Question Why does vam make my gpu so hot

[nsfwthrowaway77]

gtx1070

vam main menu:
utilization 67%, temp 77C

in scene:
utilization 100%, temp 80C

playing a normal AAA game:
utilization 80%, temp 65C

Why it is so much hotter, even with less gpu utilization?

 

[keycode]

80° for a GPU is not so tragic at all (same if playing VAM scenes in VR or by desktop mode)... eventually you can fuck off any overclock setting if you were using some crapware software for...

Just for my personal curiosity: wich PCH max temp do you read during VAM sessions? Only if your case has a good airflow without leaks or without important negative turbulence, it should be around 70°/75°c.

 

[TToby]

VaM is a very demanding game. It is a beauty of an performance-eating monster. Especially in VR it is one of the most demanding ones, I have seen so far.
It lacks some badly needed hardware optimization, and therfore a brand new version is in creation for some time and will hopefully be published sometimes in the future.
There are many things you can do wrong in VaM: using multiple light sources will eat up rendering performance, badly optimized clothings with too many polygones, hair styles with too many segments will torture your PC with severe physics calculations aso. This is something you should avoid if possible.
No offense, but your 1070 isn't the optimal GPU for VaM anymore and somewhat on the "minimum requirements" side.
I don't know If you are talking about VR or flat desktop mode, but In VR even in the starting scene there is a low, but constant rendering workload going on, that you won't have in pancake-games.
Even in the main menu VaM is showing a simple, but somewhat demanding scene with some high poly 3D meshes and textures, being constantly rendered in realtime, may it VR or 2D. This might be even more than in some well optimized AAA games with usually low poly meshes and low resolution textures.

I don't know your GPU... are those 80C normal temperature or hotspot temperatures?
For a hotspot temperature 80C is pretty normal at demanding VaM scenes. For a base temperature it is somewhat high. VRAM modules can stand a hotspot temperature of ca 100C and a bit more. But you most likely won't reach this temperature, because the GPU is automatically throttling down long before this.
As Keycode said: There are many different things that can negatively influence heat in your PC case.
As VaM is very demanding: Maybe you have to drop some GPU OC settings and enhance the speed of all your case fans up to a point that is annoyingly loud, to cool down your poor 1070. Maybe try to open your PC case for testing if you have insufficient airflow, aso. This was something I had to learn when I still had my old 980ti which should be somewhat similar to your GPU.

 

[Jiraiya]

gtx1070

vam main menu:
utilization 67%, temp 77C

in scene:
utilization 100%, temp 80C

playing a normal AAA game:
utilization 80%, temp 65C

Why it is so much hotter, even with less gpu utilization?

I think the point might have been not the total temp, but a reference to "utilisation" being 67% at 77'c in VaM or 80% at only 65'c in another game.
ie, he was seeing more use at a lower temp in a desktop game compared to VaM.

I am going to guess that the % utilisation isn't all equal when it comes to heat output. Doing some tasks might generate more heat but still count for the same or less % points according to whatever task read that data. So it's essentially non-equal testing methodology that is the issue.

What else was said about VaM hammering the CPU and GPU is also true!!! So it might be the case you have poor cooling in your PC case and waste CPU heat was being trapped and preventing the GPU from cooling down, hence higher heat in total. Did you test your CPU heat?

I would first take a careful look at your case airflow and fan layout to make sure it's optimal and carries heat out of the case rather than letting it pool around the components. This is honestly the most likely cause.

 

[keycode]

VaM is a very demanding game. It is a beauty of an performance-eating monster. Especially in VR it is one of the most demanding ones, I have seen so far.
It lacks some badly needed hardware optimization, and therfore a brand new version is in creation for some time and will hopefully be published sometimes in the future.
There are many things you can do wrong in VaM: using multiple light sources will eat up rendering performance, badly optimized clothings with too many polygones, hair styles with too many segments will torture your PC with severe physics calculations aso. This is something you should avoid if possible.
No offense, but your 1070 isn't the optimal GPU for VaM anymore and somewhat on the "minimum requirements" side.
I don't know If you are talking about VR or flat desktop mode, but In VR even in the starting scene there is a low, but constant rendering workload going on, that you won't have in pancake-games.
Even in the main menu VaM is showing a simple, but somewhat demanding scene with some high poly 3D meshes and textures, being constantly rendered in realtime, may it VR or 2D. This might be even more than in some well optimized AAA games with usually low poly meshes and low resolution textures.

I don't know your GPU... are those 80C normal temperature or hotspot temperatures?
For a hotspot temperature 80C is pretty normal at demanding VaM scenes. For a base temperature it is somewhat high. VRAM modules can stand a hotspot temperature of ca 100C and a bit more. But you most likely won't reach this temperature, because the GPU is automatically throttling down long before this.
As Keycode said: There are many different things that can negatively influence heat in your PC case.
As VaM is very demanding: Maybe you have to drop some GPU OC settings and enhance the speed of all your case fans up to a point that is annoyingly loud, to cool down your poor 1070. Maybe try to open your PC case for testing if you have insufficient airflow, aso. This was something I had to learn when I still had my old 980ti which should be somewhat similar to your GPU.

...and the most funny is.. even during the "hottest" VAM session the 1070 fans should automatically keep a standard max temp. at a reasonable limit (I have a MSI 1070 still perfect like new after years of VAM stress) ... and if a case airflow is good even that PCH component (that normally is in a critic zone of the motherboard) must get a good cooling during the graphic stress time: for sure its max temp should not be more than 60°/65° even during a parossistic vam activity ...

and about this temp value I was talking about: it's not significative by itself or for the argument in object, but just it was to understand if the case airflow is optimal or not... because graphic card fans blowing should help to keep a low temp even for this neglected PCH component (I was reading it can resist even to 90°c constant temp.... but I would not test if it is true for more than some seconds !!)...

of course to have quiet temps is not easy with modern sexy cases with charming rgb leds and lethal transparent glass front panels blocking 90% of sucking airflow. Long live ugly old meshed front panels.

 

[nsfwthrowaway77]

...

...
I am going to guess that the % utilisation isn't all equal when it comes to heat output. Doing some tasks might generate more heat but still count for the same or less % points according to whatever task read that data. So it's essentially non-equal testing methodology that is the issue.
...

VaM is a very demanding game. It is a beauty of an performance-eating monster. Especially in VR it is one of the most demanding ones, I have seen so far.
It lacks some badly needed hardware optimization,
...

Thanks for replies!
My motherboard shows 3 temps in cpuid: TMPIN0, TMPIN1, and TMPIN2; at 56C, 41C, and 43C. CPU is around 55C. I could probably use another fan, currently one large intake in the front, large output in the back, and the decent ryzen 7 cpu cooler. Opening the case only lowers temp by about 3 degrees however. My readings are desktop mode, and I do have to make video setting sacrifices for vr. 85C is the "danger zone" for 1070 as far as I've read, which made me a bit nervous. My PC did crash once, and I suspected temps, which is why I looked in the first place.

I think the bits I quoted are probably the answer. Bad optimization, specifically in a way that causes high temps.