Thermal Throttling Issue - ASUS ROG STRIX GAMING O.C RTX 3070 ti

[localhost]

Specs:
- 12900K w/ 360 AIO (stock)
- Asus Rog Strix O.C RTX 3070 ti (stock)
- 2x16GB DDR5 4800mhz (CL40)
- Corsair HX1000 psu (multi rail enabled)


Just recently bought the Asus Rog Strix OC 3070 ti.

I noticed an issue while I was playing the game Warzone. On HWInfo it shown the gpu thermal limit maximum = yes, and the max gpu temperatures reached was 65C-69C.

I did more testing this time on 3DMark Time Spy benchmark. The max gpu temperatures reached was 70C but strangely did not trigger any thermal throttle. I haven’t done anymore testing since I had to goto bed.


The question is why am I getting thermal throttled at 65-69C when playing Warzone? It hasn’t even reached the 83C limit yet. Is my new graphics card faulty? Has anyone experienced this before and know why thermal throttling is triggered at lower temp? Anyone know what to do to fix this?

 

[Jack1n]

I am not sure what the exact point at which 3xxx series cards thermal throttle, but it is much closer to 100c than it is to 70c.
That being said, your temps do seem to be a bit high for your card, do you have the 360 aio mounted in the front, as intake?

 

[joemama]

Did see that on the "Current" column with the other limits as no?
Both pictures you've taken doesn't show that currently the GPU is limited by thermal

 

[AusWolf]

Did see that on the "Current" column with the other limits as no?
Both pictures you've taken doesn't show that currently the GPU is limited by thermal

This.

@localhost As HWinfo shows "yes" as max value at basically every possible limit, I'm guessing it's a momentary glitch in the software. Can you take a picture of GPU-Z showing a constant thermal limit (the same way it's showing a Vrel limit in your screenshot)?

 

[localhost]

I am not sure what the exact point at which 3xxx series cards thermal throttle, but it is much closer to 100c than it is to 70c.
That being said, your temps do seem to be a bit high for your card, do you have the 360 aio mounted in the front, as intake?

Yes I have an AIO mounted at the front of case.

Did see that on the "Current" column with the other limits as no?
Both pictures you've taken doesn't show that currently the GPU is limited by thermal

I posted HWInfo pictures above.

GPU Thermal Limit says “no” in the current and minimum column. I have not seen it say yes before in these column.

Though I keep seeing GPU Thermal Limits “yes” in the maximum and average column. It seems to get triggered when gpu temps of 60-69C is reached while playing Warzone.

But when I do a 3DMark or Heaven benchmark I haven’t seen any thermal throttle flags yet and it reaches much higher temps 71C.

So not sure why HWInfo is saying that it’s thermal throttling when playing Warzone? But when I run benchmarks I don’t get any thermal throttling yet? Trying to figure out if it’s a bug or something wrong with my gpu? I hope someone can help me find out.

This.

@localhost As HWinfo shows "yes" as max value at basically every possible limit, I'm guessing it's a momentary glitch in the software. Can you take a picture of GPU-Z showing a constant thermal limit (the same way it's showing a Vrel limit in your screenshot)?

So would I have to play Warzone and have GPU-Z open and check if there is a “thrm” flag? If so I will do it later today and let you know what I get.

 

[elghinnarisa]

But none of these values show anything wrong with it.

The way these cards work is that they keep trying to boost until something reaches whatever limits they have defined. There is always going to be something hindering it from reaching a higher clock when this happens, and thats perfectly normal.
In your pictures it shows that the card averaged 1950Mhz, which is according to the database: https://www.techpowerup.com/gpu-specs/asus-rog-strix-rtx-3070-ti-gaming-oc.b8966 still quick a bit higher than rated boost clock. So you are in fact not struggling with something operating at a lower performance than it should, yours is in fact boosting above those numbers.

Same goes for my RTX 3070, which is currently being limited by Reliablity voltage(VRel in gpu-z), but its happily sitting at 1995Mhz throwing frames at my face no problems. It's just telling me that the reason it does not go up to, for example 2005Mhz, is because of "Reliability voltage", that is the limiting factor in my case.
In fact, that is what I did. I threw some voltage at it, and it went up to 2010Mhz, now the new limiting factor is "utilization". Which is not surprising since the game im currently playing isn't particularly GPU heavy.
Return it to normal and it drops to 1995Mhz and goes back to "VRel" limiting factor.

So your 3070ti is working just fine, at least according to the images you showed. Everything is operating as intended.

 

[eidairaman1]

Specs:
- 12900K w/ 360 AIO (stock)
- Asus Rog Strix O.C RTX 3070 ti (stock)
- 2x16GB DDR5 4800mhz (CL40)
- Corsair HX1000 psu (multi rail enabled)


Just recently bought the Asus Rog Strix OC 3070 ti.

I noticed an issue while I was playing the game Warzone. On HWInfo it shown the gpu thermal limit maximum = yes, and the max gpu temperatures reached was 65C-69C.

I did more testing this time on 3DMark Time Spy benchmark. The max gpu temperatures reached was 70C but strangely did not trigger any thermal throttle. I haven’t done anymore testing since I had to goto bed.


The question is why am I getting thermal throttled at 65-69C when playing Warzone? It hasn’t even reached the 83C limit yet. Is my new graphics card faulty? Has anyone experienced this before and know why thermal throttling is triggered at lower temp? Anyone know what to do to fix this?

Rma or replace pads, get better cooling in case, keep room temperatures under 75°Farenheit

 

[AusWolf]

So would I have to play Warzone and have GPU-Z open and check if there is a “thrm” flag? If so I will do it later today and let you know what I get.

Yes, please.

As others said above, your GPU-Z screenshot shows that everything is working normally - though it also shows a fairly light load. You need a GPU-heavy benchmark or game to really see thermal limits (if there is any).

 

[bissag]

it is driver bug, mostly happens when you alt + tab between game and desktop