@MarcinKrk

Hi

. . . . AMD vs Intel CPU, NVidia vs AMD GPU . . .

In general terms, the combination I would recommend depending on how much you want to spend is:

• Low end - Intel CPU with integrated GPU with 6 to 10 cores*, and a Nvidia RTX 2000 series GPU.
   
• Top end/future proofing - Intel CPU with integrated Iris GPU that is ARC compatible with an Intel ARC GPU** or Nvidia RTX 3000/4000 series GPU

* More cores than this does not give sufficient performance increase for the relative cost increase

**Magix are focusing on Intel and their Hypercode feature which works with the Intel ARC series of GPU however it is a bit of an unknown at the moment - see here - scroll down about 1/3rd of the page

I would avoid the following:

• Intel Xeon processors - they are not suitable for video editing.
   
• AMD CPU and GPU systems. IMHO they have always struggled against Intel and Nvidia for video editing applications.

HTH

John EB
Forum Moderator

@MarcinKrk @Former user @johnebaker

In general I agree with John Baker with one exception. The 2000 series of nvidia cards are now horrendously expensive except if you can find one in a prebuild. If you are looking for a new GPU that will give you better value with similar or better performance, that can be had from some 3060 and above GPUs IMHO but that is another minefield in it's own right with current higher designated titles now performing worse that some lower sounding titled GPUs depending on the use it is put to.

As john points out, spending beyond a certain point is an exercise in diminishing returns but that may vary from user to user as to where that point is depending on the individuals needs. Not everyone has the same needs or requirements. For future proofing though, that is tougher to predict, but for stability I personally would choose components one generation behind current offerings but at the higher end of components if I could afford to. Even that route is currently quite expensive and don't forget to factor in adequate power supply and storage and possibly a new monitor to actually see what you are editing better if it has already served you for a long time.

Because of the variation of use I find benchmarking somewhat subjective. It is perhaps interesting but often doesn't reflect real world usage.

Personally I find the more my system can do the more I push it and eventually I do find the limit of the system. I think that would be true of any system I had simply because I like to push the limits.

Ray.

@MarcinKrk I can't explain but from what I've read I sometimes wish I'd bought an Intel CPU, as far as GPU is concerned I think that all depends on the software, each one of my software's use my hardware in different ways & amounts.

Nowadays programs concentrate more on using the GPU & RAM as apposed to using the CPU for performance, because those 2 , GPU & RAM, are easier to upgrade at a later date, so buy the best you can afford, or a good basic motherboard & CPU with the intention of upgrading the GPU & RAM at a later date.., 🤷‍♂️.

PS @CubeAce Yeah a PC is never fast enough, my machine isn't fast enough but compared to most... It's hard to be satisfied with these things. the grass is always greener 🤣

Thank you all for your suggestions! It's really impressive to see all the answers coming back from you in no time!

The clear conclusion is: go for Intel + Nvidia, rather than AMD + Radeon. Also, hardware is never good enough, so despite of how much you spend, you'll still strive for something better. I feel I've been there already, too... few times ;-)

And finally - there's no Movie Studio 2023 benchmark available and looking at the benchmarks for other MAGIX products might help to make some decisions, however might not be accurate - as each of them uses different internal engine / solution.

Just in case someone is up for Benchmarking MS2023, drop me a line - I'm happy to contribute with my results ;-)

@MarcinKrk

Hi Marc.

If you are really interested in a sort of benchmark test than have a look at this topic I created some time back. That in turn links to a prior test that topic is based on.

I still have the project files if you wish to compare your own current system but as you plough your way through the topic you will find there are a lot of variables to take into account and the reason my personal interest has waned somewhat.

Personally I think component choices at the moment are difficult due to price fluctuations, the ridiculous price currently for virtually all components and the choices when thinking of which manufacturer to go with. Nothing is concrete.

These are the only comments I can find from Magix on system building and seem very generic.

Ray.

@CubeAce

Hi Ray,

Your first link goes to a non-valid source http...I still have the files. Please check. It would be good if we had a test project that we could run for benchmarking or even just checking to see if users with problems can run it.

@MarcinKrk

Hi,

See my signature for the specs on my desktop, purchased last year. As Ray mentioned, there are a lot of variables.

Here are a couple:

1. I use VPX on two monitors, one full HD, the other 4K. VPX often has problems with playback because of this. Switching to 1 monitor resolves this. I haven't tried on two FHD monitors but I should. I have 2 FHD and 1-4K on my system.
2. Frame rate differences - I'm on NTSC, so 29.97 or 30 fps. Using any Magix templates using mxv files can cause problems due to different frame rates.

As you have found out, Intel/NVIDIA is the preferred set up, with the Intel chip including integrated UHD graphics. I had to argue with the salesman/technician about this as they wanted to supply the wrong Intel chip - no integrated UHD graphics.

John CB

@browj2

Thank you for the heads up on the link. It should now be sorted.

The folder that you still have. Is it 1.59GBs in size? I'm just checking to see if I have changed the copy I have found as I have two folders but the second template folder has a video version as well. If in doubt you can download the version I have just used.

I can make the project folder available to anyone who PMs me with an email address that either wants to have a go and wants to add their findings as this test would be conducted on the newer versions of the programs than the original test.

I have just run the project in MEP 2022 with a 4K 50fps MP4 export standard rendering set to Best.. Nine minutes fifty nine seconds. Ouch!

Ray.

@CubeAce

Hi Ray,

I don't have a folder like that. I don't recall downloading that. I have other ones. You can send me a PM with the link and I'll test it out.

John CB

@browj2

Hi John.

Sorry, you confused me with "...I still have the files".

It was the 'Two Minute Tech Tips' project.

I will sort out some space on my Google Drive after I have verified all the people I've been sending stuff to have downloaded their files. It's been a busy week. I should have everything sorted within 12 hours.

Good idea, I never thought about sending links via the PM system.

Ray.

@browj2

I can make the project folder available to anyone who PMs me with an email address that either wants to have a go and wants to add their findings as this test would be conducted on the newer versions of the programs than the original test.

I have MEP2022 and MEP2023 so I could check my current setup in both, if you could share the link to MEP project.

@MarcinKrk

Hi Marc.

That would be me with the project files.

I will make the original files we used in 2019 available again but I think this time around I will make terms and conditions for the benchmarks more stringent with a verification section. Also It will be for all versions of past and present copies of MEP and VPX as well as for the newer rebadged version of MEP. That way eventually people may be able to see if an upgrade may give them an advantage or worth doing. Anyone will be able to take part that can be bothered to adhere to the terms and conditions. I will set up a new topic in about another five hours for benchmarking alone but will contact the mods first to see if this can work or is desirable as it may need some mod work to keep the topic clean and free from anything other than the results.

@Former user @johnebaker

Ray.

@CubeAce I'm up for it, you'll need to keep the export settings as one of the default settings so it's easy & everyone's the same,

Are you going to determine how a project plays back on the timeline, as far as i know the only indication on MS's screen is the little CPU indicator in the bottom left? Vegas shows an FPS at the bottom of the preview screen. 🤔

@Former user

Hi Gid.

Yes to most of the above but I'm behind as we have unexpectedly had to look after our grandson overnight.

Should have everything sorted in the next two hours.

Ray.

@CubeAce Ah, no rush, take care of family 1st 👍

@MarcinKrk @Former user @browj2 @johnebaker

Rules and application requirements are now posted here.

Please read thoroughly.

Ray.

So the request is made in this thread and queries are made here?

In the other thread you only announce the results.

Hi Marc.

Correct. Thank you for complying. Sending link.

Ray.

@CubeAce

hello, I never work with proxies and if possible without preview rendering.  The options are always turned off for me.  When I load your project, however, the proxies are activated by themselves in the film settings.

No proxies have been created yet, but the option turns on when loading your project file.

@MarcinKrk

Hi Marc.

In your instance just ignore the prompt when you open the project and then go to the Project movie settings and untick the Automatically create proxy files settings. Then just save the project at the end. It should not generate any additional files. You may find the settings are also set to play movie in loop mode and use settings as presets for new movies. You can untick those as well if you wish. I just ignore the prompt at startup. Once ignored on my system I don't get the prompt again even though the box remains ticked.

Ray.

@CubeAce As you only want results on that other page I'll post this here,

 I'm going to clean my PC, but it's the graphics that really slow it down for me,

AVC & HEVC roughly the same - 6.39mins, 🤔😒

PS. playback is stuttery on the start and end graphic/text sections,

@Former user

Hi Gid.

Those are valid results as far as I can see. Did you look at the results I posted?

The images are very interesting. Look how little work your CPU is doing and how much vram your card is using. Even the amount of motherboard ram in use is way higher than my own experience.

The times are still faster than my own thanks in main I think to your GPU. Playback is the same on my machine.

The idea is to give people looking at the results some idea on what to base their next system build on. No result is a waste. It is not a competition but a comparison data base on how the base program without third party effects involved runs and exports.

Ray.

Hey @CubeAce,

Can you share the link to the benchmark project? I think I'm ready to run some tests now (I've added my PC spec to my Profile ;-)