NVENC vs X264 with RTX 2080ti - Streaming & Gaming

The new NVIDIA RTX cards have an improved version of NVENC onboard - but does it hold up against the 'gold standard' of X264 encoding on your CPU? Let's test out the performance differences, and video quality difference and see if it's worth using!

As an Amazon Associate I earn from qualifying purchases, using the link below or other Amazon affiliate links here.
Bitcoin donations: 1PqsJeJsDbNEECjCKbQ2DsQxJWYqmTvt4E

#TechTeamGB

About TechteamGB:

TechteamGB is a long-running tech channel focused on high quality videos on PC Hardware and Consumer Electronics. We work with nearly all of the big (and small) names in technology, from Intel and AMD, to MSI, Asus, Gigabyte and more! If you have any questions, feel free to leave a comment! If you are a tech brand with some new and exciting products, email Andrew at the address below - please include details of the products you would like reviewed/showcased. Also, if you are a brand looking to sponsor/advertise, please make this clear in your email. Otherwise, enjoy the videos! Oh, and don't forget to leave a comment on the first video of mine you watch and say hello - I'd love to chat with you!

NVENC vs X264 with RTX 2080ti - Streaming & Gaming

TechteamGB

[divsta]

I think it's worth pointing out NVENC is only really truly viable when gaming at 1080p on a 1080 Ti/2080/2080 Ti. At this point you're CPU bound and will rarely see your GPU at 99% usage therefore there's a lot of headroom for NVENC. You'll still be CPU bound on a 2700X/9900K (more so on the 2700X which leaves a little more headroom for NVENC for a little less FPS ingame).

When using 1440p, 1440p ultra wide or 4K, you'll be GPU bound with very low CPU usage so it makes more sense to stream using x264. You'll generally find because your card is hitting in excess of 95% usage at these resolutions in most AAA games that NVENC streams will suffer from massive amounts of render lag.

It's very important to understand what being CPU/GPU bound is and how important FPS 0.1%, 1.0% and minimums are when it comes to streaming.

It's possible in theory to play for example Battlefield 5 at 1440p. At this point you're GPU bound as the GPU is at max usage and there's headroom on the CPU. Let's say your average FPS over the course of 30 minutes on a full multiplayer server is 140 and your minimum is 90. You could cap your FPS at 100.
Your card will hit the 100 frames and then relax, having time to do NVENC. It's not having to render the additional 40 frames you'd normally have on average so it has time for NVENC now.
Now let's say 20 grenades explode, a tank runs you over and there are planes overhead. Your FPS drops to 90. Your card is now trying the best it can to still pump out the 100 frames but it can't and it's giving it the absolute everything. On top of that you now want it to do NVENC? Not gonna happen! The stream will stutter and lag.

1080p is very easy for the 1080 Ti, 2080 and 2080 Ti and will run most AAA games in 2018/2019 in excess of well over 100 FPS when paired with a decent CPU so NVENC, especially the new NVENC implementation released with OBS 23.0 (which promises up to 60% less of a performance hit) is a very viable solution.

Unfortunately if you plan on playing anything competitively or just want the smoothest experience, the best option is still to build a dedicated computer for streaming and using a capture card to receive a clone signal from your gaming PC.

[goodfella]

why "very fast" preset? Nvidia says their Turing nvenc is better than x264 "fast". Thats what should be compared

[Cruizerman]

What do you guys think is the best solution for me to stream with? AMD Ryzen 7 1700 OC to 3.8GHz or my MSI GeForce RTX 2080 Duke OC? Atm i am trying to stream with 2560x1440 Downslaced to 1920x1080 60FPS with 8000KB/s (Faster on CPU and High Quality on GPU)

[obese]

Did you do each test pass manually, or did you use the replay feature of each game?

[streakurt]

I have rtx 3070 do will I need upgrade my 5600x to 5800x for core if I stream through nvenc ? Will 5800x give me more fps while streaming or won't be there any difference ? Because 5600x and 5800x almost same fps in games but don't have idea about streaming due core difference ?

[artishere]

i have R7 3700x with 2070 16gb but also have R5 3600 16 no GPU should i get 1660 to stream or 1050 ti and use r5 to stream

[melamauri]

What would you recommend if you would have 9900k and 2080ti?

[emilefadel]

Great video man, I’d be keen to see how NVENC fares against a fast or medium preset, as it stands right now I’d say it’s definitely a contender against very fast no doubt

[chlonex]

if you have a 9900k, do you need your graphics card to handle a stream? or would a 2080ti NVENC look better?

[wakeUK]

Really important question: is the new RTX NVENC better or the same as dual PC set-ups? I am a daily streamer. I need perfect frames, and would prefer a single-PC setup. But at the moment RTX is looking interesting.

[bigtrain]

I have dual liquid-cooled rtx2080ti's is there a way to set one of my video cards in obs so that card only those the streaming

[leexgx]

I assume they was at same bit rate you norm need at least 1000 more on nvenc to match images quolity (should be noticeable at 60fps, but I don't recommend 60fps for twitch unless you have twitch transcoders for quolity options )
--- OK you had it set to 6000 so very likely you not see muxh difference between x264 and nvenc, I am on mobile so can't tell (I find a lot of people actually mess up in the source capture side and make it blurry or have blocky Because they can't have the speed preset high enough)

[goodfella]

can you compare nvenc Turing vs pascal?

[DJHWTY]

B is NVENC. Smoother (looks like actual 60fps, rather than semi-choppy 50-55fps) and also the "buffer blocks" or whatever they're called are bigger squares in the 'B' image, which suggests a GPU has encoded the video rather than a CPU.

[cliche_AF]

Should try this again but gaming at 1440p so the CPU has even more legs. Also why would someone stream using the same process makes more sense to stream using x264 and capture using NVENC

[ZakkPanzermensch]

Does anybody knows is a difference between NVENC performance in RTX 2060 & GTX 1660 Ti?

[skymasterptgaming]

I've done the other day a test, streaming in x264 and NVENC ( i have a 1080ti with a Ryzen 1600, None overclocked) and i notice a huge difference while playing with highest GFX between both, NVENC is a much better codec for my personal experience while x264 kinda lowers a lot my gaming experience... ALthough i was having a friend watching the stream and he said that both were very good (i stream at 1780x720p 60fps (i have a ultrawide monitor and stream at that conversion), with 20mbps upload so i use 4000 bitrate)

[Zrksys]

why not 8000??? u and hardware unboxed both, u get affiliated in less then 2 weeks, especially if u are a someone with that many people from youtube. And oh yea, why recording if most of us wont use the recording, just streaming encoding?And by that i mean why not to see how good it would be for us without stressing the cpu or the gpu for recording, i understand u want to show us how it would look for the viewers.

[retakleksum]

Veryfast preset and x264 still Looks better. Test fast or medium and even you will see how much better x264 Looks.

[fidgetyjester]

I don't use Mac but I heard they have something better than x264 called Apple Prores 422. So Apple Prores 422 vs NVENC? Lol XD