Overclockers Australia Forums

OCAU News - Wiki - QuickLinks - Pix - Sponsors  

Go Back   Overclockers Australia Forums > Manufacturer-specific Forums > Intel x86 CPUs and chipsets

Notices

Reply
 
Thread Tools
Old 20th February 2017, 12:48 PM   #421
glnn_23
Member
 
Join Date: Jan 2012
Posts: 318
Default

Have tried RB 5.2ghz but am up to 1.4v and not quite there yet.
Will settle for 5.1ghz

Realbench 8 hours 1.325v.

glnn_23 - 7700k - L639F976 - 1.21v ? - Asus Max VIII Ranger - G.Skill 3600C15 - 5100/4500 - 1.325v(in bios)


Click to view full size!


Also ran Google Stressapptest in Linux Mint for 4 hours at same settings to check memory stability.
GSkill 4266c19 on this cpu/mb combo seems limited to 3600c15 at these volts.


Click to view full size!

Last edited by glnn_23; 20th February 2017 at 1:27 PM.
glnn_23 is offline   Reply With Quote
Old 20th February 2017, 10:13 PM   #422
Ratzz
Member
 
Ratzz's Avatar
 
Join Date: Mar 2013
Location: Cheltenham East 3192
Posts: 5,444
Default

Hey guys..
A couple of things I'd like to run past you.

On the image of the 7700K below (not mine) you'll note within the red area 4 metal points. Question.. what are they for? More important question.. what happens if you bridge them with liquid metal?


Click to view full size!


Now, IF bridging them is a bad thing.. seems likely?
How to prevent it? A smear of high temp silicon over them? Tape of some kind? Concrete ?


I saw a guy do a naked mount on a 7700K on youtube. The chip actually sits slightly below the socket, and when he first tried it, it wouldn't boot. He worked out that there was no pressure on the outer edges of the silicon, and it wasn't making proper contact with the CPU pins. He applied a triple layer of masking tape (??!!) in 4 strips on each side, and it worked perfectly and temps were substantially improved. I have to say, due to the chip being lower than the socket it sits in, it looks like the safest naked mount ever.

I'm thinking about it.. He used NTH1 though, and appeared not to care about those 4 metal points.



While attempting to find out more about those 4 points (unsuccessfully) I did come across an interesting discussion to which I didn't keep the link. I'll go through my history and see if I can find it, but basically a guy did some tests with an IHS, Delidded CPU, a digital micrometer, and some shims and stuff, and concluded...

The problem with the IHS is the gap varies between IHS and chip. Good chips (thermally) have very little gap, bad ones have a much larger gap.

Those which have small gaps cool pretty well out of the box, with very little TIM and close tolerances, those which don't simply have a large gap full of TIM. Of course traditional TIMS aren't particularly good thermal conductors, they simply are supposed to fill air gaps and provide a more even contact surface.

He found there are two solutions.. narrow the gap (less or no adhesive, lapping of the IHS on the surface that contacts the silicon) or bridging the gap. This is why liquid metal works better.. normal TIM fills the gap, liquid metal bridges it.

Make sense to you guys?
__________________
Don't mess with me. I can hurt you so fast you won't even feel it.

My Rig- Nokia 830, with silicone jacket upgrade.
Ratzz is offline   Reply With Quote
Old 20th February 2017, 10:36 PM   #423
glnn_23
Member
 
Join Date: Jan 2012
Posts: 318
Default

I just covered the dots with a coat of the silicon.

No need to lap the ihs

The intel tim was dry and did not contact both the chip and ihs

Removed all of the old silicon.

Once this is removed if the ihs is put in place it will spin freely on the chip.

I used the De8auer Delid mate and had a go at using a Dremel to machine it so I could relid the 7700k with it.

Not quite perfect though so I ended up using a piece of wood with a mortar on top for a weight to apply downward pressure.
The silicon was a slow curing one that I applied very very thinly and not completely all the way around.

It sat like this in a hot garage for 3 days.

Very happy with the results.
glnn_23 is offline   Reply With Quote
Old 20th February 2017, 10:40 PM   #424
shane41
Member
 
shane41's Avatar
 
Join Date: Nov 2008
Location: on the edge
Posts: 4,883
Default

The De8auer Delid mate has a reseating tool to fit the IHS back
__________________
Nothing to see here So I'm told often
shane41 is offline   Reply With Quote
Old 20th February 2017, 10:45 PM   #425
glnn_23
Member
 
Join Date: Jan 2012
Posts: 318
Default

Quote:
Originally Posted by shane41 View Post
The De8auer Delid mate has a reseating tool to fit the IHS back
That is for Skylake. Kabylake ihs slightly different. New Delid model though should be fine
glnn_23 is offline   Reply With Quote
Old 20th February 2017, 11:01 PM   #426
wraith.50
Member
 
Join Date: Sep 2016
Posts: 115
Default

Quote:
Originally Posted by Ratzz View Post
Hey guys..
A couple of things I'd like to run past you.

On the image of the 7700K below (not mine) you'll note within the red area 4 metal points. Question.. what are they for? More important question.. what happens if you bridge them with liquid metal?

pix

Now, IF bridging them is a bad thing.. seems likely?
How to prevent it? A smear of high temp silicon over them? Tape of some kind? Concrete ?


I saw a guy do a naked mount on a 7700K on youtube. The chip actually sits slightly below the socket, and when he first tried it, it wouldn't boot. He worked out that there was no pressure on the outer edges of the silicon, and it wasn't making proper contact with the CPU pins. He applied a triple layer of masking tape (??!!) in 4 strips on each side, and it worked perfectly and temps were substantially improved. I have to say, due to the chip being lower than the socket it sits in, it looks like the safest naked mount ever.

I'm thinking about it.. He used NTH1 though, and appeared not to care about those 4 metal points.



While attempting to find out more about those 4 points (unsuccessfully) I did come across an interesting discussion to which I didn't keep the link. I'll go through my history and see if I can find it, but basically a guy did some tests with an IHS, Delidded CPU, a digital micrometer, and some shims and stuff, and concluded...

The problem with the IHS is the gap varies between IHS and chip. Good chips (thermally) have very little gap, bad ones have a much larger gap.

Those which have small gaps cool pretty well out of the box, with very little TIM and close tolerances, those which don't simply have a large gap full of TIM. Of course traditional TIMS aren't particularly good thermal conductors, they simply are supposed to fill air gaps and provide a more even contact surface.

He found there are two solutions.. narrow the gap (less or no adhesive, lapping of the IHS on the surface that contacts the silicon) or bridging the gap. This is why liquid metal works better.. normal TIM fills the gap, liquid metal bridges it.

Make sense to you guys?
You don't want to get any liquid metal on any of those pads, as it will definitely cause a short of some kind.

For the person you mentioned who didn't care about the pads, that's because you said he was using NTH1 and that is non conductive, so he would not have to worry about them.

You can apply CLU onto the die and as long as you are careful, then you don't have to cover the pads at all.

But if you want to take that extra precaution, you can cover the pads with a couple of different options. One option I have seen people doing is covering it with clear nail polish and apparently that works well.

The option I prefer is to use Kapton Tape, which is a special polymer tape that can withstand very low and very high temperatures and is made for electrical applications, such as insulating and protecting electrostatic sensitive components.

This is a US site that sells all sorts of Kapton tape, but if you want to get it locally, then Jaycar sells their version of Kapton, which is quite good.

US Kapton Tape : https://www.kaptontape.com/

Jaycar Kapton Tape : https://www.jaycar.com.au/heat-resis...-50mm/p/NM2817

As for running it as a naked die, I don't think its worth the risk and hassle, for an extra I think 5 degrees or so, compared to running it with the IHS on.
wraith.50 is offline   Reply With Quote
Old 21st February 2017, 8:01 AM   #427
headin2001 Thread Starter
Member
 
headin2001's Avatar
 
Join Date: Apr 2007
Location: Sunshine Coast
Posts: 2,977
Default

Quote:
Originally Posted by Ratzz View Post
Hey guys..
A couple of things I'd like to run past you.

On the image of the 7700K below (not mine) you'll note within the red area 4 metal points. Question.. what are they for? More important question.. what happens if you bridge them with liquid metal?

pix

Now, IF bridging them is a bad thing.. seems likely?
How to prevent it? A smear of high temp silicon over them? Tape of some kind? Concrete ?


I saw a guy do a naked mount on a 7700K on youtube. The chip actually sits slightly below the socket, and when he first tried it, it wouldn't boot. He worked out that there was no pressure on the outer edges of the silicon, and it wasn't making proper contact with the CPU pins. He applied a triple layer of masking tape (??!!) in 4 strips on each side, and it worked perfectly and temps were substantially improved. I have to say, due to the chip being lower than the socket it sits in, it looks like the safest naked mount ever.

I'm thinking about it.. He used NTH1 though, and appeared not to care about those 4 metal points.



While attempting to find out more about those 4 points (unsuccessfully) I did come across an interesting discussion to which I didn't keep the link. I'll go through my history and see if I can find it, but basically a guy did some tests with an IHS, Delidded CPU, a digital micrometer, and some shims and stuff, and concluded...

The problem with the IHS is the gap varies between IHS and chip. Good chips (thermally) have very little gap, bad ones have a much larger gap.

Those which have small gaps cool pretty well out of the box, with very little TIM and close tolerances, those which don't simply have a large gap full of TIM. Of course traditional TIMS aren't particularly good thermal conductors, they simply are supposed to fill air gaps and provide a more even contact surface.

He found there are two solutions.. narrow the gap (less or no adhesive, lapping of the IHS on the surface that contacts the silicon) or bridging the gap. This is why liquid metal works better.. normal TIM fills the gap, liquid metal bridges it.

Make sense to you guys?
Long story short, running Skylake or Kabylake without the IHS is hampered by the actual socket.

In the corners of the socket, these bits are raised around 0.4mm. This stops you getting a proper mount as the cooler just rests on them and not the CPU Die. So, in order to run naked, not only do you have to shim around the CPU for even pressure on the PCB to allow the CPU to post, but you also have to either use a dremel or knife carefully to remove the tabs from the CPU socket to get the contact.

Personally, I wouldn't be doing it unless you knew the benefits were worth it, which in reality are only going to help those wanting a little more in extreme benching. And for that, you'll run more exotic cooling and not need naked die anyway.
headin2001 is offline   Reply With Quote
Old 21st February 2017, 8:21 PM   #428
Grom Hellscream
Member
 
Grom Hellscream's Avatar
 
Join Date: Jun 2004
Location: SE QLD
Posts: 2,902
Default

What sort of temp spikes do you guys see? Temps on mine are so random. I expect them to be a bit higher as I have an ITX setup in a cupboard (still working on a cooling solution so the doors are open under load atm), but when applying load the die temp will spike by 8-10 degrees instantly and stabilise at somewhat less than that. Might look into BIOS settings, I suspect additional voltage being added under load...
Grom Hellscream is offline   Reply With Quote
Old 22nd February 2017, 9:26 AM   #429
AstinGC90
Member
 
AstinGC90's Avatar
 
Join Date: Nov 2013
Location: FNQ
Posts: 392
Default

As an Intel user I am officially excited to see that AMD has bought genuine challenge to Intel this year. Amazing work from them on this architecture.

Sorry for linking WCCFT lol
http://wccftech.com/amd-ryzen-7-1700x-processor-tested/

1537cb score for R7 1700X @ 3.4Ghz, 16GB DDR4 @ 2133Mhz. Stock.
1184cb score for i7 7700K @ 5.4Ghz, 16GB DDR4 @ 3200mhz. My OC.

Regardless of the chip having double the cores this is impressive since they are only $80 USD apart to buy brand new.
__________________
I7-7700K Delided 5.2/4.8 1.42v daily OC, EKWC Hard Tube, 360mm rad x2 UV Blue, 16GB Corsair Vengeance 3.2Ghz, Aorus Z270X-G5, SamEVO500GB, 8TB, FractalDefineS, CorsairAX1200, 2 x Gigabyte GTX 1080ti EKWB 2050/12000 SLi. AsusPG279Q 165Hz, LogiG510S, LogiG502Proteus, LogiZ906, ATHM50X. Rig Pic
AstinGC90 is offline   Reply With Quote
Old 22nd February 2017, 9:37 AM   #430
mr_mordred2095
Member
 
Join Date: Mar 2004
Location: Brisbane
Posts: 3,815
Default

Quote:
Originally Posted by Grom Hellscream View Post
What sort of temp spikes do you guys see? Temps on mine are so random. I expect them to be a bit higher as I have an ITX setup in a cupboard (still working on a cooling solution so the doors are open under load atm), but when applying load the die temp will spike by 8-10 degrees instantly and stabilise at somewhat less than that. Might look into BIOS settings, I suspect additional voltage being added under load...
That's normal, even with static voltage. This is with 1.35v vcore, 1.2v vccpll

mr_mordred2095 is offline   Reply With Quote
Old 22nd February 2017, 10:27 AM   #431
mozmo
Member
 
Join Date: Jul 2001
Location: Sydney
Posts: 43
Default

Quote:
Originally Posted by AstinGC90 View Post
As an Intel user I am officially excited to see that AMD has bought genuine challenge to Intel this year. Amazing work from them on this architecture.

Sorry for linking WCCFT lol
http://wccftech.com/amd-ryzen-7-1700x-processor-tested/

1537cb score for R7 1700X @ 3.4Ghz, 16GB DDR4 @ 2133Mhz. Stock.
1184cb score for i7 7700K @ 5.4Ghz, 16GB DDR4 @ 3200mhz. My OC.

Regardless of the chip having double the cores this is impressive since they are only $80 USD apart to buy brand new.
It looks good but we haven't seen any game benchmark leaks and it's still unknown how well they overclock.

I have a feeling the chips will shine in multithread heavy benchmarks but real world gaming performance will disappoint and won't match the higher single thread performance of the 7700k especially @ 5ghz.
mozmo is offline   Reply With Quote
Old 22nd February 2017, 10:48 AM   #432
Kurosaki
Member
 
Kurosaki's Avatar
 
Join Date: Dec 2005
Location: Australia
Posts: 4,628
Default

Quote:
Originally Posted by AstinGC90 View Post

1537cb score for R7 1700X @ 3.4Ghz, 16GB DDR4 @ 2133Mhz. Stock.
1184cb score for i7 7700K @ 5.4Ghz, 16GB DDR4 @ 3200mhz. My OC.

Regardless of the chip having double the cores this is impressive since they are only $80 USD apart to buy brand new.
If you honestly think the R7 1700X is going to be $80 USB you sadly going to be seriously disappointed.

I know for a fact it will be north of exiesting 7700K prices.
Kurosaki is offline   Reply With Quote
Old 22nd February 2017, 11:01 AM   #433
AstinGC90
Member
 
AstinGC90's Avatar
 
Join Date: Nov 2013
Location: FNQ
Posts: 392
Default

Quote:
Originally Posted by Kurosaki View Post
If you honestly think the R7 1700X is going to be $80 USB you sadly going to be seriously disappointed.

I know for a fact it will be north of exiesting 7700K prices.
$80 USD more then a 7700K I was saying :-)

AMD Ryzen 7 1700X 8/16 $389 USD RRP
Core i7-7700K 4/8 $339 USD RRP

Actually $50 USD apart from a quick google.
__________________
I7-7700K Delided 5.2/4.8 1.42v daily OC, EKWC Hard Tube, 360mm rad x2 UV Blue, 16GB Corsair Vengeance 3.2Ghz, Aorus Z270X-G5, SamEVO500GB, 8TB, FractalDefineS, CorsairAX1200, 2 x Gigabyte GTX 1080ti EKWB 2050/12000 SLi. AsusPG279Q 165Hz, LogiG510S, LogiG502Proteus, LogiZ906, ATHM50X. Rig Pic

Last edited by AstinGC90; 22nd February 2017 at 11:03 AM.
AstinGC90 is offline   Reply With Quote
Old 22nd February 2017, 12:29 PM   #434
OJR
Member
 
OJR's Avatar
 
Join Date: Jan 2013
Posts: 2,593
Default

Quote:
Originally Posted by AstinGC90 View Post
$80 USD more then a 7700K I was saying :-)

AMD Ryzen 7 1700X 8/16 $389 USD RRP
Core i7-7700K 4/8 $339 USD RRP

Actually $50 USD apart from a quick google.
http://forums.overclockers.com.au/sh...postcount=1494

Things are going to be interesting....
OJR is offline   Reply With Quote
Old 22nd February 2017, 12:42 PM   #435
AstinGC90
Member
 
AstinGC90's Avatar
 
Join Date: Nov 2013
Location: FNQ
Posts: 392
Default

Quote:
Originally Posted by OJR View Post
http://forums.overclockers.com.au/sh...postcount=1494

Things are going to be interesting....
Wow nice so the pricing I found looks dead on keen to see the overclocks on this thing. Soldered spreader on these to
__________________
I7-7700K Delided 5.2/4.8 1.42v daily OC, EKWC Hard Tube, 360mm rad x2 UV Blue, 16GB Corsair Vengeance 3.2Ghz, Aorus Z270X-G5, SamEVO500GB, 8TB, FractalDefineS, CorsairAX1200, 2 x Gigabyte GTX 1080ti EKWB 2050/12000 SLi. AsusPG279Q 165Hz, LogiG510S, LogiG502Proteus, LogiZ906, ATHM50X. Rig Pic
AstinGC90 is offline   Reply With Quote
Reply

Bookmarks

Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +10. The time now is 12:42 AM.


Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2017, vBulletin Solutions, Inc.
OCAU is not responsible for the content of individual messages posted by others.
Other content copyright Overclockers Australia.
OCAU is hosted by Micron21!