News Alienware Aurora R11 Has the Weirdest Liquid GPU Loop We've Ever Seen

grimfox

Distinguished
Well now, I want to see a rad that you can integrate into a 2-4x serial liquid cooling GPU connection. Imagine a dual GPU setup where you can slot a rad into the gap between the cards. I doubt the performance would be much better than adding a 120 rad to the loop, but it would look super clean. Which is all the more reason to make one, imo.
 

Gillerer

Distinguished
Sep 23, 2013
312
33
18,890
30
When someone starts measuring temperatures in percentages, you know they don't know what they're talking about. Percentages only make any sense for temperatures if talking about absolute temperature and the heat energy contained - otherwise it's pure marketing trickery.

Would have thought better of Asetek, but I suppose it's attributable to their marketing department, or a PR company.

TH isn't doing its due diligence when it disseminates these kinds of releases without calling attention to bad statements.
 

asetekdennis

Prominent
Jan 4, 2019
9
1
510
0
When someone starts measuring temperatures in percentages, you know they don't know what they're talking about. Percentages only make any sense for temperatures if talking about absolute temperature and the heat energy contained - otherwise it's pure marketing trickery.

Would have thought better of Asetek, but I suppose it's attributable to their marketing department, or a PR company.

TH isn't doing its due diligence when it disseminates these kinds of releases without calling attention to bad statements.
We're more than happy to disclose actual temperatures. Absolutely no secrets :)

The reason we've chose percentage changes is that we're comparing up against the former GPU Blower card that Alienware used, in their chassis.

Numbers aren't totalitarian, so you'll get different actual numbers depending on cases and setups. What we decided to do is eliminate that and focus on a strict A/B scenario.

I'm also pretty sure that there will be more content (from Alienware and Ourselves) on the Rad Card in the near future :)

so the gpu itself is mounted on one slot , and the radiator+ pump on another slot ? similar to this though https://ibb.co/x10qDdQ
It is similar. Though without touting our own horn, slightly more advanced :)

Pump is in the coldplate (ie. on the GPU) and the radiator and fan for the rad on the actual Rad Card.
 

Math Geek

Champion
Ambassador
what an interesting idea. never really thought of such a thing as this.

if it manages to get out into the wild could be interesting to work with. for now, pics will have to be enough as i won't be buying this pc just for the novel cooling technique. :)
 
May 16, 2020
3
1
15
0
We're more than happy to disclose actual temperatures. Absolutely no secrets :)

The reason we've chose percentage changes is that we're comparing up against the former GPU Blower card that Alienware used, in their chassis.

Numbers aren't totalitarian, so you'll get different actual numbers depending on cases and setups. What we decided to do is eliminate that and focus on a strict A/B scenario.

I'm also pretty sure that there will be more content (from Alienware and Ourselves) on the Rad Card in the near future :)



It is similar. Though without touting our own horn, slightly more advanced :)

Pump is in the coldplate (ie. on the GPU) and the radiator and fan for the rad on the actual Rad Card.
i see , saw that card on an ebay listing , then this article pops up .. was going ,hmm like seen this concept before ..
 

PapaCrazy

Distinguished
Dec 28, 2011
264
62
18,890
9
Or they could go all Cray on it and fill the whole bloody computer with cooling fluid and just pump that around through a refrigeration unit. Old, proven tech.
Hm, that's interesting. Wonder if someone could jimmy-rig that with an aquarium PC. Kind of reminds me of an attempt by Microsoft to put servers under ocean water. Tides supply a perfectly reliable pump system and free energy... speaking of old, proven tech.
 

PapaCrazy

Distinguished
Dec 28, 2011
264
62
18,890
9
something like this ?
That's pretty much the idea. I checked out Part 4 and the results are a little underwhelming. I think the idea has more potential though, maybe with higher pressure and more efficient flow over components and through the radiator. What a mess though. Linus seemed over it by the end.
 

Gillerer

Distinguished
Sep 23, 2013
312
33
18,890
30
We're more than happy to disclose actual temperatures. Absolutely no secrets :)

The reason we've chose percentage changes is that we're comparing up against the former GPU Blower card that Alienware used, in their chassis.

Numbers aren't totalitarian, so you'll get different actual numbers depending on cases and setups. What we decided to do is eliminate that and focus on a strict A/B scenario.

I'm also pretty sure that there will be more content (from Alienware and Ourselves) on the Rad Card in the near future :)
This comment demonstrates what I expected.

Numbers aren't totalitarian, but mathematics and physics as we understand them work according to set rules.

The percentage is meaningless - not because changing test conditions or test variance - but because the origin point is arbitrary. You can get wildly varying percentage values by choosing to measure temperature changes against any conceivable origin. You can have your pick of (at least):
  • 0°C (80°C to 60°C yields "-25%" compared to the freezing point of water, for some reason)
  • 0 K (80°C to 60°C, or 353.15 K to 333.15 K yields "-5.7%", compared to absolute zero energy state)
  • room ambient temperature (like 20°C): this is the unreachable limit using air or water cooling (80°C to 60°C yields "-33%" of potential cooling)
  • case ambient (maybe 30°C?): this is the unreachable limit using air or water cooling with the radiator fans drawing air from inside the case (80°C to 60°C yields "-40%" of potential cooling)
None of the above percentages mean anything. It's pure marketing talk.

*

EDIT: To add to the above, you can say it's "fine" to say 80°C to 60°C is "-25%", but let's look at the math that arises from using 0°C as origin:

60°C / 80°C = 0.75, so 60°C is 75% of 80°C.
0.75 - 1 = -0.25, so 60°C is 25% less than 80°C.
So far, so good, right?

Suppose we then choose a comparison point other than 80°C. Let's pick -1°C.

60°C / -1°C = -60.00, so 60°C is -6000% ("negative six-thousand percent") of -1°C.
-60.00 - 1 = -61.00, so 60°C is supposedly 6100% less than -1°C.

Not only that but there's a discontinuity at comparison point of 0°C (division by zero), so the ratio would be undefined - you couldn't compare 60°C to 0°C. As the comparison point approaches 0°C from the positive side, the ratio approaches +∞, but approaching the same from the negative side it approaches -∞.

Once percentage calculations using legal values start producing this kind of garbage, it's a pretty good hint that something is wrong. And what's wrong is that the origin is set in the middle of the temperature line.
 
Last edited:

ASK THE COMMUNITY

TRENDING THREADS