F5F Stay Refreshed Power Users Overclocking 4790K high temps Custom Loop

4790K high temps Custom Loop

4790K high temps Custom Loop

Pages (3): Previous 1 2 3
M
mkfrankie
Junior Member
44
05-19-2016, 09:56 AM
#21
Jossrik shared his observations about system behavior during light loads. He noted minimal activity at startup, yet the system attempted to run. When Windows was launched, the core temperature reached a high value, suggesting the operating system was active despite inactivity. He mentioned using media players and Chrome while running Windows, which contributed to elevated temperatures.

He referenced a guideline for radiator sizing: approximately 120mm per heat source plus one extra for safety. With three sources, this equates to around 481°C, placing him in the higher end of expectations. He preferred larger radiators and chose a model rated for over 1400W, which comfortably exceeded the 900°C range he was concerned about.

Jossrik explained that high temperatures in the 40s to 50s on a pair of 980Tis models running under water with modifications were acceptable after an hour of operation. This allowed ambient air to effectively cool the CPU and GPU, keeping those components within safe limits. He clarified that a strong pump would not significantly increase heat absorption from the block, as the cooler coolant maintained stable temperatures.

He also addressed concerns about the cooling system, noting that while the radiator could handle more heat than expected, the actual thermal load depended on the overall setup and usage patterns. He emphasized that ambient conditions played a key role in maintaining safe temperatures.

Jossrik advised seeking specific product details if the radiator model was different from what he assumed, and stressed that the system's performance was largely dependent on proper cooling design rather than just component ratings.
M
mkfrankie
05-19-2016, 09:56 AM #21

Jossrik shared his observations about system behavior during light loads. He noted minimal activity at startup, yet the system attempted to run. When Windows was launched, the core temperature reached a high value, suggesting the operating system was active despite inactivity. He mentioned using media players and Chrome while running Windows, which contributed to elevated temperatures.

He referenced a guideline for radiator sizing: approximately 120mm per heat source plus one extra for safety. With three sources, this equates to around 481°C, placing him in the higher end of expectations. He preferred larger radiators and chose a model rated for over 1400W, which comfortably exceeded the 900°C range he was concerned about.

Jossrik explained that high temperatures in the 40s to 50s on a pair of 980Tis models running under water with modifications were acceptable after an hour of operation. This allowed ambient air to effectively cool the CPU and GPU, keeping those components within safe limits. He clarified that a strong pump would not significantly increase heat absorption from the block, as the cooler coolant maintained stable temperatures.

He also addressed concerns about the cooling system, noting that while the radiator could handle more heat than expected, the actual thermal load depended on the overall setup and usage patterns. He emphasized that ambient conditions played a key role in maintaining safe temperatures.

Jossrik advised seeking specific product details if the radiator model was different from what he assumed, and stressed that the system's performance was largely dependent on proper cooling design rather than just component ratings.

A
adricrono
Junior Member
48
05-19-2016, 01:02 PM
#22
Oh yeah, about the new rule of thumb for radiators—didn't think much about it before, but now people seem more informed. As for Prime95, it could definitely be the reason behind what's happening. Most of the old articles are a bit outdated, so you're probably not seeing much fresh info.
A
adricrono
05-19-2016, 01:02 PM #22

Oh yeah, about the new rule of thumb for radiators—didn't think much about it before, but now people seem more informed. As for Prime95, it could definitely be the reason behind what's happening. Most of the old articles are a bit outdated, so you're probably not seeing much fresh info.

T
Tuetme
Senior Member
418
06-04-2016, 11:35 PM
#23
Yeah, on that radiator, that has to be a very high estimate. I can't picture how they figured out that number, but it seems pretty off. I’d really like to buy one just to check how it performs and what the results are.

There isn’t really a standard guideline because it tends to oversimplify TDP for graphics cards. Since mid-range cards usually use about 150 watts and top-tier ones can be double that, it’s tough to suggest a good radiator without knowing the exact TDP. We’re still talking about 'one graphics card,' but its power consumption could be twice as much or more, depending on the model.

Here’s where I try to generalize builds, though I still prefer doing the math to be sure. Of course, you should use quality fans if possible. On average, there’s around a 25% difference in volume between a 140mm radiator and a 120mm one, assuming their thicknesses are similar.
T
Tuetme
06-04-2016, 11:35 PM #23

Yeah, on that radiator, that has to be a very high estimate. I can't picture how they figured out that number, but it seems pretty off. I’d really like to buy one just to check how it performs and what the results are.

There isn’t really a standard guideline because it tends to oversimplify TDP for graphics cards. Since mid-range cards usually use about 150 watts and top-tier ones can be double that, it’s tough to suggest a good radiator without knowing the exact TDP. We’re still talking about 'one graphics card,' but its power consumption could be twice as much or more, depending on the model.

Here’s where I try to generalize builds, though I still prefer doing the math to be sure. Of course, you should use quality fans if possible. On average, there’s around a 25% difference in volume between a 140mm radiator and a 120mm one, assuming their thicknesses are similar.

L
lamperiere
Junior Member
7
06-21-2016, 03:22 PM
#24
rubix_1011 :
Yeah, on that radiator, that has to be a very high estimation. I can't imagine how they calculated that value, but that seems pretty unrealistic. I almost want to buy one just to test it and see what results I get.
There isn't really much of a rule of thumb because it overly generalizes TDP on graphics cards. Considering that mid-range cards run around 150watts and high end cards can double that, it becomes difficult to give a suggested radiator suggestion without knowing TDP. We're still discussing '1 graphics card' but it can be valued at double the energy draw and output, depending on model.
Here is kind of where I generalize builds, but I still like to do the calculations to make certain. Of course, you want quality fans to be used on the rads, if possible. On average, there is around 25% difference in volume on a single 140mm rad vs. a 120mm rad, given their thicknesses are the same.
Stock CPU - 1x120mm rad
Overclocked CPU - 2x120mm rad (or equivalent)
Highly OC CPU - up to 3x120mm rad (or equivalent)
GPU under 250watts - 2x120mm rad (or equivalent)
GPU over 250watts - 3x120mm rad (or equivalent)
My next build I want a broadwell extreme, maybe skylake extreme cpu (I know, it won't be for a while, but it's never too soon to start saving, and my system will last a while in everything I play). Sounds like I'm gonna have to go with a caselabs case
But I gotta say, I've never enjoyed computers more than tinkering with the water loop. I gotta measure my case and see if I can't fit a thicker radiator in there, I had 54mm radiators in there, but they were too thick and I had to get crummy fans, so I bought this one, maybe something in the middle, this radiator didn't score well on the flow test that I read, but everything better was much thicker, and that has a big effect on the temps. If I can find a middle of the road rad that I'm happy with I'll send you the one I have so you can play with it, just send it back when you're done.
L
lamperiere
06-21-2016, 03:22 PM #24

rubix_1011 :
Yeah, on that radiator, that has to be a very high estimation. I can't imagine how they calculated that value, but that seems pretty unrealistic. I almost want to buy one just to test it and see what results I get.
There isn't really much of a rule of thumb because it overly generalizes TDP on graphics cards. Considering that mid-range cards run around 150watts and high end cards can double that, it becomes difficult to give a suggested radiator suggestion without knowing TDP. We're still discussing '1 graphics card' but it can be valued at double the energy draw and output, depending on model.
Here is kind of where I generalize builds, but I still like to do the calculations to make certain. Of course, you want quality fans to be used on the rads, if possible. On average, there is around 25% difference in volume on a single 140mm rad vs. a 120mm rad, given their thicknesses are the same.
Stock CPU - 1x120mm rad
Overclocked CPU - 2x120mm rad (or equivalent)
Highly OC CPU - up to 3x120mm rad (or equivalent)
GPU under 250watts - 2x120mm rad (or equivalent)
GPU over 250watts - 3x120mm rad (or equivalent)
My next build I want a broadwell extreme, maybe skylake extreme cpu (I know, it won't be for a while, but it's never too soon to start saving, and my system will last a while in everything I play). Sounds like I'm gonna have to go with a caselabs case
But I gotta say, I've never enjoyed computers more than tinkering with the water loop. I gotta measure my case and see if I can't fit a thicker radiator in there, I had 54mm radiators in there, but they were too thick and I had to get crummy fans, so I bought this one, maybe something in the middle, this radiator didn't score well on the flow test that I read, but everything better was much thicker, and that has a big effect on the temps. If I can find a middle of the road rad that I'm happy with I'll send you the one I have so you can play with it, just send it back when you're done.

E
EzechielDLC
Member
67
06-22-2016, 10:40 PM
#25
I just upgraded to a Skylake 6700k, 16GB of TridentZ memory and a Gigabyte Z170N Wifi board...waiting on the GTX 1080's to see if I want one.
I was previously on an i7 2600 (non-k) and 8GB RAM. While this ran well, I have noticed the Skylake chip is a lot snappier on the same exact Windows 10 install.
E
EzechielDLC
06-22-2016, 10:40 PM #25

I just upgraded to a Skylake 6700k, 16GB of TridentZ memory and a Gigabyte Z170N Wifi board...waiting on the GTX 1080's to see if I want one.
I was previously on an i7 2600 (non-k) and 8GB RAM. While this ran well, I have noticed the Skylake chip is a lot snappier on the same exact Windows 10 install.

B
booopsmufc04
Junior Member
17
06-22-2016, 11:53 PM
#26
I recently upgraded to a Skylake 6700k, 16GB TridentZ memory, and a Gigabyte Z170N Wi-Fi board...holding off on the GTX 1080 until I decide.
Before that, I had an i7 2600 (non-K) with 8GB RAM. It worked fine, but the Skylake chip seems much faster with the same Windows 10 setup.
I've been weighing whether to stick with my current configuration or switch to a Skylake system. I already own a M.2 Samsung 950 Pro and recently sold my Strix 980ti, so I'm considering getting the 1080 and selling the 980ti. It’s been good for me—used EVGAs from a 970, and it still has its 10-year warranty.
Wouldn’t it be better to start fresh from the old 1155 (for the 2600k) to the 1151? I don’t blame you if you hate reinstalling Windows, but I do it often because I try a lot of things and eventually get bogged down. Still, I’ve found that 10 installations have kept my system much cleaner than any before.
B
booopsmufc04
06-22-2016, 11:53 PM #26

I recently upgraded to a Skylake 6700k, 16GB TridentZ memory, and a Gigabyte Z170N Wi-Fi board...holding off on the GTX 1080 until I decide.
Before that, I had an i7 2600 (non-K) with 8GB RAM. It worked fine, but the Skylake chip seems much faster with the same Windows 10 setup.
I've been weighing whether to stick with my current configuration or switch to a Skylake system. I already own a M.2 Samsung 950 Pro and recently sold my Strix 980ti, so I'm considering getting the 1080 and selling the 980ti. It’s been good for me—used EVGAs from a 970, and it still has its 10-year warranty.
Wouldn’t it be better to start fresh from the old 1155 (for the 2600k) to the 1151? I don’t blame you if you hate reinstalling Windows, but I do it often because I try a lot of things and eventually get bogged down. Still, I’ve found that 10 installations have kept my system much cleaner than any before.

Pages (3): Previous 1 2 3