🎓 Dr. Freemo 🇳🇱 @freemo@social.cleverlibre.org

@kranfahrer Yea mine was 100% continuous load.

After 15 minutes of continuous load I saw the water temperature still had 5 C more before it maxed out. The GPU junction temperature itself had a lot more breathing room at 25C more before it maxed out.

So yea this thing is specced to run at full speed continuously.

@lucifargundam sure send me the address and ill send one to ya

@lupyuen Hows that going for you?

@kranfahrer Ok managed to get it a bit hotter.

Before the tests were running long enough to heat up the cooling fluid, so the CPU was staying cool.

However if i keep it at 100% workload for 15 minutes or so the liquid in the coolers slowly heat up to 65 C and max out and the GPU itself gets to 77C at that point. Keeping it maxed out it appears to be as hot as I can get it.

That said if the GPU relaxes the load for even a second or two the liqud quickly cools to 50 C after a second or two.

@kranfahrer oh really, I will definately consider that for my next build then.. for what i do i have no need to do my own watercooling (the GPU watter cooling is how these cards come)... but i might do it just for coolness factor at that point.

@kranfahrer By the way I only know these are liquid cooled, I'm not sure if its water though. I suspect it uses gas transition cooling and uses some sort of very low boiling temperature liquid like heat pipes do.

@kranfahrer I dont usually overclock myself as I want a long-life out of my equipment and need it to be reliable.

I am running a different benchmark and this one can only get it to 52C... so yea seems 60C is really the max I can push it.

@kranfahrer So running a benchmark for a few minutes to max out one of the GPU's it reaches 60C. Couldnt really get it too much past that without overclocking.

I mostly compute datascience/big data stuff. AI too.

@kranfahrer downloading and installing basemark to see how well this does on a graphics benchmark (I usually use these for computation not graphics, and they are designed for that technically)

@kranfahrer let me run a GPU benchmark on it for you real fast see what I get.

@kranfahrer The tachometer on all of them are at their lowest. So less than 10% load on them right now.

@kranfahrer Thats correct, thats just the temp right now, there isnt much running on it.

@kranfahrer I run this at full load all the time, just no need for it right now. It works well at full load the water cooling on the 4 gpus makes a huge difference.

If your curious how hot I can run them under load I can run a test for ya, but I have no need to do so for my self.

@kranfahrer so yea based on that it would appear my GPUs are running at a cool 25 C right now and my CPU around 63 C

@kranfahrer

Here is a dump of all my onboard sensors to give you an idea of the temperatures:

amdgpu-pci-4500
Adapter: PCI adapter
vddgfx: 750.00 mV
fan1: 919 RPM (min = 400 RPM, max = 3300 RPM)
edge: +27.0°C (crit = +70.0°C, hyst = -273.1°C)
(emerg = +75.0°C)
junction: +27.0°C (crit = +105.0°C, hyst = -273.1°C)
(emerg = +110.0°C)
mem: +27.0°C (crit = +95.0°C, hyst = -273.1°C)
(emerg = +100.0°C)
power1: 9.00 W (cap = 220.00 W)

amdgpu-pci-0c00
Adapter: PCI adapter
vddgfx: 750.00 mV
fan1: 928 RPM (min = 400 RPM, max = 3300 RPM)
edge: +24.0°C (crit = +70.0°C, hyst = -273.1°C)
(emerg = +75.0°C)
junction: +25.0°C (crit = +105.0°C, hyst = -273.1°C)
(emerg = +110.0°C)
mem: +25.0°C (crit = +95.0°C, hyst = -273.1°C)
(emerg = +100.0°C)
power1: 3.00 W (cap = 220.00 W)

k10temp-pci-00cb
Adapter: PCI adapter
Tctl: +63.9°C
Tdie: +36.9°C

nvme-pci-4100
Adapter: PCI adapter
Composite: +43.9°C (low = -273.1°C, high = +72.8°C)
(crit = +75.8°C)
Sensor 1: +43.9°C (low = -273.1°C, high = +65261.8°C)
Sensor 2: +53.9°C (low = -273.1°C, high = +65261.8°C)

nvme-pci-0900
Adapter: PCI adapter
Composite: +42.9°C (low = -273.1°C, high = +72.8°C)
(crit = +75.8°C)
Sensor 1: +42.9°C (low = -273.1°C, high = +65261.8°C)
Sensor 2: +50.9°C (low = -273.1°C, high = +65261.8°C)

amdgpu-pci-4800
Adapter: PCI adapter
vddgfx: 775.00 mV
fan1: 885 RPM (min = 400 RPM, max = 3300 RPM)
edge: +28.0°C (crit = +70.0°C, hyst = -273.1°C)
(emerg = +75.0°C)
junction: +30.0°C (crit = +105.0°C, hyst = -273.1°C)
(emerg = +110.0°C)
mem: +29.0°C (crit = +95.0°C, hyst = -273.1°C)
(emerg = +100.0°C)
power1: 5.00 W (cap = 220.00 W)

amdgpu-pci-0f00
Adapter: PCI adapter
vddgfx: 750.00 mV
fan1: 949 RPM (min = 400 RPM, max = 3300 RPM)
edge: +24.0°C (crit = +70.0°C, hyst = -273.1°C)
(emerg = +75.0°C)
junction: +25.0°C (crit = +105.0°C, hyst = -273.1°C)
(emerg = +110.0°C)
mem: +26.0°C (crit = +95.0°C, hyst = -273.1°C)
(emerg = +100.0°C)
power1: 11.00 W (cap = 220.00 W)

k10temp-pci-00c3
Adapter: PCI adapter
Tctl: +63.0°C
Tdie: +36.0°C
Tccd1: +62.8°C

nvme-pci-4200
Adapter: PCI adapter
Composite: +42.9°C (low = -273.1°C, high = +72.8°C)
(crit = +75.8°C)
Sensor 1: +42.9°C (low = -273.1°C, high = +65261.8°C)
Sensor 2: +53.9°C (low = -273.1°C, high = +65261.8°C)

@kranfahrer Ahh ok, VERY NICE cooler you have going there. My CPU is air cooled but all 4 GPUs are water cooled. The water cooling however comes built-in and sealed with the cards so no jumps or reservoirs needed.

I'm honestly not sure what temperature they run at, all I can tell you is I have yet to find a game capable of maxing it out to the point that it thermally throttles. I'll have to check next time what they run at.

You have me seriously considering that little cooling setup you have going though, whats the brand and model?

@icedquinn I dont even have the video bonding stuff configured, thought supposedly its supported on linux.I used it for GPGPU stuff. I'm not a big gamer but if i play a game im likely running it off a single GPU. Though with that said these are so overpowered I can put any game at max settings and not max these things out.

I havent tried 2077 yet though, might be worth getting it just to see how the GPUs fair.

@kranfahrer Are we talking about the weather or are you asking about the temperature of the GPUS?

@kranfahrer It works, why do you ask?

@kranfahrer Not too bad today at all, 73 F