Water Cooling of CPU...Too Risky or just Paranoid?

[Indianatone]

It is usually 70F to 75F + by the evening up here, room has a vaulted ceiling and gets fairly warm. Need a portable AC unit in the summer to cool the room adequately. Currently room temp is 71.4F - 21.9C and the water cooler at idle shows as low as 13C to 21C and the air cooled machine shows 14C to 20C. It is 10F - -11.8C outside Central Indiana brrr.

 

[Lady Fitzgerald]

It is usually 70F to 75F + by the evening up here, room has a vaulted ceiling and gets fairly warm. Need a portable AC unit in the summer to cool the room adequately. Currently room temp is 71.4F - 21.9C and the water cooler at idle shows as low as 13C to 21C and the air cooled machine shows 14C to 20C. It is 10F - -11.8C outside Central Indiana brrr.

Something must be wrong with the way you are measuring your temps. There is no way an air or water cooler can cool below room temperature (unless you are using an evaporative water cooler).

 

[Indianatone]

I use the program Core Temp to monitor the temps.
Core Temp
With the Windows 7 Gadget....its so pretty....:roflmao:
I also have hardware monitor, CPU Zand GPU Z on here too.

 

[Lady Fitzgerald]

Something else is wacky, then. I also use CoreTemp (without the gadget; I have my temps posted in the system tray) and it does a reasonably accurate job of posting CPU temperatures (I've checked it against Speccy and Asus' AI Suite II). As I said before, there is no way any normal air or water cooler can cool a CPU down to, let alone below, the temperature of the air it is using.

 

[kbrady1979]

Something else is wacky, then. I also use CoreTemp (without the gadget; I have my temps posted in the system tray) and it does a reasonably accurate job of posting CPU temperatures (I've checked it against Speccy and Asus' AI Suite II). As I said before, there is no way any normal air or water cooler can cool a CPU down to, let alone below, the temperature of the air it is using.

This ^^

I built an Ivy Bridge system when it was first launched, in May of 2012, and paired the 3570K with the Intel All in One water cooler and it has been flawless. I left it at stock clock speeds for a month and then started OC'ing. Has been @ 4.4-4.6GHz since then.......working on 2 years now. I will never build another high performance rig without a closed loop system.

I will repeat what was said above......any component can fail, and will given half a chance, but the amount of people who have had a quality AIO unit spring a leak has to be a fraction of a percent. If it were common, it would be splashed across every forum on the internet. You also have the "idiot factor" to add in.......you know, those people who can't walk and breathe at the same time.....unfortunately, they can leave reviews as well.

 

[Lady Fitzgerald]

Something else is wacky, then. I also use CoreTemp (without the gadget; I have my temps posted in the system tray) and it does a reasonably accurate job of posting CPU temperatures (I've checked it against Speccy and Asus' AI Suite II). As I said before, there is no way any normal air or water cooler can cool a CPU down to, let alone below, the temperature of the air it is using.

This ^^

I built an Ivy Bridge system when it was first launched, in May of 2012, and paired the 3570K with the Intel All in One water cooler and it has been flawless. I left it at stock clock speeds for a month and then started OC'ing. Has been @ 4.4-4.6GHz since then.......working on 2 years now. I will never build another high performance rig without a closed loop system.

I will repeat what was said above......any component can fail, and will given half a chance, but the amount of people who have had a quality AIO unit spring a leak has to be a fraction of a percent. If it were common, it would be splashed across every forum on the internet. You also have the "idiot factor" to add in.......you know, those people who can't walk and breathe at the same time.....unfortunately, they can leave reviews as well.

What the cottonpickin' heck does that have to do with what I posted? I was talking about Tony's unrealistically low temps, not the merits of water cooling over air cooling.

For the record, I stipulate that water cooling is superior to air cooling and that water cooling is reasonably safe. However, water cooling does require more maintenance than air cooling and does have a slight chance of leakage. I feel that, as long as air cooling is capable of cooling adequately, why subject myself to the additional risk, no matter how slight, maintenance, and expense of water cooling?

 

[linnemeyerhere]

I know I'm headed towards my first full water loop in my main full tower rig one of these days nothing about the connections on that or a sealed loop system worry me about leakage if tested properly but I agree water and electronics aren't a match made in heaven if one touches the other.

LF, I agree there isn't a way in hell a running chip can be cooled with air or water to below ambient once left running a few minutes. To me if the temps show below ambient then it's an indicator of inaccurate software or sensors or both.

 

[sygnus21]

To me if the temps show below ambient then it's an indicator of inaccurate software or sensors or both.

Or extreme cooling... as in Nitrogen



Hi Cookie breaks several i5 world records using GIGABYTE Z87x-OC | TweakTown

But yeah outside of the extremes, no.

 

[Layback Bear]

Lady Fitzgerald your post is not ignored at least by me and linnemeyerhere.
If the air used to cool the system is room temp then that is the coolest your system will be.
I would suspect it's a sensor. Some sensors work poorly at low temps.
The manufacture really doesn't worry about low temps so the sensors are calibrated to be more accurate reading higher temps where the concern is.
Idle temps mean very little. Put the computer under a load and then compare the readings to others computers at the same ambient temps. That will allow one to know if their is a cooling problem.

Air or water cooling is just one of the choices that one makes.
Both have their plus and minus. Which ever one chooses please pick a quality one because your cpu deserves it.

 

[linnemeyerhere]

Quality is the name of the game on any cooler and just like PSU's they can't be too powerful !

 

[GeneO]

Gene, a 35% OC is fairly heavy.

But easily achievable with good air cooling, which was my point. You don't need liquid cooling for that.

 

[linnemeyerhere]

With Noctua and other air coolers the weight and stress on the mobo's would give me pause when transporting anywhere but once set up and stable they can do very well.

 

[Indianatone]

Jeannie,
I looked at the temp in the bios and it is reading 40C and I installed Asus temperature thermal radar and the CPU temp was 40C. Core Temp is not reading the FX cpu correctly. I have indeed taken out the water cooler and popped in the Hyper 212 with the Corsair fans and stuck the Cooler Master fan in the rear of the case. All seems to be the same. Same idle temps.

 

[GeneO]

Alan,

4.5 Ghz on a Sandy is a 36% OC and doesn't even require a heavy cooler - a Hyper 212 would do. For Ivy and Haswell, this is no longer true- you would need a heavier air cooler for probably an overclock < 30% and liquid above that, I agree.

 

[GeneO]

Jeannie,
I looked at the temp in the bios and it is reading 40C and I installed Asus temperature thermal radar and the CPU temp was 40C. Core Temp is not reading the FX cpu correctly. I have indeed taken out the water cooler and popped in the Hyper 212 with the Corsair fans and stuck the Cooler Master fan in the rear of the case. All seems to be the same. Same idle temps.

Asus measures the temp from some probe external to the CPU die. It is usually 10c lower than the core temperature (at least on Intel CPU).

 

[Britton30]

The temps on CPUs are read from a few pins/contacts on them. The values read from there goes through an algorithm to arrive at a value, core temp. Intel has a TJmax temp which they compute by drilling a hole in the CPU lid and take a temp in its exact center.
AMD uses what they call Package temp, I don't know what it is other than the average temp across all cores.

 

[Computer0304]

If you are going to switch to air cooling you should try looking at the Thermaltake FrioOCK or a similar performance cooler with 2+ fans.. It could be a it overpowered but that is a good thing in case you want to overclock in the furture.

 

[GeneO]

The temps on CPUs are read from a few pins/contacts on them. The values read from there goes through an algorithm to arrive at a value, core temp. Intel has a TJmax temp which they compute by drilling a hole in the CPU lid and take a temp in its exact center.
AMD uses what they call Package temp, I don't know what it is other than the average temp across all cores.

The BIOS temps are not the core temps nor the Tcase (It is Tcase that hypothetically measures a spot within the heatspreader -not TjMax with a "hole drilled" . Nobody implements a Tcase probe). What the BIOS reports as CPU temperature is measured with a probe located near the CPU , most likely under it, by an onboard chip like the Nuvoton. The probe measures around 10C cooler than the core temperatures. The core and package temperatures that you can readout from the CPU chip are from Intel probes within the CPU die. It is these core temps that TJMax is measured against. Fans are controlled by the BIOS based on the Nuvoton Temp probe.

 

[Britton30]

The temps on CPUs are read from a few pins/contacts on them. The values read from there goes through an algorithm to arrive at a value, core temp. Intel has a TJmax temp which they compute by drilling a hole in the CPU lid and take a temp in its exact center.
AMD uses what they call Package temp, I don't know what it is other than the average temp across all cores.

The BIOS temps are not the core temps nor the Tcase (It is Tcase that hypothetically measures a spot within the heatspreader -not TjMax with a "hole drilled" . Nobody implements a Tcase probe). What the BIOS reports as CPU temperature is measured with a probe located near the CPU , most likely under it, by an onboard chip like the Nuvoton. The probe measures around 10C cooler than the core temperatures. The core and package temperatures that you can readout from the CPU chip are from Intel probes within the CPU die. It is these core temps that TJMax is measured against. Fans are controlled by the BIOS based on the Nuvoton Temp probe.

Yeah, I screwed up there, I meant Tcase instead. Intel does "drill" a hole in the IHS to make up their standards.

These terms are related to processor temperature for desktop and mobile systems based on Intel® Processors. To allow optimal system operation and long-term reliability, the processor must not exceed the maximum case temperature specifications as defined by the applicable thermal profile.
Tcase is the temperature measurement using a thermocouple embedded in the center of the heat spreader. This initial measurement is done at the factory. Post-manufacturing, Tcase is calibrated by the BIOS, through a reading delivered by a diode between and below the cores.
Tjunction is synonymous with core temperatures, and calculated based on the output from the Digital Thermal Sensor (DTS) using the formula Tjunction = (Tjunction Max – DTS output).

Source

 

[kbrady1979]

Something else is wacky, then. I also use CoreTemp (without the gadget; I have my temps posted in the system tray) and it does a reasonably accurate job of posting CPU temperatures (I've checked it against Speccy and Asus' AI Suite II). As I said before, there is no way any normal air or water cooler can cool a CPU down to, let alone below, the temperature of the air it is using.

This ^^

I built an Ivy Bridge system when it was first launched, in May of 2012, and paired the 3570K with the Intel All in One water cooler and it has been flawless. I left it at stock clock speeds for a month and then started OC'ing. Has been @ 4.4-4.6GHz since then.......working on 2 years now. I will never build another high performance rig without a closed loop system.

I will repeat what was said above......any component can fail, and will given half a chance, but the amount of people who have had a quality AIO unit spring a leak has to be a fraction of a percent. If it were common, it would be splashed across every forum on the internet. You also have the "idiot factor" to add in.......you know, those people who can't walk and breathe at the same time.....unfortunately, they can leave reviews as well.

What the cottonpickin' heck does that have to do with what I posted? I was talking about Tony's unrealistically low temps, not the merits of water cooling over air cooling.

For the record, I stipulate that water cooling is superior to air cooling and that water cooling is reasonably safe. However, water cooling does require more maintenance than air cooling and does have a slight chance of leakage. I feel that, as long as air cooling is capable of cooling adequately, why subject myself to the additional risk, no matter how slight, maintenance, and expense of water cooling?

By saying "This ^^" I meant I was in agreement with you on the point of not being able to get the temps below ambient air temperature.

The rest of the post was to try to give my opinion of water cooling to the OP since this is HIS thread and he seemed to want advice.

Clear enough??