area. If the PC is crashing, there is no software fix in Resolve to address that. Sometimes such situations happen on Macs, but it's much less common. You rarely see people with a Dell, HP or Puget Systems professional workstation asking for hardware and system-layer troubleshooting help here. In that sense, those machines are similar to Macs.
Offline
User avatar

Rakesh Malik

  • Posts: 3378
  • Joined: Fri Dec 07, 2012 1:01 am
  • Location: Vancouver, BC

Re: Mac Studio performance compared to 40xx GPUs

PostWed Oct 18, 2023 9:17 pm

joema4 wrote:I totally agree. I currently have six Macs and have lots of experience at both MacOS and Windows system-level debugging, and I formerly did digital hardware design. I like many aspects of MacOS, but IMO my M1 Ultra Mac Studio is no more reliable than a professionally built Windows workstation.


Corporate IT folks usually detest macs, not because they're less reliable than the Dell, HP, Lenovo, etc workstations that are more common, but because they're not any more reliable... and support is much more clumsy. You have a hard drive die on your Dell NAS, you call a toll-free number and it SHALL be back to 100% operation the same day, even if you call at 2am, because that's in the contract. That's something Apple didn't figure out.

(1) MacOS is not tested with the same rigor as Windows. You can't run SQL Server or other high-end multithreaded apps on a MacOS server, because that doesn't exist. That kind of real-world stress uncovers various difficult transient bugs, such as multithreaded race conditions. Those fixes are rolled into all Windows platforms.


Yeah, I think a lot of people have no idea how much data Microsoft gathers just from automated bug reports alone, never mind the dedicated testing systems.

NT architect Dave Cutler came from Digital Equipment Corp, so had great sensitivity to real-world operational conditions. He mandated that high-quality monitoring and support tools be part Windows NT from the very beginning. Those are still there and more extensive than ever.


I've always preferred UNIX-ish environments (except Irix, which I detested due to the bug-ridden documentation) from a usability point of view, but there's no arguing the reality that VMS and MVS dominated the high-reliability niches for decades for a reason. Other than Nonstop Unix, there were no UNIX-like OSs that were even close.

By contrast Apple's focus has traditionally been on application developers. Apple takes good care of those application developers, and the XCode Instruments tracing tool is excellent. However it is designed only for app developers, and is not suited for troubleshooting an operational, deployed system, especially a remote one.


Actually, it's more accurate to say that Apple's focus has mostly been on APP developers. Mobile computing is what remade Apple after the G5 got steamrolled by x86. Without mobile computing, Apple Silicon would not exist at all right now.

That may be why it's much more common on this forum to see posts from people with hardware and reliability problems on a self-built Windows PC. It's great this forum is so welcoming and helpful on hardware and Windows system-layer issues, but those people really should be getting system-layer help from a place more targeted to the problem area. If the PC is crashing, there is no software fix in Resolve to address that. Sometimes such situations happen on Macs, but it's much less common. You rarely see people with a Dell, HP or Puget Systems professional workstation asking for hardware and system-layer troubleshooting help here. In that sense, those machines are similar to Macs.


Yes, that is quite true. Windows machines from reputable manufacturers are every bit as stable as macs, and the fact that there are typically even more technical questions from mac users on most forums reflects that... particularly since the gaming and VFX world are VERY heavily dominated by x86, most of the time with Windows at the workstation and Linux on the renderfarm, though the on-prem renderfarms are steadily being replaced by cloud based renderfarms like AWS.

But then a lot of people blamed Black Magic for its cameras being unreliable because they bought inadequate media instead of approved media, so that's just par for the course. The bug was the media, not the camera, but some people just refused to take Black Magic's advice, and that's what you get.

That has always been Apple's real secret; there are no DIY macs anywhere, yet the reliability comparisons always emphasize the DIY windows machines instead of reputable ones.

In fact, I only ever retired one computer because it was simply unreliable rather than because I had quite literally outgrown it, and that one computer was a mac. Every other computer upgrade I've made has been driven by a need for more computing power, and no other reason. I'm itching to upgrade my current machine for the same reason; it's plenty reliable, but I need more horsepower for Houdini, Redshift, and Terragen.
Rakesh Malik
Cinematographer, VFX Artist, photographer, adventurer, martial artist
http://WinterLightStudios.ca
System:
Asus ProArt 16/64GB/12 core Zen5/nVidia RTX 4070 8GB
Nuke/Houdini/Resolve
Offline
User avatar

Yasser Saeed

  • Posts: 736
  • Joined: Tue Feb 16, 2016 3:46 pm
  • Location: Egypt

Re: Mac Studio performance compared to 40xx GPUs

PostWed Oct 18, 2023 10:18 pm

Thanks Joe Marler and Rakesh Malik for explaining in much more details than I ;)

My 9 years old AIO HP Z1 workstation was working great but 5 years later I replaced it with HP Z2 Mini G5 workstation just becouse needed more power .. and soon I will be replacing my Z2 Mini G5 with Z2 Mini G9 just becouse I also need more power. Even my first HP workstation, the Z600 which I purchased in 2009 is still working flawlessly. I truly never faced any hardware problems with any of my HP workstations that is why I always upgrade from the same brand but I am sure other reputable workstations brands like Dell and Boxx will be as good as HP.

When dependability and reliability is the most important factor in sensitive fields like space, medicine, military, research, manufacturing etc... do you see them rely on Macs? From what I can see, it is either Unix or Windows or both, but never Macs. Why?
HP Z2 Mini G9 Workstation
OS: Windows 11 Pro
Display: HP Z27 QHD DreamColor
CPU: i9-13900K 24 core
GPU: RTX 4000 SFF 20GB Ada Gen
RAM: 96GB DDR5-5200
NVMe SSD: OS 4TB, Data 4TB
NLE: DaVinci Resolve Studio 20 Beta
Offline

Dejan Špagnut

  • Posts: 84
  • Joined: Sun Jun 21, 2015 2:45 pm
  • Location: Belgrade, Serbia

Re: Mac Studio performance compared to 40xx GPUs

PostWed Oct 18, 2023 11:16 pm

Yasser Saeed wrote:When dependability and reliability is the most important factor in sensitive fields like space, medicine, military, research, manufacturing etc... do you see them rely on Macs? From what I can see, it is either Unix or Windows or both, but never Macs. Why?


In most cases that has nothing to do with reliability. Many organisations prefer to be able to buy the same necessary equipment from multiple sources or manufacturers in order not to be dependent on the single source or manufacturer for that equipment. You simply can’t do that with Macs. Only manufacturer is Apple, therefore that disqualifies them fast in these situations.
Offline
User avatar

Rakesh Malik

  • Posts: 3378
  • Joined: Fri Dec 07, 2012 1:01 am
  • Location: Vancouver, BC

Re: Mac Studio performance compared to 40xx GPUs

PostThu Oct 19, 2023 12:38 am

Yasser Saeed wrote:When dependability and reliability is the most important factor in sensitive fields like space, medicine, military, research, manufacturing etc... do you see them rely on Macs? From what I can see, it is either Unix or Windows or both, but never Macs. Why?


One major factor is that Apple does not understand enterprise. I attended a presentation by an Apple dolt who was showing us Apple's NAS solution and couldn't figure out why no one wanted it, even though it was less expensive than the far more common Dell/EMC NAS solution.

He was looking at the contract price, but ignoring the contents of the contract -- the Apple price was for hardware. The Dell/EMC price was for hardware, network and power buildout, installation, configuration, and five years of same day on-site tech support.

For anyone buying a NAS, that was a no-brainer; it was blindingly obvious that you could not trust the Apple solution, because Apple wouldn't stand behind it.

Apple is a phone manufacturer. It's not an enterprise services provider. Dell, HP, Lenovo, etc are enterprise service providers.

In the real world, nothing is perfect, and when you have an organization with hundreds or thousands of computers that have to work, Apple is a complete non-starter because the Genius Bar is not a viable solution, or even close.
Rakesh Malik
Cinematographer, VFX Artist, photographer, adventurer, martial artist
http://WinterLightStudios.ca
System:
Asus ProArt 16/64GB/12 core Zen5/nVidia RTX 4070 8GB
Nuke/Houdini/Resolve
Offline
User avatar

Yasser Saeed

  • Posts: 736
  • Joined: Tue Feb 16, 2016 3:46 pm
  • Location: Egypt

Re: Mac Studio performance compared to 40xx GPUs

PostThu Oct 19, 2023 10:56 am

Dejan Špagnut wrote:
Yasser Saeed wrote:When dependability and reliability is the most important factor in sensitive fields like space, medicine, military, research, manufacturing etc... do you see them rely on Macs? From what I can see, it is either Unix or Windows or both, but never Macs. Why?


In most cases that has nothing to do with reliability. Many organisations prefer to be able to buy the same necessary equipment from multiple sources or manufacturers in order not to be dependent on the single source or manufacturer for that equipment. You simply can’t do that with Macs. Only manufacturer is Apple, therefore that disqualifies them fast in these situations.


Well, I think in sensitive fields it has to do with reliability and dependability becouse so many big organisations generally prefer to stick with one or two reputable brands. If your point was valid, then you would see such organisations buying from so many brands (including Macs) whether reputable or not for the sake of not depending on a single source!
HP Z2 Mini G9 Workstation
OS: Windows 11 Pro
Display: HP Z27 QHD DreamColor
CPU: i9-13900K 24 core
GPU: RTX 4000 SFF 20GB Ada Gen
RAM: 96GB DDR5-5200
NVMe SSD: OS 4TB, Data 4TB
NLE: DaVinci Resolve Studio 20 Beta
Offline

Dejan Špagnut

  • Posts: 84
  • Joined: Sun Jun 21, 2015 2:45 pm
  • Location: Belgrade, Serbia

Re: Mac Studio performance compared to 40xx GPUs

PostThu Oct 19, 2023 11:51 am

I think that you have overly romantic view on all of this :D

NASA buys what is good for NASA, you should buy what is good for you. Why would you in video postproduction even care what a hospital or space agency buys? And yes, big organisations change vendors all the time, for reasons you would never imagine.
Next

Return to DaVinci Resolve

Who is online

Users browsing this forum: Bing [Bot], fndpena, panos_mts and 266 guests