Monitoring Temperatures with HWInfo64 - Best Practice?

[RevGAM]

29 minutes ago, IkeaGnome said:

And length. A lot of AIOs don't have long enough tubes to reach the front of the case and the bottom. Especially once you add in a GPU. 

Yes, that's exactly what happened to me. Even though the Frozen Notte has 450mm long tubes, my GC prevented me from having the tubes down on the side AND on the front. I had to get a vertical mount to overcome that problem, which still isn't perfect because a lot of vertical mounts put the GC too close to the side panel, which is usually solid glass or metal, as is my 5000x. Mine allows me to adjust the angle so that the fans blow upwards at a 45^o angle but it's still very long and comes within less than an inch of the rad's bottom 2 fans when mounted on the front.

 

@freeagent**cough cough**

1 hour ago, freeagent said:

For one, it looks absolutely ridiculous.. and on my H100 tubes down didn't do anything meaningful. I originally attempted it to quiet the pump, sounded like a midget with a rattle can in my ginormous case at the time.

I didn't realize that your sense of aesthetics is so refined.
So, was it rattling when the tubes were up? That's what it seems but your explanation is a bit garbled. How'd you finally resolve the issue?

[freeagent]

1 hour ago, RevGAM said:

I didn't realize that your sense of aesthetics is so refined.

I am much more sophisticated than you give me credit for, I just enjoy it here down in the dirt with my mismatched fans 

 

1 hour ago, RevGAM said:

So, was it rattling when the tubes were up?

The pump started to rattle a week or two after I started to run it, it did not matter the orientation.

 

2 hours ago, RevGAM said:

How'd you finally resolve the issue?

 

I didn't. Corsair sucks and I have refused to buy their junk since the TX-850 I had took a massive dump.. kits of their ram that don't work with each other. That company is a joke. Just a fancy Thermaltake.

[NorKris]

34 minutes ago, freeagent said:

Corsair sucks 

i have to shot in that  not much(if anything) beats cmd pro's + icue + elgato  profiles and changing it fast  

[RevGAM]

1 hour ago, freeagent said:

I am much more sophisticated than you give me credit for, I just enjoy it here down in the dirt with my mismatched fans 

 

The pump started to rattle a week or two after I started to run it, it did not matter the orientation.

 

 

I didn't. Corsair sucks and I have refused to buy their junk since the TX-850 I had took a massive dump.. kits of their ram that don't work with each other. That company is a joke. Just a fancy Thermaltake.

Aw, come on, I don't look down on you, bro.
 

I can't blame you for boycotting Corsair. They're overpriced and have just average customer service. I mean, they ARE better at that than Asus, but that's a pretty low bar.

 

[Agall]

On 2/10/2024 at 6:54 AM, NorKris said:

 wow

Ewh

On 2/10/2024 at 8:25 AM, RevGAM said:

Does anyone else have any suggestions?

I mean, I could certainly add in clocks and electricity for other tests, but I'm not sure how useful (and which) for this test. Any feedback from anyone?

Cinebench and strategically blocking fans to see which temperature sensors respond best to suboptimal conditions. Establish a low point and high point, find which sensor responded best to that simulation condition.

 

In general, most temperature sensors are indirect measurements, since there's not a thermocouple/TRD actually on the die, so its using an approximation based on assumptions, which also has its own deviation at points. 

 

When testing temperature, you'd want performance and wattage metrics as well. I think there's enough data in Cinebench R23 to justify using that as a control, though with an AIO, you'd likely need to do 10 minute runs versus single runs to properly heat soak the AIO.

[RevGAM]

1 hour ago, Agall said:

Ewh

Cinebench and strategically blocking fans to see which temperature sensors respond best to suboptimal conditions. Establish a low point and high point, find which sensor responded best to that simulation condition.

 

In general, most temperature sensors are indirect measurements, since there's not a thermocouple/TRD actually on the die, so its using an approximation based on assumptions, which also has its own deviation at points. 

 

When testing temperature, you'd want performance and wattage metrics as well. I think there's enough data in Cinebench R23 to justify using that as a control, though with an AIO, you'd likely need to do 10 minute runs versus single runs to properly heat soak the AIO.

The only fans are the AIO and one exhaust...

[Agall]

1 minute ago, RevGAM said:

The only fans are the AIO and one exhaust...

Exactly. A great way to establish a control temperature.