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
...unfortunately few of us can afford the multi four and even five digit price tags for the cards that are best suited to the task. So, we have to take other precautions as you mention. Water cooling, yes that will help keep temperatures down but again adds to the complexity and cost (though not as bad as it did six years ago). I have enough room in the case for such a system if I can find an aftermarket one with a backplate.
I've installed several additional fans in my case to promote even more airflow in support the Titan-X, but the biggest benefit is the 400mm one one on the side panel which draws in fresh air right next to the GPU (and has a filter screen). Part of the reason I'll never get rid of this case, even if I build an entirely new system as there are so few cases with this feature today (most have that [in my book useless] window on the left panel now).
So does MSI afterburner work with any brand or just MSI cards?
Ok. Here are a couple choices ive found.
https://www.amazon.com/EVGA-GeForce-Scanner-Graphics-08G-P4-5671-KR/dp/B076S4RH6K/ref=sr_1_1?s=pc&ie=UTF8&qid=1530321338&sr=1-1&refinements=p_n_feature_keywords_four_browse-bin:6030981011,p_n_feature_four_browse-bin:6066318011,p_n_feature_keywords_five_browse-bin:6147187011,p_89:EVGA|Asus,p_n_condition-type:2224371011,p_72:1248879011&th=1
https://www.amazon.com/ASUS-GeForce-Turbo-Graphic-TURBO-GTX1080-8G/dp/B01HDUVJ2M/ref=sr_1_5?s=pc&ie=UTF8&qid=1530321338&sr=1-5&refinements=p_n_feature_keywords_four_browse-bin:6030981011,p_n_feature_four_browse-bin:6066318011,p_n_feature_keywords_five_browse-bin:6147187011,p_89:EVGA|Asus,p_n_condition-type:2224371011,p_72:1248879011
Thoughts?
Hmmm, interesting. Back in the old days when we still used slings, arrows, spears, and discrete transistors, the plethora of transister designation numbers were due to the uncertainty of the manufacturing process. They'd make a bunch of transisters, designed with a desired response curve in mind, then put them on a testing machine and chart its actual amplification characteristics and throw them into buckets depending on the shape of their response curves, and label them "2N3055" or some other related number in the series depending on the actual shape of the response curve.
Check system review sites, and the benchmark thread here - I seem to recall that the 1070Ti was actually faster, at least for some things (possibly including rendering), than the plain 1080.
Newegg is running their pre-fourth of July sale right now. Definitely take a look.
www.newegg.com
I got this one.
https://www.amazon.com/EVGA-GeForce-Scanner-Graphics-08G-P4-5671-KR/dp/B076S4RH6K/ref=sr_1_1?s=pc&ie=UTF8&qid=1530321338&sr=1-1&refinements=p_n_feature_keywords_four_browse-bin:6030981011,p_n_feature_four_browse-bin:6066318011,p_n_feature_keywords_five_browse-bin:6147187011,p_89:EVGA|Asus,p_n_condition-type:2224371011,p_72:1248879011&th=1
It arrives tomorrow and im super excited!
How many watts power supply is the minimum for it to be compatible? Will it run with a 500w, or does it require higher? Would higher be better?
...according to Nvida the recommended power supply is 500w (the card peaks at 180w). I would still got a little higher particularly if you are considering other future upgrades.
I realize that's a popular belief, but do you have any references to support any of that? Because there are a lot of people who successfully use their GPU's on the standard fan profiles for rendering iray for years, and I don't know of any facts to support the belief that their cards would have lasted longer if they were just gaming and not doing iray. Electronics have thermal limits because of physical characteristics of the silicon, etc. And they have continuous ratings, because that's what they can withstand continuously. For some reason some like to believe that using something at the continuous ratings for a longer period is worse. I don't think the facts support that. My GPU's fans keep the temps at or below about 80C. Whether I'm rendering for 20 minutes or 20 hours. 80C max. That's not a damaging temperature. Just check the equipment specs.
Guess what im currently toying with!?
Im on an 850.
I used to think so too but there must be a reason why nVidia reduces the warranty to 90 days (normally 3 years for gaming cards) on cards designed specifically for mining:
"What’s being stripped away? Details are scarce, but the manufacturers are removing HDMI and/or Display ports as image outputs from the mining-specific hardware. They will also come with reduced 90-day warranty periods due to the intensive 24-hour operation that the bitcoin mining GPUs are likely to see."
https://www.ccn.com/nvidia-amd-to-release-cheaper-bitcoin-mining-gpus/
EVGA specs say it draws less than 250 watts. They say a 500watt PS is minimum, though I'm sure that assumes worst case and that you're using a ton of other stuff simultaneously. Your best bet IMO is to buy a $30 power meter and measure what your computer actually requires. My 1080ti specs say it draws 250watts max, and with a GTX1070 in parallel my computer barely ever gets to 400watts, and I have a 750 watt PS.
Often you'll hear that you need to get a massive power supply. Well, if you're planning on adding a lot of stuff later on then maybe. But most of your components draw relatively little power. The GPU is typically the biggest by far.
Sapat, if there is any way that you can ditch that old card of yours and summon up the funds, defenitely look into it. Im still just doing random tests. What I can see already is improved render speeds, better navigation when dealing with the geometry on screen, far more stability as far as memory use is concerned. Ive already rendered a couple sets that the old card wouldve just crashed on or dumped to cpu and wouldve taken a ridiculous amount of time to render.
Ill toss up some examples...
I did come attempt one render on a set however that still threatened to maximize the cards recources and dump to cpu if not straight out crash it. But only one. It had atmospheric fog and bloated texture files, so...
This render I stopped at 28% after 10 minutes. What gets me about this render is that I didnt hide or delete a single item in the set, I didnt reduce a single texture size on the set or the figure. Anyone who knows this set also knows that it has alot of glossy surfaces. I attempted to render this set with the 770 a couple times. I had to delete or hide as many items and surfaces as I possibly could and I had to resize the textures at least twice. Even after that I never finished the render because it was super grainy and taking an impracticle amount of time to get anywhere with it. That was without any figures or extras in the scene.
Heres something just for fun. Complete render time = 43 secs.
I have both a 980 TI and a 1080 TI in a very well ventilated case. I also have a very intensive render that will run both cards to thermal limit in about 10 minutes on the default fan settings. At thermal limit (85 C for both) the clocks throttle back - eventually to less than half speed. With a custom fan setting the 980 TI stays under 75 C and the 1080 TI stays close to 65 C and both will run at full clock speed and 97% gpu utilization.
I don't see any point in running the cards hot if I don't have to.
So maybe this is a no-brainer, but should I run my fans at 100% while I render or is that unnecessary? I never have anything super complex going on...
Well it's obvious that they run 24/7 but how well they're cooled is hard to tell from a picture. Wonder if there are some data out there about this.
...very nice.
IMO the most important thing is case cooling, for if it's too hot inside the case even the best or fastest CPU or GPU coolers/fans are not worth much. Room temperature also matters.
I have a GTX 1070 with two dynamic cooling fans, in a well cooled case (2 x 120 mm fans in and 3 x 140 mm out). The card is usually at 60-65 C and the GPU fans run at 50-60% speed no matter how long time I render, depending on room temperature which is usually between 25-30 C. The 2 x 140 mm top fans in the case are also running at high speed when I render (the case has a manual switch where you can set their speed to low or high).
I haven't noticed any difference in GPU clock speeds or temperature whether the scene is complex or not so it doesn't look like that matters.
First off, a lot of mining setups are not engineered well. The airflow is not directed correctly. In fact, some are not even set up in proper cases but simply mounted on a backplane and running in ambient room air. That’s going to shorten the life of any electronics. In an intelligently designed pc, the case itself assists with cooling as in constrains the airflow brought in and routed through by intelligently placed fans. This is just simple engineering at work and it’s a beautiful thing. My pc case for example has cool air from almost floor level pulled in by a fan and that airflow is directed up and over what needs to be cooled before being pushed out of the case by a strategically placed exhaust fan at the top rear of the case. Now if I had 3 or more gpu’s in my case I would consider a liquid cooling setup for the gpu’s necessary because you can only do so much with engineered airflow. However for a single gpu ,or two gpu’s in a well designed case with intelligent fan placement and direction, ambient room temperature of about 68f-75f is perfectly fine for both hours of gaming and long rendering stretches. I think a lot of people get alarmed when the thermal sensors cause the fans to ramp up but that’s what they were designed and built to do.
Pretty sure it does, at least it works onmy cards and i have NO idea of their actual make (PC was made up for me by a company).
To say the least...
Consider this: a GTX 1080ti uses 250watts, which means it generates 250 watts of heat. So four of those next to each other is like a 1,000 watt toaster oven on HIGH. Certainly not what the GPU designers designed for.
Also, keep in mind that many/most of these mining operations are do-it-yourself operations where a group of guys hear that they can make a ton of money if they buy a bunch of GPU's. So they build some racks and get some power supplies and get it running as quickly as possible. It's not like they can buy a pre-made and pre-engineered mining rack system from Dell or HP or NVIDIA. They rent some space and throw something together and hope for the best. Totally do-it-yourself. Just imagine trying to figure how to cool the equivalent of 20 toaster ovens on HIGH, when all you have is an office building space with standard A/C vents in the ceiling. That's why youtube has tons of "hey, here's how I cooled my mining rig" videos.
It's not the continuous rendering that causes problems. Its about putting your PC in a pizza oven and hoping the GPU fans will keep it cool while it's rendering.
Firstly, congratulations on your new card - good to see you enjoying the new speed.
I have a very similar setup to Nanffuak who posted earlier, and I also use a custom fan profile. I have used MSI Afterburner for quite a long time, but I've recently had to stop using it and go with the software that came with my card because Riva Tuner Statistics Server is used to display the graphs and I'm having problems with it since the update to Windows 10 1803. But if you're not runing that build of Windows 10 then I would definitely recommend it. I usually runthe fans at 30% at 50 Celsius and at 60% at 70 Celsius and this is enough to keep them at around 70 Celsius during a prolonged render. If it's a very hot day then I turn the GPU fans up for long renders.
If video card makers wanted they could use something other than solder to connect the GPU to the video card which is the Achilles heel of the video card most failure is due to the constant cooling and reheating of solder connecting this to the card. When this starts to happen your GPU becomes disconnected from video board and that's it for your video card (time to by another). Keeping the card cool lengthens the life of these video cards solder points. If solder was a good solution we would have them on expensive desktop CPU sockets but most good motherboard manufactures don't solder CPU's because this is the hottest part of the system and the most expensive. Consequenlty, we have sockets that make the direct connection to CPU without solder and really good air or water cooling for the CPU. Solder is a cheap way for companies to attach something to a board and it breaks down! Do CPU and GPU wear out? Yes they do but most of the time it is bad solder contacts or bad resistors on the boards that go bad that bring the CPU motherboard system or video card down. Yes I'm over simplifing but it would take pages and pages to expain why things are engineered to break just ask any good engineer about the problem Henry Ford ran into for making a car that could run too long without breaking.
Oh, Kyoto Kid yes MSI Afterburner will work with any video card.
Firstly, congratulations on your new card - good to see you enjoying the new speed.
I have a very similar setup to Nanffuak who posted earlier, and I also use a custom fan profile. I have used MSI Afterburner for quite a long time, but I've recently had to stop using it and go with the software that came with my card because Riva Tuner Statistics Server is used to display the graphs and I'm having problems with it since the update to Windows 10 1803. But if you're not runing that build of Windows 10 then I would definitely recommend it. I usually runthe fans at 30% at 50 Celsius and at 60% at 70 Celsius and this is enough to keep them at around 70 Celsius during a prolonged render. If it's a very hot day then I turn the GPU fans up for long renders.
Good to know, thanks. Im pretty sure that I have fan controllers somewhere in this computer.
Regarding the real causes of failure of GPU's, here's a quote from a Former Chief Product Architect at AMD, when asked "What are the most common reasons for GPU cards to fail?". His answer (I've highlighted some pertinent parts):
Basically he's echoing the points I've been making. As long as you don't operate outside the continuous ratings (by overclocking or damaging the cooling system), your GPU should be fine.