Question / Help CPU Core count Vs. clock speed

[Makkusu]

I have a project and I been looking at x264 benchmarks to start making my mind about what CPU I would like to put in my dedicated streaming rig. I am thinking about a Xeon E5-2630 or an i7 5820/30k since that is in the budget that I want to spend on a cpu. The advantages I have on the E5 is at the end of the year I'll have a new budget to burn and pair it up with another 2630 to double my core count. With the I7s I can overclock them and that is cheaper and I don't say free because of the cooler that I will be using (Noctua NH-D14/15) but that is still several hundreds less than getting another cpu/mobo. I have 2 options for capture cards since I own a SC-512N1-L/DVI and a C985 Liver Gamer HD.

What I want is to be able to stream 720/60 and 1080/30 and provide the best quality possible and push encoding/compression as much as I can. The main factor here is quality and cost should not be a factor at all. I have enough upload to hit the recommended caps around 7 times, so I am thinking of streaming at 3500kbps or lower depending on my viewers.

Is there any true quality benefit (visible) on streaming at the resolution I want on a preset of medium or better and will how will 8 slower cores will compare to 6 faster cores (I will be pushing 4.2Ghz or better) or even 16 cores against 6 faster cores (this one I can guess, but I don't like guessing). I am also taking hyperthread out of the evacuation because its obvious is there.

I am not partnered with anyone but I do want to have a good quality withing my means. What I want to know is what will get me to encode like a boss and if I get to encode like a boss, will it make a visible difference to the viewers and how do these x264 presets matter, if they do at all for my application.

EDIT: forgot to mention I will not only use this PC to stream but also record. Not record while streaming but during game play from my main rig and consoles.

 

[Boildown]

I don't think anyone has ever posted the results of any kind of testing of this, at least not with OBS when I've been browsing these forums.

As far as will it be noticeable if you go all out and drop your x264 preset a few notches, don't fool yourself, the differences will be minor once you go past Medium. The law of diminishing returns kicks in pretty good. That said, no one that I know of has really quantified it with OBS. You might want to Google search for x264 results and hope that it applies.

I think two E5-2630s will be significantly faster. Not sure its "worth it" though.

 

[Makkusu]

That is why I ask if someone has this data or knows where I can find some sort of information on the matter. The first thing I did was look for x264 information, benchmarks, etc (like I mentioned in my original post) but I don't want to 'hope' it applies, that in a way would be fooling myself and ignoring your advice.

I have read in other posts about going below medium and also what it says on https://obsproject.com/forum/threads/technical-explanation-of-obs-settings.642/ about:

x264 CPU Preset
Changing this setting to improve quality is not recommended as the improvements are very minor and come at a great CPU cost. Going faster than veryfast will also disable many x264 optimizations and your stream quality will suffer.

My guess is as good as yours about the x2 E5-2630 in this application. I wouldn't mind bringing some data in but reading different mods and devs says that below medium is not really that much of an improvement makes it easier to make a decision for me as far as what I want to do. I want to keep my bitrate as low as possible with the best quality I can give within reason.

I will wait and keep researching for a couple more weeks. This factors in the decision I will make as far as the parts that I am buying for this workstation.

Thanks for your reply.