Overclock.net › Forums › Cooling › Specialized Cooling › Peltiers / TEC › CPU temp v.s. Chilled water temp
New Posts  All Forums:Forum Nav:

CPU temp v.s. Chilled water temp

post #1 of 13
Thread Starter 
Hey guys, I did a simple test to find out the CPU/Core temperature with different water temp.

Test setups:
i7 960 @4GHz, 1.32V, IHS removed and polished silicon die
xspc rasa water block, polished to 1500 grit
D5 pump on setting 5
ice in the loop...

I added ice to the pot and then let the water temp raise, and kept recording the temps. The radiator is not important here.

picture.php?albumid=4893&pictureid=26447
picture.php?albumid=4893&pictureid=26448

Test results:

picture.php?albumid=4893&pictureid=26446

Conclusion:
The thermal resistance between cpu die and water block plays a significant role when the cpu is stressed. The large amount of heat dissipating through small die-block contact area results in 30C core temp when water is at zero degrees.

Please correct me if anything is wrong.
Obsidian
(22 items)
 
CPUMotherboardGraphicsRAM
i7-3930K ASUS P9X79 Deluxe XFX HD 7970 Black Edition Double Dissipation Patriot Viper Xtreme Series 4x 4GB DDR3 2000 9-... 
Hard DriveHard DriveHard DriveCooling
OCZ RevoDrive3 X2 OCZ Vertex 4 WD 640GB  Swiftech 360 Radiator 
CoolingCoolingCoolingCooling
MCP 35X pump Swiftech Apogee XT waterblock Foxrena's Water Rack KOC TEC and Fan controller 
CoolingOSMonitorKeyboard
KOC 62mm TEC block Windows 7 Ultimate Dell UltraSharp IPS 27'' 2560x1440 Razor Black Widow Ultimate 
PowerCaseMouseAudio
Corsair AX1200i Corsair 800D Logitech G9X Logitech Z4 
  hide details  
Reply
Obsidian
(22 items)
 
CPUMotherboardGraphicsRAM
i7-3930K ASUS P9X79 Deluxe XFX HD 7970 Black Edition Double Dissipation Patriot Viper Xtreme Series 4x 4GB DDR3 2000 9-... 
Hard DriveHard DriveHard DriveCooling
OCZ RevoDrive3 X2 OCZ Vertex 4 WD 640GB  Swiftech 360 Radiator 
CoolingCoolingCoolingCooling
MCP 35X pump Swiftech Apogee XT waterblock Foxrena's Water Rack KOC TEC and Fan controller 
CoolingOSMonitorKeyboard
KOC 62mm TEC block Windows 7 Ultimate Dell UltraSharp IPS 27'' 2560x1440 Razor Black Widow Ultimate 
PowerCaseMouseAudio
Corsair AX1200i Corsair 800D Logitech G9X Logitech Z4 
  hide details  
Reply
post #2 of 13
Interesting graph but I admit that I am not exactly following your study.

Are you studying the benefit of the removal of the IHS?
I do not understand how the significance of removing the IHS from the CPU is reflected by the graph.

Anyway, I have some data of my own and I have taken the liberty to plot them onto your graph. As I only have four data points each for load and idle, extrapolation is not very accurate.
post #3 of 13
Thread Starter 
Quote:
Originally Posted by windfire;14267678 
Interesting graph but I admit that I am not exactly following your study.

Are you studying the benefit of the removal of the IHS?
I do not understand how the significance of removing the IHS from the CPU is reflected by the graph.

Anyway, I have some data of my own and I have taken the liberty to plot them onto your graph. As I only have four data points each for load and idle, extrapolation is not very accurate.


Mmm.... You data is interesting. But very reasonable.

These data actually give a lot of information. I think the slope and intercept of the fitting line depends on a number of factors:
1. room ambient temperature (affecting intercept)
2. the thermal resistance between the core and water (affecting slope and intercept). It can be affected by the waterblock, thermal paste and flow speed, etc.
3. How much heat is generated from the CPU (intercept difference of the load and idle line)
4. thermal conductance between CPU and adjacent components. (slope and intercept). Those components can become very hot when stressed, and cooling of this area is important, too.


Obviously radiator doesn't play a role in this graph.

So comparing the data points from your system and mine. Both CPUs put out about the same amount of heat. You cooling of the surrounding components around the CPU is much better than mine. And you thermal conductance between the core and water is a little worse than mine, probably because I use direct die method.
Those are just my speculations, correct me if I am wrong. smile.gif
Edited by foxrena - 7/19/11 at 7:08pm
Obsidian
(22 items)
 
CPUMotherboardGraphicsRAM
i7-3930K ASUS P9X79 Deluxe XFX HD 7970 Black Edition Double Dissipation Patriot Viper Xtreme Series 4x 4GB DDR3 2000 9-... 
Hard DriveHard DriveHard DriveCooling
OCZ RevoDrive3 X2 OCZ Vertex 4 WD 640GB  Swiftech 360 Radiator 
CoolingCoolingCoolingCooling
MCP 35X pump Swiftech Apogee XT waterblock Foxrena's Water Rack KOC TEC and Fan controller 
CoolingOSMonitorKeyboard
KOC 62mm TEC block Windows 7 Ultimate Dell UltraSharp IPS 27'' 2560x1440 Razor Black Widow Ultimate 
PowerCaseMouseAudio
Corsair AX1200i Corsair 800D Logitech G9X Logitech Z4 
  hide details  
Reply
Obsidian
(22 items)
 
CPUMotherboardGraphicsRAM
i7-3930K ASUS P9X79 Deluxe XFX HD 7970 Black Edition Double Dissipation Patriot Viper Xtreme Series 4x 4GB DDR3 2000 9-... 
Hard DriveHard DriveHard DriveCooling
OCZ RevoDrive3 X2 OCZ Vertex 4 WD 640GB  Swiftech 360 Radiator 
CoolingCoolingCoolingCooling
MCP 35X pump Swiftech Apogee XT waterblock Foxrena's Water Rack KOC TEC and Fan controller 
CoolingOSMonitorKeyboard
KOC 62mm TEC block Windows 7 Ultimate Dell UltraSharp IPS 27'' 2560x1440 Razor Black Widow Ultimate 
PowerCaseMouseAudio
Corsair AX1200i Corsair 800D Logitech G9X Logitech Z4 
  hide details  
Reply
post #4 of 13
It's hard to say anything remotely certain, but just a stab in the dark....

The core isn't the only thing that gets hot in the CPU, but by removing the IHS, you have greatly reduced the heat transfer surface of the chip so I can only assume that the additional heat that was transferred into the IHS instead of through the core will now entirely be going through the core...

Perhaps the addition of thermal pads surrounding the core might help this.
    
CPUMotherboardGraphicsRAM
e5645 @ 4.2Ghz P6X58D-E 5450 x2 6Gb Corsair Dominator 1600 
CoolingOSPowerCase
ALPENFOHN K2 Windows 7 Pro Antec Truepower Trio 650W Dell Optiplex 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
e5645 @ 4.2Ghz P6X58D-E 5450 x2 6Gb Corsair Dominator 1600 
CoolingOSPowerCase
ALPENFOHN K2 Windows 7 Pro Antec Truepower Trio 650W Dell Optiplex 
  hide details  
Reply
post #5 of 13
One factor you have left out of this conversation is the fact that BOTH of those chips have built in IMC's which both throw off heat on an overclock and as a result skew the temp readings from the cores. (This isn't true pre i3,5,7) The IHS mitigates this to an extent.

You are also forgetting that all temp sensors that read core temps are inherently flawed. It was more pronounced in the Wolfdale days but still holds true. AMD's temp sensors have always been wacky.

Temp sensors actually vibrate and when cold they tend to do so more slowly which also throws off the temps. Hell you can even over tighten a water block and get them to show all kinds of weird readings.

There is really only two ways to get a measurable temp for comparison of a TEC system on a given chip. You need a thermocouple in the base of the block and then compare that to the IHS temps. If you really want to get an idea of the core temps just add 10c to the IHS temps.
    
CPUMotherboardGraphicsRAM
4930k EVGA X79 Dark EVGA GTX 980 Hydrocopper GSKILL 
Hard DriveHard DriveOptical DriveCooling
6 x 300 GB Western Digital Velociraptors Raid 0 Western Digital Velociratpr 10k LG Custom Water Cooled 
OSMonitorKeyboardPower
Windows 7 Ultimate 64 AOC - 27" Widescreen Flat-Panel IPS Razer  EVGA 1300w G2 
CaseMouseAudio
Danger Den Torture Rack Razer Creative Sound Blaster X-FI Titanium HD 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
4930k EVGA X79 Dark EVGA GTX 980 Hydrocopper GSKILL 
Hard DriveHard DriveOptical DriveCooling
6 x 300 GB Western Digital Velociraptors Raid 0 Western Digital Velociratpr 10k LG Custom Water Cooled 
OSMonitorKeyboardPower
Windows 7 Ultimate 64 AOC - 27" Widescreen Flat-Panel IPS Razer  EVGA 1300w G2 
CaseMouseAudio
Danger Den Torture Rack Razer Creative Sound Blaster X-FI Titanium HD 
  hide details  
Reply
post #6 of 13
Thread Starter 
Quote:
Originally Posted by Crabid;14295243 
It's hard to say anything remotely certain, but just a stab in the dark....

The core isn't the only thing that gets hot in the CPU, but by removing the IHS, you have greatly reduced the heat transfer surface of the chip so I can only assume that the additional heat that was transferred into the IHS instead of through the core will now entirely be going through the core...

Perhaps the addition of thermal pads surrounding the core might help this.

Yes, you are right the cooling around the die is important. The thermal padding is a very good idea.

The IHS actually doesn't do much to transfer heat from the CPU to water block except from the silicon die. The edge is glued to the PCB with some kind of silicone or polymer, which I believe is not good heat conductor.

The tin between the die and IHS is especially pesky (thermal conductivity of tin is only 1/5 of pure copper), and often the soldering is poor. I saw a bunch of pores in the solder after I removed the IHS, some as large as 2mm...
Obsidian
(22 items)
 
CPUMotherboardGraphicsRAM
i7-3930K ASUS P9X79 Deluxe XFX HD 7970 Black Edition Double Dissipation Patriot Viper Xtreme Series 4x 4GB DDR3 2000 9-... 
Hard DriveHard DriveHard DriveCooling
OCZ RevoDrive3 X2 OCZ Vertex 4 WD 640GB  Swiftech 360 Radiator 
CoolingCoolingCoolingCooling
MCP 35X pump Swiftech Apogee XT waterblock Foxrena's Water Rack KOC TEC and Fan controller 
CoolingOSMonitorKeyboard
KOC 62mm TEC block Windows 7 Ultimate Dell UltraSharp IPS 27'' 2560x1440 Razor Black Widow Ultimate 
PowerCaseMouseAudio
Corsair AX1200i Corsair 800D Logitech G9X Logitech Z4 
  hide details  
Reply
Obsidian
(22 items)
 
CPUMotherboardGraphicsRAM
i7-3930K ASUS P9X79 Deluxe XFX HD 7970 Black Edition Double Dissipation Patriot Viper Xtreme Series 4x 4GB DDR3 2000 9-... 
Hard DriveHard DriveHard DriveCooling
OCZ RevoDrive3 X2 OCZ Vertex 4 WD 640GB  Swiftech 360 Radiator 
CoolingCoolingCoolingCooling
MCP 35X pump Swiftech Apogee XT waterblock Foxrena's Water Rack KOC TEC and Fan controller 
CoolingOSMonitorKeyboard
KOC 62mm TEC block Windows 7 Ultimate Dell UltraSharp IPS 27'' 2560x1440 Razor Black Widow Ultimate 
PowerCaseMouseAudio
Corsair AX1200i Corsair 800D Logitech G9X Logitech Z4 
  hide details  
Reply
post #7 of 13
Thread Starter 
Quote:
Originally Posted by Elloquin;14295377 
One factor you have left out of this conversation is the fact that BOTH of those chips have built in IMC's which both throw off heat on an overclock and as a result skew the temp readings from the cores. (This isn't true pre i3,5,7) The IHS mitigates this to an extent.

You are also forgetting that all temp sensors that read core temps are inherently flawed. It was more pronounced in the Wolfdale days but still holds true. AMD's temp sensors have always been wacky.

Temp sensors actually vibrate and when cold they tend to do so more slowly which also throws off the temps. Hell you can even over tighten a water block and get them to show all kinds of weird readings.

There is really only two ways to get a measurable temp for comparison of a TEC system on a given chip. You need a thermocouple in the base of the block and then compare that to the IHS temps. If you really want to get an idea of the core temps just add 10c to the IHS temps.

Thanks a lot for the information. I didn't know that the sensor is bad. It's very interesting that the temp sensor vibrates. Do you know why they make it that way? I thought it would be convenient to fabricate some temperature sensitive resistor or transistor on the die.
I will look into this issue...
Obsidian
(22 items)
 
CPUMotherboardGraphicsRAM
i7-3930K ASUS P9X79 Deluxe XFX HD 7970 Black Edition Double Dissipation Patriot Viper Xtreme Series 4x 4GB DDR3 2000 9-... 
Hard DriveHard DriveHard DriveCooling
OCZ RevoDrive3 X2 OCZ Vertex 4 WD 640GB  Swiftech 360 Radiator 
CoolingCoolingCoolingCooling
MCP 35X pump Swiftech Apogee XT waterblock Foxrena's Water Rack KOC TEC and Fan controller 
CoolingOSMonitorKeyboard
KOC 62mm TEC block Windows 7 Ultimate Dell UltraSharp IPS 27'' 2560x1440 Razor Black Widow Ultimate 
PowerCaseMouseAudio
Corsair AX1200i Corsair 800D Logitech G9X Logitech Z4 
  hide details  
Reply
Obsidian
(22 items)
 
CPUMotherboardGraphicsRAM
i7-3930K ASUS P9X79 Deluxe XFX HD 7970 Black Edition Double Dissipation Patriot Viper Xtreme Series 4x 4GB DDR3 2000 9-... 
Hard DriveHard DriveHard DriveCooling
OCZ RevoDrive3 X2 OCZ Vertex 4 WD 640GB  Swiftech 360 Radiator 
CoolingCoolingCoolingCooling
MCP 35X pump Swiftech Apogee XT waterblock Foxrena's Water Rack KOC TEC and Fan controller 
CoolingOSMonitorKeyboard
KOC 62mm TEC block Windows 7 Ultimate Dell UltraSharp IPS 27'' 2560x1440 Razor Black Widow Ultimate 
PowerCaseMouseAudio
Corsair AX1200i Corsair 800D Logitech G9X Logitech Z4 
  hide details  
Reply
post #8 of 13
I'd be interested to know what temperature the CPU really is at. Almost 70C with chilled water seems strange.
post #9 of 13
Thread Starter 
Quote:
Originally Posted by Willhemmens;14298709 
I'd be interested to know what temperature the CPU really is at. Almost 70C with chilled water seems strange.

From the graph, the core is 70C when water is at 36C. So it's not chilled water at that temperature. With 20C chilled water, the core will be below 55C.
Obsidian
(22 items)
 
CPUMotherboardGraphicsRAM
i7-3930K ASUS P9X79 Deluxe XFX HD 7970 Black Edition Double Dissipation Patriot Viper Xtreme Series 4x 4GB DDR3 2000 9-... 
Hard DriveHard DriveHard DriveCooling
OCZ RevoDrive3 X2 OCZ Vertex 4 WD 640GB  Swiftech 360 Radiator 
CoolingCoolingCoolingCooling
MCP 35X pump Swiftech Apogee XT waterblock Foxrena's Water Rack KOC TEC and Fan controller 
CoolingOSMonitorKeyboard
KOC 62mm TEC block Windows 7 Ultimate Dell UltraSharp IPS 27'' 2560x1440 Razor Black Widow Ultimate 
PowerCaseMouseAudio
Corsair AX1200i Corsair 800D Logitech G9X Logitech Z4 
  hide details  
Reply
Obsidian
(22 items)
 
CPUMotherboardGraphicsRAM
i7-3930K ASUS P9X79 Deluxe XFX HD 7970 Black Edition Double Dissipation Patriot Viper Xtreme Series 4x 4GB DDR3 2000 9-... 
Hard DriveHard DriveHard DriveCooling
OCZ RevoDrive3 X2 OCZ Vertex 4 WD 640GB  Swiftech 360 Radiator 
CoolingCoolingCoolingCooling
MCP 35X pump Swiftech Apogee XT waterblock Foxrena's Water Rack KOC TEC and Fan controller 
CoolingOSMonitorKeyboard
KOC 62mm TEC block Windows 7 Ultimate Dell UltraSharp IPS 27'' 2560x1440 Razor Black Widow Ultimate 
PowerCaseMouseAudio
Corsair AX1200i Corsair 800D Logitech G9X Logitech Z4 
  hide details  
Reply
post #10 of 13
Quote:
Originally Posted by foxrena;14298910 
From the graph, the core is 70C when water is at 36C. So it's not chilled water at that temperature. With 20C chilled water, the core will be below 55C.

Yeah I admit I read the chart wrong, was pretty late last night.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Peltiers / TEC
Overclock.net › Forums › Cooling › Specialized Cooling › Peltiers / TEC › CPU temp v.s. Chilled water temp