[RESOLVED] NVIDIA-graphic chips confusion

Support forum for DGDecNV
Post Reply
DAE avatar
Jagutherrschaften
Posts: 10
Joined: Wed Nov 03, 2010 6:04 am

[RESOLVED] NVIDIA-graphic chips confusion

Post by Jagutherrschaften »

Hey.

I've read several threads in this forum and Doom9's as well but I am still not completely sure about some questions:

I see the GT 610 (Fermi chipset) performs as good in the benchmarks as more expensive Kepler-based cards like the GeForce GTX 650 Ti for example.

1) Now, does that mean I can get a GT 610 and won't have any disadvantages when encoding video with megui like this: I demux TS with tsMuxer, open demuxed (something.264) file with dgindexnv and save index file, open avs script creator and open dgi-file, ...some steps like cropping etc...., encode video using x264 (automatic 2pass)?

2) I am sorry if this is a dumb question, I just couldn't find the info I need: The new Kepler-based cards have more CUDA shaders than the Fermi-based cards. Does this make any difference?

Conclusion: If I want to have the best solution for encoding x264-video the way I described above, would I get a GT 610, a GT 640 (making sure it's a version with a Kepler-based chip) or an even higher model like the GeForce GTX 650 Ti? As I said, I read the benchmark-thread, but I am not completely sure if the benchmark covers all kinds of jobs a video card has to do when really encoding a video (and maybe deinterlacing it).

Thanx a lot for any helpful answer. I am enjoying this way of encoding videos a lot, I started working with videos a long time ago when it was normal to see the PC run over night to encode a simple SD-recording to divx :lol:.
User avatar
admin
Posts: 4551
Joined: Thu Sep 09, 2010 3:08 pm

Re: NVIDIA-graphic chips confusion

Post by admin »

As far as video decoding is concerned, it is only the VP engine version that matters. VP5 is on both the 610 and the 650, so the performance will be very similar.

I have not performed any testing or benchmarks to know if the deinterlacing performance differs between the two. Actually, I do not know if the post-processing is all performed in the VP engine. I suspect that it is, however.
DAE avatar
Jagutherrschaften
Posts: 10
Joined: Wed Nov 03, 2010 6:04 am

Re: NVIDIA-graphic chips confusion

Post by Jagutherrschaften »

Thank you for your answer. If you suspect right this would mean no difference at all in encoding video right?

What made me stumble was this post: http://forum.doom9.org/showthread.php?p ... ost1530397

And the quote:
Despite the GT 520's advanced VPU engine, the lack of shaders limits its post processing capabilities. With all post processing options enabled, the GT 520's GPU load was always between 60 and 80%. The memory controller load (DRAM bandwidth usage) was between 20 and 40%. Despite the headroom apparently available, NVIDIA indicated that there weren't enough shaders available to implement the more advanced deinterlacing algorithms.
User avatar
admin
Posts: 4551
Joined: Thu Sep 09, 2010 3:08 pm

Re: NVIDIA-graphic chips confusion

Post by admin »

I just don't know at this time. We need people to benchmark the deinterlacing to be sure.
DAE avatar
Jagutherrschaften
Posts: 10
Joined: Wed Nov 03, 2010 6:04 am

Re: NVIDIA-graphic chips confusion

Post by Jagutherrschaften »

OK, thank you. If I buy a new card soon, I am available for running some benchmarks if needed.
Post Reply