Installed EKWB 3090FE block - expected ballpark temps?

I do love the O11d in general, but airflow isn't optimal. I see a lot of people using bottom rad as intake and top as exhaust but in my mind all your doing it taking the heat out of 1 rad and putting it the other. EDIT just seen your Reddit post, i really dont like the side distro block as it takes away a space for 3 fans to put some good airflow into the case, even if it's at the side, it's better than nothing.

The front panel fan mod really helps, but it loses the look of the O11d.

All in all, your temps are still great, it just depends on how far you want to chase the numbers :)
 
Olly said:
EK block on my 3090 tuf never goes over 45 degrees full load. that's with a 5900x in the same loop and 2x360mm rads 60mm thick
Click to expand...

What's a full load? overclocked mining, overclocked benching, gaming or web browsing? with an overclocked CPU in the loop with a 1Ghz+ overclock at 1.37v?

Don't want to sound picky but "full load" doesn't mean much without something to reference against.
 
@Olly indeed the O11D is a fine case, and can really make a showpiece of a system sat on a desk. As I’m finding out however, as standard airflow isn’t great. Yep there sure is compromise with the side distro - I find it aesthetically pleasing for a nice glow and view of system out the front when I’m spending all day at the desk working, yet perhaps would go for the front distro if I was doing it again (but still don’t find the front gives as clean a look, being able to see the pump and all). A front panel mod is not for me - as you say, you lose the look.

I think my remaining options are to see if rear is better as intake, and play with fan and pump rpm. Not a priority to chase numbers, main thing is keeping the components safe in this system, which cost far more than previous pc, which was soft tubing and more of build and forget - didn’t really pay much attention to temps. Just a little concerned when I was running workloads for several hours yesterday the water temps saturates to about 34 deg c, about +12 over ambient.
 
Trunkey said:
What's a full load? overclocked mining, overclocked benching, gaming or web browsing? with an overclocked CPU in the loop with a 1Ghz+ overclock at 1.37v?

Don't want to sound picky but "full load" doesn't mean much without something to reference against.
Click to expand...

Furmark and cpu stress tester running and left for an hour. 3090 has a small overclock on it and the 5900x is running about 4.4 @ 1.28v
 
Okay finally remouted FE block. Used longer screws around the core first time thinking no way smaller screws were one to use. Seems they were, temps improved to where I expected.

Consuming 525w (400w GPU / 125w CPU) in stress tests will result in GPU sitting around 60 degrees, hot but what's expected with fans at 1300RPM on a 360mm Slim rad after an hour reaching balance. normal gaming load, so around 425w between CPU and GPU in AC Valhalla and will settle in low 50's at same 1300 RPM. Pretty content with that in a HTPC. Memory with +1000 settles in around mid 60's.

As a bonus, did have slight coil whine on stock air cooler. All gone, seems pressure on the caps helped.
 
Radox-0 said:
Okay finally remouted FE block. Used longer screws around the core first time thinking no way smaller screws were one to use. Seems they were, temps improved to where I expected.

Consuming 525w (400w GPU / 125w CPU) in stress tests will result in GPU sitting around 60 degrees, hot but what's expected with fans at 1300RPM on a 360mm Slim rad after an hour reaching balance. normal gaming load, so around 425w between CPU and GPU in AC Valhalla and will settle in low 50's at same 1300 RPM. Pretty content with that in a HTPC. Memory with +1000 settles in around mid 60's.

As a bonus, did have slight coil whine on stock air cooler. All gone, seems pressure on the caps helped.
Click to expand...
I got lucky too, coil whine was greatly reduced :D
 
grumpybeard said:
Hey all,

finally managed to get the EKWB 3090FE block installed. Shared on Reddit, including screenshots from Furmark tests. Really not too sure what to make of some of the comments regarding performance, which sees GPU max out at 54 deg C on core and 70 on memory junction under load. This compares with 104 deg C with stock cooler.

I’m fine accepting card may not be best performer, looking for some expert opinion here :)

https://www.reddit.com/r/watercooli...urce=share&utm_medium=ios_app&utm_name=iossmf
Click to expand...
Seems a bit warm but what rpms you on? Just pwm? I've upped mine in latest rig. Made no diff to sound either
 
If your water delta is 11⁰c at sub 1000rpm I wouldn't worry too much. I prefer silence so have a water delta around 15⁰c under load currently but can drop that to 7⁰c at full speed. It's surprising how much heat a gpu will dump into a loop. By far the best rads for sub 1000 rpm are thermolab black ice SR series but too wide for most cases at 133mm for a 120mm fan.
 
bigjimmyauk said:
If your water delta is 11⁰c at sub 1000rpm I wouldn't worry too much. I prefer silence so have a water delta around 15⁰c under load currently but can drop that to 7⁰c at full speed. It's surprising how much heat a gpu will dump into a loop. By far the best rads for sub 1000 rpm are thermolab black ice SR series but too wide for most cases at 133mm for a 120mm fan.
Click to expand...
When you say about how much heat a gpu can dump into a loop have you guys ever compared a same setup with 2 x aios? I've always wondered whats better
 
slim01 said:
When you say about how much heat a gpu can dump into a loop have you guys ever compared a same setup with 2 x aios? I've always wondered whats better
Click to expand...

There's not much in it tbh. You have a double failure point. And usually find that low task stuff a single loop will be cooler/on par with a dual loop with less noise. Massively more expensive too especially when using £200 pump and res combos
 
The temps looks actually nice.

The Zotac one's runs hotter even under water... Considering that all their cards are reference designs that are further brought down design wise.... Using the cheapest components they can find and the most generic vrm design they can use that won't cause the gpu to run unstable...

With that I am already steering away from them.

But Nvidia and AMD's new founders design seem to be real good.
 
I cant complain about my FE currently mining (Only cause I'm going to bed and there's more point than Folding or CETI) at 114.7M and 35c GPU 72c mem junction & coolant temp 28c CPU idling at 24C across all ten cores. Room Temp ??? Shorts and T-shirt though.
 
bigjimmyauk said:
There's not much in it tbh. You have a double failure point. And usually find that low task stuff a single loop will be cooler/on par with a dual loop with less noise. Massively more expensive too especially when using £200 pump and res combos
Click to expand...
Double failure point with two aios however at least 100% of the cooling doesn't stop. The only time I've ever had gpu and cpu on a single loop was with a cheapy thermaltake lcs system (well 250 quid or so). The first time the pump broke, I was at the pc and saw the alarm and managed to turn off in time. The second, a friend was on the pc and I came into a room of smoke. This was in the old days though where amds had no auto shutdown or even thermal throttling I believe.

Since then I've always been wary of single pump solutions. However as said I think everything nowadays would just turn off before any damage done if pump stopped working?
 
Trunkey said:
I cant complain about my FE currently mining (Only cause I'm going to bed and there's more point than Folding or CETI) at 114.7M and 35c GPU 72c mem junction & coolant temp 28c CPU idling at 24C across all ten cores. Room Temp ??? Shorts and T-shirt though.
Click to expand...
That's insane. Your pc basically sounds like it's at room temp? I'd only wear shorts and t-shirt up to 21c at least anyway lol
 
slim01 said:
Double failure point with two aios however at least 100% of the cooling doesn't stop. The only time I've ever had gpu and cpu on a single loop was with a cheapy thermaltake lcs system (well 250 quid or so). The first time the pump broke, I was at the pc and saw the alarm and managed to turn off in time. The second, a friend was on the pc and I came into a room of smoke. This was in the old days though where amds had no auto shutdown or even thermal throttling I believe.

Since then I've always been wary of single pump solutions. However as said I think everything nowadays would just turn off before any damage done if pump stopped working?
Click to expand...

There such a multitude of safety features these day. Bios options, aquacomputer sensors that can turn your pc off. Your well protected tbh
 
bigjimmyauk said:
There such a multitude of safety features these day. Bios options, aquacomputer sensors that can turn your pc off. Your well protected tbh
Click to expand...
Yeah for sure. I watched that tech support off linus did with jayz2cents vs Steve where they sabotaged a pc with all kinds of issues and they had to try and fix the fastest.
Jay didn't even realise the cpu was installed with no thermal paste whatsoever. Only when he did a benchmark he realised the cpu was under clocking (obviously to keep temps right down). You could prob run a cpu with no cooler nowadays for at least a few mins anyway lol. Old amds would burn up in seconds
 
You must log in or register to reply here.
Back
Top Bottom