OTOY Forums

Hey, so I'm pretty perplexed about what's going on.

So I have 2*2080 Ti's

Yesterday I set a heavy scene to render, only to realize both my GPU's stay at around 50-55 degrees Celsius. This is bizarre because over the last year having these GPUs, they easily get to 70-80 degrees when rendering heavy stuff, sometimes even 85-90. I render almost everyday, it's been a very strict pattern.

It's very very confusing, degrees are very low, fans are very low.

The numbers are confusing as well - While rendering, Octane shows 100% GPU usage. Windows shows 0-2% GPU usage. GPU-Z shows 100% GPU usage.
The base clock keeps jumping frantically between 600-1300mhz while rendering. (Maybe that's normal I've never really monitored base clock speeds)

I haven't changed anything in settings or nothing, only thing is I updated to Octane 2020.1.5 R4 and had to update GPU drivers for it to work. But I'm pretty sure I have rendered some stuff since and all had fans working at high speed and GPU heating up to the normal 70-80c. Just yesterday it all changed.

(Another thing I noticed, the only game I play is Modern Warfare, and I also noticed that starting yesterday, my FPS has tanked and I can only play at 1080p unlike the normal 4k I always used to play with no issues, and generally something feels off with the gameplay - maybe it's just the lower fps but I really have no idea since I never play any games besides this one)

So yeah, I'm at a loss. I have no idea what to look for, where to even start figuring out what's the issue, or even if there is one. I feel completely dumb.

I'm not complaining that the tempratures are low obviously, I just have a feeling it implies the GPU is being underused by the system. As if it's not really working to its full potential.


Anyway, maybe there's something in the new updates I've missed or something obvious I'm missing that I havent looked at.

Hope anyone could be of any help because I feel like my computer has been stolen and I have no clue where to start looking.
Thanks!

Gpu usage measure is correct in LiveViewer, not windows's. You should use another software to see GPU usages, as "ASUS GPU Tweak II" or so.

Thanks for replying Aoktar, I'm using both MSI Afterburner and GPU-Z, you think those are not enough?

Also, is there any new updates about GPU's not heating up in Octane? Or any known things people do when they encounter this issue?
I mean I tried googling it but like, where do I even start figuring out what's the problem?

I'm not deep into CORE part, it's a low-level topic for the Otoy team. But I don't think that's a worrying topic if GPU is staying cold while you see high usages. It's fine in my poor opinion. Anyway ask to Otoy team or help if you don't be satisfied with that

I believe what is happening is your scene is going out of core - check to see if your VRAM exceeds the 11GB of your 2080Tis (or the 22 if NVLinked). If this is the case data has to be moved from main CPU Memory during render and as a result your GPUs might have a series of spikes instead of one continuous 100% usage.

Thanks for your input plasmapeppers
So I tried to render out different lighter scenes and its the same thing. I also disabled out-of-core options, and same.

I ran octanebench again, and i got a score of 328 (2*2080Ti, Threadripper 2920 12-cores 3.5Ghz)
This is extremely low, last time I ran the bench score a few months ago I got 575.
I definitely think there's something not working properly. However, can't understand what.

I'm thinking of reverting to previous Octane versions, but it's a headache I'm trying to avoid, especially since I didn't really have any problems after updating. They just randomly started yesterday and I didn't do anything different.

Ok. So. I rolled back the Nvidia drive, and we're back to normal. As bizarre the issue was, as simple the solution was.

Oh yes, the loud noise of fans that symbolize the extreme effort of electric pulses moving between metal and plastic.