Overclock.net › Forums › Intel › Intel CPUs › Skylake Delidded
New Posts  All Forums:Forum Nav:

Skylake Delidded - Page 78

post #771 of 821
I havent been in the skylake delid threads, but over past 5 years, most that delidded in the 3770k/4770k threads were getting much better temp drops with liquid metal. On that point we will just have to agree to disagree.

But agree on pumpout, so tim isnt a good option any way.

When I ran bare die, I only got about 5C better temps vs clp with ihs on (though the temp drop will always be proportional to power tested at), and the slightest amount of over pressure would cause ram channels etc to drop out from likely bending the pcb causing poor socket contact on some of pins, for me the hassle of bare die isnt worth it, given the majority of improvement of 22C drop (at 185W cpu power) of just replacing tim1 and leaving ihs.

As for difference between between 5 w/mk paste vs 40 wmk liquid metal, that is 8x conductance difference, also when you hit 20+ probably start getting into diminishing returns (how high do you need to no longer be the limiting factor). Whereas 40 vs 73 is only 1.8X as high and that assumes there is that difference given both are primarily gallium with small amounts tin/indium etc (and gallium is 40 w/mk).
My System
(15 items)
 
  
CPUMotherboardGraphicsRAM
6900K 4.4ghz  MSI X99A Krait Titan XP Gskill 32GB 3200 C14 
Hard DriveHard DriveCoolingOS
Samsung 960 Pro 1TB OS Samsung 850 pro 1TB games D5, XT45 280 rad, Nemesis 280 rad,  Windows 10 64bit 
MonitorPowerCase
Acer Predator XB271HU 144hz Seasonic Prime Titanium 850 Fractal Design R5 
  hide details  
Reply
My System
(15 items)
 
  
CPUMotherboardGraphicsRAM
6900K 4.4ghz  MSI X99A Krait Titan XP Gskill 32GB 3200 C14 
Hard DriveHard DriveCoolingOS
Samsung 960 Pro 1TB OS Samsung 850 pro 1TB games D5, XT45 280 rad, Nemesis 280 rad,  Windows 10 64bit 
MonitorPowerCase
Acer Predator XB271HU 144hz Seasonic Prime Titanium 850 Fractal Design R5 
  hide details  
Reply
post #772 of 821
Wow y'all are so over my head in the specifics.
Intel 6600k
(12 items)
 
  
CPUMotherboardGraphicsRAM
Intel Skylake I5 6600K Asus Z170-A Powercolor RX 480 Red Devil G.Skill Ripjaw V 
Hard DriveCoolingOSMonitor
SSD and Various Sata Corsair H110i GT Win 10 1x 29, 2x24 plus TV 
KeyboardPowerCaseMouse
Corsair K70 Lux EVGA 1050W 80+ Gold Certified Fully-Modular ATX... Thermaltake Tower 900 Logitech G502 Proteus 
  hide details  
Reply
Intel 6600k
(12 items)
 
  
CPUMotherboardGraphicsRAM
Intel Skylake I5 6600K Asus Z170-A Powercolor RX 480 Red Devil G.Skill Ripjaw V 
Hard DriveCoolingOSMonitor
SSD and Various Sata Corsair H110i GT Win 10 1x 29, 2x24 plus TV 
KeyboardPowerCaseMouse
Corsair K70 Lux EVGA 1050W 80+ Gold Certified Fully-Modular ATX... Thermaltake Tower 900 Logitech G502 Proteus 
  hide details  
Reply
post #773 of 821
Quote:
Originally Posted by opt33 View Post

I havent been in the skylake delid threads, but over past 5 years, most that delidded in the 3770k/4770k threads were getting much better temp drops with liquid metal. On that point we will just have to agree to disagree.

But agree on pumpout, so tim isnt a good option any way.

When I ran bare die, I only got about 5C better temps vs clp with ihs on (though the temp drop will always be proportional to power tested at), and the slightest amount of over pressure would cause ram channels etc to drop out from likely bending the pcb causing poor socket contact on some of pins, for me the hassle of bare die isnt worth it, given the majority of improvement of 22C drop (at 185W cpu power) of just replacing tim1 and leaving ihs.

As for difference between between 5 w/mk paste vs 40 wmk liquid metal, that is 8x conductance difference, also when you hit 20+ probably start getting into diminishing returns (how high do you need to no longer be the limiting factor). Whereas 40 vs 73 is only 1.8X as high and that assumes there is that difference given both are primarily gallium with small amounts tin/indium etc (and gallium is 40 w/mk).

Im shocked you got issues running bare die. For me the ekwb waterblock sits on the cpu socket rather then on the standoffs, it puts minimal pressure on the die itself and im sure with maybe another .05mm of pressure it would lower temps even more.

Most decent pastes arent 5 w/mK though, anything halfway decent that we buy/get is closer to 8 and higher. Even hydronaut and kryonaut i believe are in the 10-11 range while gc extreme is about 8.5. The basic cheap stuff is probably 5 but never seen anyone use that.

For me going bare die dropped me about 7°C but temps after load drop instantly and avg temps also dropped. Can't just look at max temps although thats how people check it out.

Truth is theres very few TIMs that do a decent job on the die. Again there is so many variables between cpus let alone identical cpus its pretty crazy. Could be that your IHS and block were fat flatter then mine are so you noticed bigger temps.

Personally i always start by saying, in my case. Its not a guarantee anyone will drop 20-25°C switching to CLU. Theres ton of people who switched over and saw minimal temp drops. Personally i think lapping the IHS and waterblock/heatsink prob gets more of a temp drop then switching between TIM to LM.
    
CPUMotherboardGraphicsRAM
R7 1700X asus crosshair 6 Sapphire R9 390 Nitro CMK16GX4M2B3200C16 
Hard DriveHard DriveCoolingCooling
Kingston HyperX 3K Seagate ST2000DM006 EKWB Supremacy Evo EK-XRES 100 Revo D5 - Acetal 
CoolingCoolingCoolingCooling
Alphacool GPX r9 390 m01 alphacool xt45 240 Alphacool xt45 360 Alphacool D5 PWM 
OSMonitorKeyboardPower
Windows 10 LG 29UM68 Corsair Strafe Corsair RM1000 
CaseMouseMouse Pad
Thermaltake X5 Corsair M65 Pro RGB Razer Goliathus Speed 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
R7 1700X asus crosshair 6 Sapphire R9 390 Nitro CMK16GX4M2B3200C16 
Hard DriveHard DriveCoolingCooling
Kingston HyperX 3K Seagate ST2000DM006 EKWB Supremacy Evo EK-XRES 100 Revo D5 - Acetal 
CoolingCoolingCoolingCooling
Alphacool GPX r9 390 m01 alphacool xt45 240 Alphacool xt45 360 Alphacool D5 PWM 
OSMonitorKeyboardPower
Windows 10 LG 29UM68 Corsair Strafe Corsair RM1000 
CaseMouseMouse Pad
Thermaltake X5 Corsair M65 Pro RGB Razer Goliathus Speed 
  hide details  
Reply
post #774 of 821
Quote:
Originally Posted by bluej511 View Post

As long as it doesnt move and break the bond i would think you can leave it on for a long time. If you remove the block/heatsink you might need to reapply but thats about it.

one quick question, how is cleaning? can I just use alcohol like standard paste?
post #775 of 821
Quote:
Originally Posted by bluej511 View Post

Im shocked you got issues running bare die. For me the ekwb waterblock sits on the cpu socket rather then on the standoffs, it puts minimal pressure on the die itself and im sure with maybe another .05mm of pressure it would lower temps even more.

Most decent pastes arent 5 w/mK though, anything halfway decent that we buy/get is closer to 8 and higher. Even hydronaut and kryonaut i believe are in the 10-11 range while gc extreme is about 8.5. The basic cheap stuff is probably 5 but never seen anyone use that.

For me going bare die dropped me about 7°C but temps after load drop instantly and avg temps also dropped. Can't just look at max temps although thats how people check it out.

Truth is theres very few TIMs that do a decent job on the die. Again there is so many variables between cpus let alone identical cpus its pretty crazy. Could be that your IHS and block were fat flatter then mine are so you noticed bigger temps.

Personally i always start by saying, in my case. Its not a guarantee anyone will drop 20-25°C switching to CLU. Theres ton of people who switched over and saw minimal temp drops. Personally i think lapping the IHS and waterblock/heatsink prob gets more of a temp drop then switching between TIM to LM.

yeah, some EK waterblocks are designed for bare die, probably why you have less issues than I had, Im not using EK. I ran bare die 8 years ago on a cpu that had thicker PCB without issue, but my 4790k was thin enough that along with my ghetto mount of my koolance 380i, just didnt like hassle vs gain.

Agreed, no doubt differences in cpus cause differences in results. I tested 5 waterblocks ? 8 years ago, got about 6-7C difference from another reviewer. We finally switched blocks/cpus and then realized it was the bow differences in 2 different samples of same waterblocks plus the shape differences of IHS on 2 cpus causing the difference in temps...so agree on not being adamant about what somewhere else will see with their particular cpu.
My System
(15 items)
 
  
CPUMotherboardGraphicsRAM
6900K 4.4ghz  MSI X99A Krait Titan XP Gskill 32GB 3200 C14 
Hard DriveHard DriveCoolingOS
Samsung 960 Pro 1TB OS Samsung 850 pro 1TB games D5, XT45 280 rad, Nemesis 280 rad,  Windows 10 64bit 
MonitorPowerCase
Acer Predator XB271HU 144hz Seasonic Prime Titanium 850 Fractal Design R5 
  hide details  
Reply
My System
(15 items)
 
  
CPUMotherboardGraphicsRAM
6900K 4.4ghz  MSI X99A Krait Titan XP Gskill 32GB 3200 C14 
Hard DriveHard DriveCoolingOS
Samsung 960 Pro 1TB OS Samsung 850 pro 1TB games D5, XT45 280 rad, Nemesis 280 rad,  Windows 10 64bit 
MonitorPowerCase
Acer Predator XB271HU 144hz Seasonic Prime Titanium 850 Fractal Design R5 
  hide details  
Reply
post #776 of 821
Quote:
Originally Posted by Damasterjj View Post

one quick question, how is cleaning? can I just use alcohol like standard paste?

Its easy, i do a quick stress test get it hot, shut down then clean it up. I only cleaned it up after a couple weeks and it came right off, idk how it would do after 2 years though lol.
    
CPUMotherboardGraphicsRAM
R7 1700X asus crosshair 6 Sapphire R9 390 Nitro CMK16GX4M2B3200C16 
Hard DriveHard DriveCoolingCooling
Kingston HyperX 3K Seagate ST2000DM006 EKWB Supremacy Evo EK-XRES 100 Revo D5 - Acetal 
CoolingCoolingCoolingCooling
Alphacool GPX r9 390 m01 alphacool xt45 240 Alphacool xt45 360 Alphacool D5 PWM 
OSMonitorKeyboardPower
Windows 10 LG 29UM68 Corsair Strafe Corsair RM1000 
CaseMouseMouse Pad
Thermaltake X5 Corsair M65 Pro RGB Razer Goliathus Speed 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
R7 1700X asus crosshair 6 Sapphire R9 390 Nitro CMK16GX4M2B3200C16 
Hard DriveHard DriveCoolingCooling
Kingston HyperX 3K Seagate ST2000DM006 EKWB Supremacy Evo EK-XRES 100 Revo D5 - Acetal 
CoolingCoolingCoolingCooling
Alphacool GPX r9 390 m01 alphacool xt45 240 Alphacool xt45 360 Alphacool D5 PWM 
OSMonitorKeyboardPower
Windows 10 LG 29UM68 Corsair Strafe Corsair RM1000 
CaseMouseMouse Pad
Thermaltake X5 Corsair M65 Pro RGB Razer Goliathus Speed 
  hide details  
Reply
post #777 of 821
Quote:
Originally Posted by opt33 View Post

yeah, some EK waterblocks are designed for bare die, probably why you have less issues than I had, Im not using EK. I ran bare die 8 years ago on a cpu that had thicker PCB without issue, but my 4790k was thin enough that along with my ghetto mount of my koolance 380i, just didnt like hassle vs gain.

Agreed, no doubt differences in cpus cause differences in results. I tested 5 waterblocks ? 8 years ago, got about 6-7C difference from another reviewer. We finally switched blocks/cpus and then realized it was the bow differences in 2 different samples of same waterblocks plus the shape differences of IHS on 2 cpus causing the difference in temps...so agree on not being adamant about what somewhere else will see with their particular cpu.

I tested mine with a straight edge and its def very very convex, even tested it with pressure paper and it def puts more pressure on teh cpu socket then the actual die. For skylake i def wouldnt go bare die without the aqua computer skylake shim though.

My 4690k def hasnt bent but then again theres def not much pressure on it so prob why.
    
CPUMotherboardGraphicsRAM
R7 1700X asus crosshair 6 Sapphire R9 390 Nitro CMK16GX4M2B3200C16 
Hard DriveHard DriveCoolingCooling
Kingston HyperX 3K Seagate ST2000DM006 EKWB Supremacy Evo EK-XRES 100 Revo D5 - Acetal 
CoolingCoolingCoolingCooling
Alphacool GPX r9 390 m01 alphacool xt45 240 Alphacool xt45 360 Alphacool D5 PWM 
OSMonitorKeyboardPower
Windows 10 LG 29UM68 Corsair Strafe Corsair RM1000 
CaseMouseMouse Pad
Thermaltake X5 Corsair M65 Pro RGB Razer Goliathus Speed 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
R7 1700X asus crosshair 6 Sapphire R9 390 Nitro CMK16GX4M2B3200C16 
Hard DriveHard DriveCoolingCooling
Kingston HyperX 3K Seagate ST2000DM006 EKWB Supremacy Evo EK-XRES 100 Revo D5 - Acetal 
CoolingCoolingCoolingCooling
Alphacool GPX r9 390 m01 alphacool xt45 240 Alphacool xt45 360 Alphacool D5 PWM 
OSMonitorKeyboardPower
Windows 10 LG 29UM68 Corsair Strafe Corsair RM1000 
CaseMouseMouse Pad
Thermaltake X5 Corsair M65 Pro RGB Razer Goliathus Speed 
  hide details  
Reply
post #778 of 821
Quote:
Originally Posted by bluej511 View Post

Its easy, i do a quick stress test get it hot, shut down then clean it up. I only cleaned it up after a couple weeks and it came right off, idk how it would do after 2 years though lol.

I want something that will last, and easy to clean:)
post #779 of 821
Hi Guys, I recently delidded my i7-6700K with really good results. After reading about skylake delidding on the forum and having a glass of wine (maybe two...) I went for the blade method. I first practiced on an old Core2 processor I had laying around.

I was very happy with the result, a massive 15 celsius drop with Gelid Extreme Thermal Paste.

I have a custom loop from EK, and I was more inclined to a temperature drop rather than taking my OC any higher since its my main rig and I need it for everyday usage.

CPU: i7-6700K
On Die-TIM: Gelid Extreme
IHS-TIM: Gelid Extreme
Core Voltage 1.35v
Temps Before Delid: 84 Average (84-85-87-80)
Temps After Delid: 69 Average (71-69-71-65)
Temp drop: 15 Celsius
Room Temperature: 23 Celsius


Before delidding. Real Temp was missreading the CPU Speed so I addeed a screenshot of CPUID.
The core voltage on the Bios is set to 1.35, CPUID read 1.376.


After delidding


Original TIM
post #780 of 821
Howdy, I'm just chipping in to point out that there is no need to create vents for the minuscule volume of air "trapped" beneath the IHS. It's not getting hot under there such that your CPU is overheating as a result. Nor any other perceived negative effects.
CPUs with epoxy type thermal coupling required those holes to facilitate curing of the epoxy when it was applied by the manufacturer. You do not need them unless you are also using a thermal compound which requires exposure to air to cure.
With the exception of lapping, don't grind away sections of, or drill holes in your IHS in the hope of improving it.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Intel CPUs
Overclock.net › Forums › Intel › Intel CPUs › Skylake Delidded