Daz 3D is part of 
Connect
DAZ Productions, Inc.
7533 S Center View Ct #4664
West Jordan, UT 84084
Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2025 Daz Productions Inc. All Rights Reserved.
Comments
Yep, just like I said, the cooler can be the only real difference sometimes. Some chips might clock better, these slightly chips tend to be given better coolers and marketed as the higher end cards. So there is often a relationship between the two because the board makers want to market their best stuff to gamers who like to overclock their PCs.
And yes, blower coolers can actually be better when going with a multiple GPU setup. The reason why is pretty simple, blower coolers eject the hot air out the back of the system, which helps keep this hot air away from the other GPUs. Any other fan cooler setup will feature a situation where one GPU is blowing air right on top of the other.
BUT, that doesn't mean you have to use them, and the benefits are kind of minor. Neither of my cards are blower styles, and I listed my temps already.
I am running a render right now. My EVGA SC2 is hitting 70C, my MSI Gaming X is only 63. That's pretty much their normal temp even after an hour or two. Sometimes the EVGA might go as high as 74 on a hot day. The EVGA is positioned on top. I once tried swapping the two cards and having the EVGA on the bottom, but this didn't work out so well, as both cards went over 70, so I moved them back. I'm sure some can do better, but these temps are totally acceptable. Also, being 1080tis, they are going to be hotter than a lot of other cards. So I am sure combos of most newer cards can be lower and even easier to manage, with the exception of the 2080ti.
You don't actually need blower cards for a multi GPU setup. If you have good airflow in the case the radial fan coolers will work just fan.
Heat doesn't seem to be an issue for me either for some reason. My 2080 is on top, 1070 on the bottom, there is like an inch between the fans on my 2080 and the circiut board of the 1070. I have not seen my card get hotter than 51 yet lol. I was worried it would be an issue, so far so good. It might be because all the huge fans on my case and the case itself is ginormous, it makes my 1920 x 1080 screen look small.
@kenshaw011267, @TheKD You guys are probably right. I put together a first gen thread ripper. not knowing that AMD chips aren't crazy-hot like they used to be back in the day, so I went kinda crazy on water cooling and fans. I run 2 1080tis, I can slightly smell them when they get hot, and my case makes creaking noises like one of those old electric-coil heaters, but I've had no problems. I would not press my luck with 3 or 4, though.
I have actually been thinking about eGPUs for that reason. Anyone have an opinion?
The cars should never be getting that hoy. They are supposed to throttle down if they get that hot. When was the last time you dusted out your case? The creaking sound seems unusual, can't ever recall hearing any case I've owned making any noise. Do you have CPU-Z and GPU-Z installed? How hot are the GPU's and CPU getting?
BTW TR chips are hot enough that a AIO makes sense. The air coolers for TR are massive and they don't fit in a lot ofcases.
eGPU's are a fairly expensive solution for a desktop. The enclosure cost several hundred dollars and then you need the card itself.
With TR you should have more than enough PCIE lanes for more cards. The question would be how much ventilation does the case provide? TBH if I was in a position where I wanted to add 1 or 2 GPU's I'd perhaps look into getting a high airflow case, the biggest issue there is that a lot of full towers, and I'd certainly want a full tower for arig with 4 GPU's. are positively ancient and come with ODD bays in front which prevent simply loading the whole front of the case with intake fans. You might be able to make the Lian LI 011 dynamic XL work for you. It's got lots of room for fans, side intake but people say it works even in an air cooled system, and its huge as it was designed for custom loop rigs.
Does anyone know if Daz Studio/ Iray supports the Nvidia Tesla K80 GPU accelerator cards? On paper it looks nice for $499 on clearance:
Tesla K80 GPU Accelerator
A dual GPU board that combines 24 GB of memory with blazing fast memory bandwidth and up to 2.91 Tflops double precision performance with NVIDIA GPU Boost, the Tesla K80 GPU is designed for the most demanding computational tasks. It's ideal for single and double precision workloads that not only require leading compute performance but also demands high data throughput.
Model
Brand
NVIDIA
Series
Tesla K80
Model
900-22080-0000-000
Interface
Interface
PCI Express 3.0 x16
Chipset
Chipset Manufacturer
NVIDIA
GPU
2 x Kepler GK210
Core Clock
560MHz Core - Boostable to 876MHz
CUDA Cores
4992 (2496 per GPU)
Memory
Memory Clock
10 GHz
Memory Size
24GB (12GB per GPU)
Memory Interface
384-bit
Memory Type
GDDR5
The main drawbacks to eGPUs (other than the added cost of the enclosure as @kenshaw011267 mentioned and - of course - having a motherboard with Thunderbolt 3 or similar to connect it) are latency from the length of the physical cabling needed and reduced bandwidth (current designs max out at PCI-E x4 speeds.) This makes them kind of impractical for real-time rendering focused tasks like gaming. BUT for Iray in particular, neither of these issues really matters that much.
So if rendering in apps like Iray is really the only thing you care about, I'd say they're worth a look.
Unfortunately all Kepler Cuda generation 3.x GPUs either have been or are about to be officially dropped from Iray due to their advanced age (see the Nvidia Iray changelog thread.)
I figured if they aren't throttling, then they're not too hot, but I should really use CPU-Z/G like you suggested. I checked out the case you indicated, and that looks like a definite contender. And it's beautiful too...
Something to keep in mind about hardware temps in general...
Specific critical computing components (like CPU/GPU dies and VRM controller ICs) may be marketed as having very high temperature tolerances. Thereby implying that component temperatures in excess of - say 80c - within your system may be perfectly acceptable. However this ignores the fact that there are many THOUSANDS of other chips mounted to a graphics card alone that may have no where near those tolerances for operating temps. Especially over an extended period of time. And since SMDs (like CPU/GPU dies and VRM controllers) sink a sunbstantial amount of their heat back into the metal traces connecting them to other components...
It's not often talked about, but one of the main things that sets "professional" level hardware (such as Quadro GPUs) apart from consumer level versions is the amount of time and expense put into making sure that each and every board-level component included (beyond just the sexy brand name ones) is capable of handling whatever levels of excess heat may inadvertantly come its way during intensive operations. Both for longevity of the overall product and because of power consumption.
Operating temperatures have a direct effect on power efficiency. And, depending on the specific board component, having an ambient case temperature of +-10c (even well below the critical temp limits of any one major component) can make a BIG difference in terms of overall power consumption. And power consumption - while seemingly not very important from a cost-value perspective given current electrical usage costs - becomes important again from a product longevity standpoint because additional power consumption leads to additional heat production (in a vicious spiral if left totally unchecked.)
All of which is to say that lower case temps = better