Cooling Quick hands on the Cryorig Cu C7

[VegetableStu]

Cryorig C7 cu
Cryorig C7
Noctua NH-L9i
Noctua NH-L9x65
ID-Cooling IS-40 v3
Thermolab LP53
Silverstone AR06
Silverstone AR11 was out of stock but I'll get it next week.

Can't wait!! Also hoping things get to the point where Dan and kmpkt joins the list

 

[IntoxicatedPuma]

the moment i get my thermal greasy hands on one they will be there.

 

[IntoxicatedPuma]

Just to compare with what Hudzzy posted earlier, I'll give my numbers today for 4.3ghz @1.173v using a 4690k. When fan speed was set to 50% it quickly failed this test, so I am using max fan speed for these. I also have turned off speed step/power saving modes so the CPU is running at 4.3ghz always. Motherboard is an Asus B85M-Gamer, with 2x4g HyperX Savage 2133mhz.
I'll give my numbers for AIDA 64 as Hudzzy has done already, and also numbers for Prime95 test as well.

All tests are done in an open air bench, I'm also in China and have an ambient air temp today of 21 Celsius, 60% humidity. I don't have clips so I'm using the stock fans, and using the Cryorig included thermal paste. I will try on MX-4 later and see if it makes a difference. Temps given are CPU package only, taken from HWMonitor Pro.

Prime95 Test:
Cryorig C7: FAILED (79C over ambient @ 112.2W Power Package)
Cryorig C7 cu: PENDING.....
Noctua NH-L9x65 (stock fan): 73C over ambient @ 121.0W Power Package
Noctua NH-L9x65 (NF-A9x14): 69C over ambient @ 120.4W Power Package
Noctua NH-L9x65 (NF-A9): 68C over ambient @ 114.1W Power Package


AIDA 64 Extreme FPU only (10 minute run time)
Cryorig C7: 57C over ambient @ 83.9W Power Package
Cryorig C7 cu: PENDING.....


uh ok.......I'm gonna look into this and see what's going on...
edited: I removed the C7 cu scores until I figure out what is happening. Right now mine is not performing as well as it would be expected.

edit 1: Looks OK? The Cu cooler isn't even warm to the touch but the paste application looks OK after pulling it off. I'll re-apply and try again.

edit 2:
Cleared the CPU & heat sink with Arcticlean remover & then with the purifier. Re-applied the Cryorig paste. Same results. I can't see any defect on the cooler, so I'm trying again on different CPU settings (4.0ghz @ 1.085v) to see how it compares.

 

[Hudzzy]

Just to compare with what Hudzzy posted earlier, I'll give my numbers today for 4.3ghz @1.173v using a 4690k. When fan speed was set to 50% it quickly failed this test, so I am using max fan speed for these. I also have turned off speed step/power saving modes so the CPU is running at 4.3ghz always. Motherboard is an Asus B85M-Gamer, with 2x4g HyperX Savage 2133mhz.
I'll give my numbers for AIDA 64 as Hudzzy has done already, and also numbers for Prime95 test as well.

All tests are done in an open air bench, I'm also in China and have an ambient air temp today of 21 Celsius, 60% humidity. I don't have clips so I'm using the stock fans, and using the Cryorig included thermal paste. I will try on MX-4 later and see if it makes a difference. Temps given are CPU package only, taken from HWMonitor Pro.

Prime95 Test:
Cryorig C7: FAILED (79C over ambient @ 112.2W Power Package)
Cryorig C7 cu: PENDING.....
Noctua NH-L9x65 (stock fan): 73C over ambient @ 121.0W Power Package
Noctua NH-L9x65 (NF-A9x14): 69C over ambient @ 120.4W Power Package
Noctua NH-L9x65 (NF-A9): 68C over ambient @ 114.1W Power Package


AIDA 64 Extreme FPU only (10 minute run time)
Cryorig C7: 57C over ambient @ 83.9W Power Package
Cryorig C7 cu: PENDING.....


uh ok.......I'm gonna look into this and see what's going on...
edited: I removed the C7 cu scores until I figure out what is happening. Right now mine is not performing as well as it would be expected.

edit 1: Looks OK? The Cu cooler isn't even warm to the touch but the paste application looks OK after pulling it off. I'll re-apply and try again.

edit 2:
Cleared the CPU & heat sink with Arcticlean remover & then with the purifier. Re-applied the Cryorig paste. Same results. I can't see any defect on the cooler, so I'm trying again on different CPU settings (4.0ghz @ 1.085v) to see how it compares.

Application looks ok to me.

BTW I used spreader included in MX-4 package to evenly spread paste in all runs, but normally I just use pea method and it's working fine for me.

 

[tinyitx]

The spread pattern looks OK. But it is possible that even with a good looking spread pattern, the contact pressure can still be insufficient.

 

[IntoxicatedPuma]

Here's an alternative theory: Don't point with the vents aimed at the RAM 1 module like a total noob.

 

[loader963]

The lp53 guys always preach that as well lol. Love to see some direct numbers against it and the c7cu.

 

[IntoxicatedPuma]

You'll have them in an hour.

 

[IntoxicatedPuma]

EDIT: One important thing I forgot to mention, the ambient temps when testing Cu C7 are around 20 degC as it's almost summer in my city now. Ambient temps during my Alu testing were around 10 degC back in November.

Is it possible for you to retest your Alu C7? I am not seeing major improvements on my C7 cu over the Alu one. I remember reading (i wish I could find the article now) that a thermodynamics student at a university had studied ambient temps on CPU's, and found that up until Haswell they made no real impact on Delta T but Haswell and later would have a gradually increasing impact on it. Unfortunately I can't find the article or remember how much the difference was

 

[Hudzzy]

Is it possible for you to retest your Alu C7? I am not seeing major improvements on my C7 cu over the Alu one. I remember reading (i wish I could find the article now) that a thermodynamics student at a university had studied ambient temps on CPU's, and found that up until Haswell they made no real impact on Delta T but Haswell and later would have a gradually increasing impact on it. Unfortunately I can't find the article or remember how much the difference was

I did retest my Alu and Cu C7 yesterday in case you missed it Quick hands on the Cryorig Cu C7. The thing is now I'm running out of Kryonaut, if I do another test then I don't have enough paste to finish the test Also replacing the cooler requires alomost taking apart my whole case, too much hassle involved...

 

[IntoxicatedPuma]

The lp53 guys always preach that as well lol. Love to see some direct numbers against it and the c7cu.

NH-L9x65 Stock - 52C over ambient
NH-L9x65 NF-A9 - 54C over ambient
NH-L9x65 NF-A9x14 - 56C over ambient
Cryorig C7 - 61C over ambient
Cryorig C7 cu -62C over ambient
Thermolab LP53 + NF-A9 - 63C over ambient
Scythe Jitte - 64C over ambient
ID-Cooling IS-40 v3- 65C over ambient
Thermolab LP53 + NF-A8 - 69C over ambient
Thermlab LP53 stock fan - 73C over ambient
TT Engine 27 - 79C over ambient

Tested at 4.0ghz & 1.085v, as 4.3ghz & 1.173v was too much for most of them to handle.

 

[Hudzzy]

NH-L9x65 Stock - 52C over ambient
NH-L9x65 NF-A9 - 54C over ambient
NH-L9x65 NF-A9x14 - 56C over ambient
Cryorig C7 - 61C over ambient
Cryorig C7 cu -62C over ambient
Scythe Jitte - 64C over ambient
Thermlab LP53 - 73C over ambient
TT Engine 27 - 79C over ambient

The LP53 results were on par with my past experience, had one few months ago and it (lp53+A9x14) performed slightly worse than my non-coated Alu C7+A9x14 combo

 

[IntoxicatedPuma]

Did you pull the fan from a noctua cooler or buy a retail one? I found they aren't the same and the ones that come on the coolers can run about 500rpm faster, as you can see on the L9x65 it helped by a decent amount (stock vs a9x14)

 

[Hudzzy]

Did you pull the fan from a noctua cooler or buy a retail one? I found they aren't the same and the ones that come on the coolers can run about 500rpm faster, as you can see on the L9x65 it helped by a decent amount (stock vs a9x14)

Mine is a retailed model.

 

[loader963]

I think I had read that is was a few degrees different somewhere else as well, but the more reviews the better.

My lp53 would go into mid-upper 80s (21C ambient) with a factory settings, SL delidded 8700k while gaming. Would love to have a 10* drop if I put her back into my sff build.

 

[Thehack]

Reminder that testing should be controlled. That is, control for noise or temps. You generally trade noise for temps, so one of those should be controlled.

Scenario A: allow temps to reach a certain point, let's say 80C at a certain power consumption. Measure the noise it outputs.

Scenario B: normalize the noise outputs and record temperature. Adjust fan curve to reach a certain point, ie 50dB and then record temperature.

I realize this makes it a bit difficult and time consuming, but it adds context besides just temperature.

 

[IntoxicatedPuma]

I test at 50% fan noise and 100%. I don't match noise levels because noise level doesn't always mean the fan will be obnoxious. The TT engine for example has a quite soft hum while the Scythe Jitte is obnoxious. I do record fan speed and the sound however.

 

[annasoh323]

Reminder that testing should be controlled. That is, control for noise or temps. You generally trade noise for temps, so one of those should be controlled.

Scenario A: allow temps to reach a certain point, let's say 80C at a certain power consumption. Measure the noise it outputs.

Scenario B: normalize the noise outputs and record temperature. Adjust fan curve to reach a certain point, ie 50dB and then record temperature.

I realize this makes it a bit difficult and time consuming, but it adds context besides just temperature.

Personally, I have no love lost if noise isn't controlled for, at least not for these tests. I want to know overall cooling effectiveness first and then worry about noise later. Temperature and noise are the two "outputs" I would say while factors like heatsink, fan, fan voltage levels, thermal paste type and application, system hardware and software specs, and ambient temperature are all "inputs." To me, noise is the output I don't really care about right now. I would phrase our problem statement (or, at least the one I want to see ) as, "What is cooling effectiveness of the C7 Cu compared to its nearest competitors?" Controlling noise/temps as you have suggested in your post would be necessary if the problem statement were (again, as I see it), "What is the cooling effectiveness with respect to noise of the C7 Cu compared to its nearest competitors?" or, "What are the noise/cooling characteristics of the C7 Cu compared to its nearest competitors?"

TL;DR: My take is to 1) control the inputs (testing setup including paste, settings, stress test, and ambient temps) 2) check the temperature and 3) let the noise be what it is.

 

[Thehack]

Personally, I have no love lost if noise isn't controlled for, at least not for these tests. I want to know overall cooling effectiveness first and then worry about noise later. Temperature and noise are the two "outputs" I would say while factors like heatsink, fan, fan voltage levels, thermal paste type and application, system hardware and software specs, and ambient temperature are all "inputs." To me, noise is the output I don't really care about right now. I would phrase our problem statement (or, at least the one I want to see ) as, "What is cooling effectiveness of the C7 Cu compared to its nearest competitors?" Controlling noise/temps as you have suggested in your post would be necessary if the problem statement were (again, as I see it), "What is the cooling effectiveness with respect to noise of the C7 Cu compared to its nearest competitors?" or, "What are the noise/cooling characteristics of the C7 Cu compared to its nearest competitors?"

TL;DR: My take is to 1) control the inputs (testing setup including paste, settings, stress test, and ambient temps) 2) check the temperature and 3) let the noise be what it is.

You're probably in the minority. You can always crank the fan to 100% and get some better temps but most people need the context of noise to compare it with other coolers.

 

[loader963]

The problem with noise is it is so subjective. What I may find tolerable another may not. Plus the case itself will have just as much effect as the cooler’s fan in most examples. Unless your testing on an open bench, which is something I prefer reviewers to avoid.