-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
7950x onboard graphics?
- Thread starter subbytna
- Start date
I would disable it, no need for the on board IGPU if you have a dedicated 1.
Normally you would disable it in the bios and then it does not appear in Windows at all.
I have it disabled on my 9800X3D
Man of Honour
- Joined
- 23 Mar 2011
- Posts
- 18,750
- Location
- West Side
I always disable the onboard graphics if I have a dedicated GPU , stops programs asking which graphics to use.
Is there some sort of toggle taht can be used? Mind if I ask how you did this and is it easy to switch between the two?
Open Windows settings and search for graphic settings:
In System-->Display-->Graphics you can choose which GPU is used for each app or browse for exe files. This works very well and you do not need to have a display connected to the GPU to use it.
I'm a little confused as to what you are doing here but even with the dGPU in its lowest idle power state + iGPU in use is going to be using the same or more power than just using the dGPU.
Generally the power use of the iGPU if enabled but not in use is very low to pretty much nothing, although on AMD there can be some complications with the IO die resulting in very slight extra power use.
Interestingly with newer nVidia GPUs even the high end ones if properly idling will use less power for the whole board than typical for an iGPU (at least on higher end setups - lower end systems can be another story).
Last edited:
Are you sure about that?
I haven't done any power testing with AM5 Ryzen IGP, but certainly with Intel, the IGP can easily beat higher-end nvidia GPUs on power for desktop use, even accounting for the idling nvidia GPU. That's especially with media playback where the GPU memory is hitting higher clocks. If the clocks were low, then perhaps not.
I'm generalising a bit - I added a caveat in there as there is a range of iGPUs from very low end low power Intel ones through to AMD RDNA based stuff which uses a fair bit of power at idle comparatively due to the integration with the IO die, etc.
I'm just testing it out and it does use less power for me so far.
HWiNFO64 values:
dGPU (Idle: ~10-16W):
Netflix in Firefox: ~40W
YouTube in Firefox: ~40W
VLC playing 1080p Video: ~40W
iGPU (IO die Idle: ~18-25W): IO die increase
Netflix in Firefox: ~10-15W
YouTube in Firefox: ~10-15W
VLC playing 1080p Video: ~10-15W
It might also help a little when gaming and playing a video on the second screen.
Last edited:
My 4080 Super uses around 7 watt idle, 13 watt average playing high-resolution video in YouTube and around 11 watt playing Netflix or 1080p video in VLC/MPC. The difference in power consumption at the wall is probably negligible once including conversion losses, etc.
Last edited:
Your 4080 video playback uses less than my 6900XT's idle. Anyway, using the iGPU drops the power a tiny bit for me and its easy to setup so why not.
Those are very low numbers for that class of card, I think you got a particularly good sample there!
There seems to be some oddities with the 4000 and 5000 series cards i.e. https://forums.overclockers.co.uk/t...power-draw-comparison.19004318/#post-37904584
There are loads which have power consumption like mine and a bunch that are higher/all over the place.
Several people have found multiple instances of the same model GPU i.e. 5070 when plugged into the same system have a range of ~50% difference in idle power.
Review on my GPU here https://www.techpowerup.com/review/gigabyte-geforce-rtx-4080-super-gaming-oc/41.html shows a fair bit higher power usage than mine - mine shows ~7 watt average measured with HW Monitor and IIRC 6.86 watt average at idle measured from the supply. Their 6900XT has 7 watt idle though video playback, etc. gets over 40 watt.
EDIT: In my case with the airflow the fans on the GPU never come on doing stuff like watching video even multi-monitor and 4K video, possible they were getting some fan activation especially if measuring idle after benchmarking, etc.
EDIT: In my case with the airflow the fans on the GPU never come on doing stuff like watching video even multi-monitor and 4K video, possible they were getting some fan activation especially if measuring idle after benchmarking, etc.
Last edited:
Yeah, it is complicated, results can vary a lot. There can be substantial differences if cards use different memory too, both at idle and with video playback. High refresh and multi monitor too, though that would apply regardless of which graphics is being used, except I'd expect a high idle to result in less delta between idle and video playback, because the memory clocks are already high.
Cheers Fred, been wondering if this can be done but never really looked into it too much. Now I have Chrome set to my 7800X3D iGPU and it works well, also did a test running OBS on the iGPU as OBS can use a lot of GPU and can cause stuttering in the preview that ends up on the stream itself, OBS runs great with this and its looking pretty smooth although I wish I had been able to test it with my 3060Ti as that card never really had enough grunt for Very High in-game settings and still run smooth in OBS at the same time.
And people always talk about DayZ being able to be ran on a potato so I just tested out this theory playing it with only the 7800X3D iGPU.. only on Poor settings is it kinda playable, was getting around 35fps-70fps in normal play, looking up into the sky I got to see 100fps
Much appreciated!
