I have many identical 1080ti FE’s. None of them perform the same from a temp perspective. The silicon GPU dies have vastly different yields and performance from die to die and wafer to wafer, just like CPU’s.
So each GPU will have very different over clocking levels and will run at different temps if you let them. Pulling the same temp on all GPU’s (this is what I do) will cause your performance numbers to shift. However, from my experience the average will remain relatively the same depending on the number of cards you have (some will go up to 780-790, some will drop to 720).
I understand why you would not want the SATA connector. A single PCI riser can draw more power than it is rated for based on the PCIe specifications. But why would you not want to use the 4 pin Molex? its rated at 11amps and safely power 2 PCI risers.
Using the splitter would be to much power on the PCI power cable and VERY dangerous.
Run your fan at 100%, will bring temps down to mid to low 60C, unless you have airflow issues. If its still high then reduce power as suggested by CitricAcid.
I was using nvdiai inspector and when i set the temps to say 65c it wouldnt auto control the fan to get it to this temperature.
Same problem with afterburner.
Currently running at base clocks with 100% power and fan at 81%
Temps are sat at 65c trying to balance fan speed to temps just reducing to 71% at the moment seeing what affect that has on temps
What are you talking about? I have many GPU’s fans running at 100% for almost two years now. I have never seen high failure rate if you do preventative maintenance… But yes I have fans in my spare part inventory and replace one occasionally.
Why would you not run your fans at 100% on a mining farm? That makes no sense.
You need to switch that around. Run a constant fan speed and auto the power level to hit the desired temp. Additional external airflow is helpful on the FE versions but not necessary.
My rigs average 450-480Sol/sec at a set temp of 63C.
Sorry my rigs do 750-780Sol/sec not 450-480Sol/sec… it was late
Hmmm… joined to post a link to an outside site that shows a blank page. Nothing bad reported or happened on my Mac, but curious if it will behave differently for Windoze…
The 1080ti’s that are triple fan cooled, I run at 100% power, on auto fan control temp set to 75 and the fans are running between 60% and 62%.
Something to keep in mind, whatever power you dedicate to the GPU the fans take part of that. So if you are running at 90% power with fans at 100%, you are taking power to run the fans at 100% that is not being used for H/s.
Hi CitricAcid,
Oops I just noticed my error, my rigs average 750-780Sol/sec not 450-480Sol/sec… is was late.
Yeah I don’t use any of the triple fan models for that reason. I have FE blowers, they are more efficient and cost less.
Unless you are water cooling, the rate limiting factor on power is cooling. With fan at 100% my chosen temp set point becomes the rate limiting factor. So at any given point in a day my rigs are doing the maximum hash they can given the ambient temp.
With my OC levels many GPU’s start becoming unstable at a set temperature of 64C, this is why water cooling is not worth it (the ROI does not work out). This is actually a good thing, I cycle through my GPU as resale, I would not want to buy a GPU that has been mined at 75C! So I wont sell a GPU that has been used in those conditions. I have log files for every hour a GPU has run so I can show its usage history over its life, and they never see higher than 65C. Replace the fan, clean the heat exchange fins, and I can resell a GPU with a clear conscience.
I appreciate the information. I am glad the FE GPU’s work for you, they did not work for my vision of how a rig should run.
all of my rigs are in a 65F server room with a 52% humidity setting.
Way back in the beginning (after my AMD Fury rig) I purchased 2 1080Ti FE GPU’s and 2 1080ti EVGA GPU’s, for $$ reasons and testing. Technically spec’s put the FE card above the 3 fan GPU.
Even at 100% fan, the FE card ran hotter and at a 30 H/s to 50 H/s below the 3 fan 1080 TI at stock and overclock settings. This was slight variations in H/s difference when testing fans on auto and 100%, but the triple fan GPU always performed better. just want to add that this is how it was for me.
Has far as Hybrids go, I disagree that the ROI does not work out. the price difference between a EVGA 1080ti 3 fan GPU and a 1080ti Hybrid is usually USD $30 to $50. The temps never ago above 48c. I get a slightly higher H/s over the 3 fan, a more consistent rate, and a better resale value.
I don’t have a 65F ambient so I have to work with 70F+… except in winter.
Each of my rigs has a secondary box fan for supplemental air flow. This is more to direct the heat toward the cooling system intakes but also cooled the GPU’s a few C.
For me: I found that you must maintain a ratio between the GPU OC and Memory OC. There is a pretty significant difference when you have the right ratio (just like a CPU). I have code that sweeps the Mem OC for each GPU OC on each card. Once I have identified the Max GPU OC then the Mem OC can be set to the optimum. Once I am at this point the temp is the rate limiting factor for my H/s.
I set my GPU’s at 62-63C because I discovered that once the OC is optimized the GPU’s will start to become unstable with power and temp. For example: when my GPU’s were set to 62C and the winter temps plummeted the software tried to increase power to maintain a 62C temp. However, several GPU’s on different rigs started to become unstable. I didn’t expect this and my software did not have a power limit set for reboots (they do now). Then (again cold) I had GPU’s start to drop hash rate (not hang just drop hash). The software compensated with increased power. At some point the power could not compensate for the lower temps and the rig rebooted due to lower GPU performance (I allow 10% drop from the GPU moving average). However, the power was now set rather high, and once rebooting the temps hit my limit of 70C and caused a reboot before the power level could be dropped (reboot loop). That was harder to fix but my rigs text me when they reboot more than 3 times in 15 min so I caught it pretty quickly. Now I sweep power setting as well after GPU OC and Mem OC have been set. This gives me a power limit for each GPU as well.
So my point being that I have run into more issues with cold rather than heat so water cooling would not increase my earnings since I am running cold now. Running the fan at 100% on my rigs ensures I get the max hash rate I can with the absent temp and a power limits set for each GPU. However, in all honestly I have not tried a fan sweep… but that would be a very difficult matrix to interpret.
There is also a harmonic present on Nvidia GPU’s between the GPU OC and the Mem OC. Its rather hard to find but its ~20H/s, so worth it. I have just started writing beta code to find the oscillation and compensate for it.
Yes, you can heat your house with rigs but you need to take the fire risk very seriously. The first year I tried this I had 5KW of rigs but it was not enough and the furnace still ran. Now I keep ~8-10KW of rigs in my home in the winter and the furnace literally does not run.
I have a basement room so you cant really hear the fans on the upper levels. I have an air handler on my HVAC system so air is always moving. In the winter I just have the air intake get all its air from the rigs. I then either have cool air from the house enter the floor in the rig room or when its too warm, cool air from outside, or a mix of the two. Unless it gets really cold I almost always have a some fresh air coming in from outside but it does not need to go through a heat exchanger. The only drawback to this is I don’t get the humidity exchange that would happen with the heat exchanger so I have to add extra humidity into the house to compensate.
Another drawback is when there is an unexpected warm winter day, it can be hard to bring the temps down and you will be opening windows to let the heat out. My rigs auto temperature regulate so when this happens I can send a command from my phone and lower the set GPU temps on the rigs to something in the low 40C range. This dramatically reduces the power and my hash rate, I use it a bit like a thermostat. Works quite well, I just need to automate it now.
A word of caution, rigs in your home can be a fire hazard. This is where you and your family live so safety comes before profits. Be absolutely certain you have your rigs setup in a safe way (no SATA connector anywhere except the HD. Go way overboard on safety. I have a webcam and IOT smoke alarm in with the rigs. If the smoke alarm goes off I get notified and power to all rigs is cut immediately. I am looking at installing a fire extinguishing system in that room as well, but for now the room has no flammable items in it. No wood or paper, all metal shelves, cinder block walls, and fire retardant ceiling.
Depending on which EVGA model you get, they seems to draw less power, produce less heat and generate about the same sols/sec as other brands. Seems that the other AIB cards like 8 + 8 power as opposed to 8 + 6 that the SC cards do.
For example, my EVGA 1080 Ti SC cards use 224W, average low 50’sC and produce ~750 sols/sec or average around 3.4 sols/watt. I run them at 90% power, +100 GPU and +600 memory.
The other cards (MSI, Gigabyte, Zotac) run low 60’s, use 240W+ and produce similar sols/sec. I run those at 85-90% power, + 100 GPU and +0-600 memory.
I’m sure I can tweak more if I had the time and patients, but overall the EVGA cards just perform better for me overall.