EXTREME Overclocking Forums
Home | Reviews | Forums | Downloads | RealTime Pricing Engine | Folding Stats Contact Us


Go Back   EXTREME Overclocking Forums > Getting Started > Essential Tutorials, Guides, & Info

Welcome Guest Visitor! Please Register, It's Free and Fun To Participate!
The EXTREME Overclocking Forums are a place for people to learn how to overclock and tweak their PC's components like the CPU, memory (RAM), or video card in order to gain the maximum performance out of their system. There are lots of discussions about new processors, graphics cards, cooling products, power supplies, cases, and so much more!

You are currently viewing our boards as a "guest" which gives you limited access to view most discussions. You need to register before you can post: click the register link to proceed. Before you register, please read the forum rules. By joining our free community you will have access to post topics, communicate privately with other members (PM), respond to polls, upload your own pictures, and access many other special features. Registration is fast, simple, and absolutely free! To start viewing messages, select the forum that you want to visit from the selection below.

After you have registered and read the forum rules, you can check out the FAQ for more information on using the forum. We hope you enjoy your stay here!

Note To Spammers: We do not allow unsolicited advertising! Spam is usually reported & deleted within minutes of it being posted, so don't waste your time (or ours)!


Please Register to Post a Reply
 
Thread Tools
Old 07-07-2010, 06:45 AM   #61
OptyTrooper
Banned!
Banned
Don't ask why unless you want to join them.
Quote:
Originally Posted by ShrimpBrime View Post
Yes Opty and your using dual air. You have a lot of variables in your case defining a good seating or not. Air flow and room temps being almost as important. With dual liquid, your temps probably would never exceed 45c even over clocked.

So I suppose any methods would work. Just as long as there is some kind of cooling. And I imagine the two stock Phenom HSF's I sent you are not doing your system justice...
I suppose running 6 fans and having to wear a sweatshirt in my computer room may have something to do with my cooling...

The stock HSF's do fine actaully. I rarely do anything that will load all 8 cores or even 4. It's nice to know I have them if I need them though. I'm thinking very seriously about selling it on CL and building a 16 core just in case I need to run a small city or something.
United States  Offline
    Register to Reply to This Post
Old 07-07-2010, 12:30 PM   #62
ShrimpBrime
7GHZ CLUB
ShrimpBrime's Avatar
Senior Member
 
Posts: 7,310
Last Seen: Yesterday
From: nowhere
iTrader: 63 / 100%
Ya those Phenom HSF's are pretty decent with a little upgrade on the fan.

It's hot here in chicago this week so far. 90+ F and the humidity is killer like usual. I wonder how the Tri Phenom II 720BE core is doing. Off the top of my head, it's running 1.4v @ 3500mhz and under 40c @ load. HWMonitor shows 105w TDP. It's currently folding on liquid. The HSF it'd be well over 40c at load. That's all I know. Spread method or not. But since I'm not at home, I can't get a screeny. sorry.
United States  Offline
    Register to Reply to This Post
Old 07-07-2010, 11:35 PM   #63
WiCKeD
Still alive and well
WiCKeD's Avatar
Senior Moderator
 
Posts: 5,092
Last Seen: 07-10-2017
Age: 37
From: San Diego, CA
iTrader: 24 / 100%
I have played with all the methods and there generally is not a noticeable difference between them, once you get a good seal. The dot or line methods were popularized, because they guarantee good coverage in the middle and eliminate human error. You can put too little compound using the spread method.

Videos of how it spreads using an acrylic plate are one thing, but the TIM is going to spread more evenly once you apply real torque by screwing it down and it is heated.

Personally, I combine both the spread and line method. Spread only enough to fill the pitting in the heatspreader to form a smooth surface. Thin line to make sure there is contact with the core and to fill any major deformations. A small amount of excess will be squeezed out, but eventually the whole spreader is covered.
Aauria Hibedi  Offline
    Register to Reply to This Post
Old 07-07-2010, 11:49 PM   #64
Confessed
Poor Gamer
Confessed's Avatar
Senior Member
 
Posts: 1,016
Last Seen: 05-27-2011
From: North Texas
iTrader: 0 / 0%
Quote:
Originally Posted by drdeathx View Post
For the record.... The pea has been proven to be best method. Those who say it is negligible and didn't see a difference probably have a low end heatsink and or did not do the pea correctly. Yes the difference is not much but hey, 1 or 2 degrees means a bit to some when overclocking. I have put my waterblock on without reapplying TIM and there was a 4 degree difference(thats it). As it may be, the small pea has been proven over.. and over.. and over to be the best method on a flat heatsink.
I've tried them all and the credit card method always wins for me.
United States  Offline
    Register to Reply to This Post
Old 07-08-2010, 03:34 AM   #65
BigE4u
Yessss!
BigE4u's Avatar
Senior Member
 
Posts: 5,492
Last Seen: 01-09-2017
From: Texas
iTrader: 50 / 100%
Quote:
Originally Posted by Confessed View Post
I've tried them all and the credit card method always wins for me.
You "bride" your thermal paste to make sure it works right for you? j/k

Seriously, everyone has their favorite way of applying their own , it just boils down to what works best for you and your application.

I remember a time or two, i've always applied my thermal solution using the "spread" method, then placing a very small extra "dot" in the center of the IHS to guarantee the complete coverage i need for the cores.
United States  Offline
    Register to Reply to This Post
Old 07-08-2010, 04:35 AM   #66
mimart7
Je plie, donc je suis
mimart7's Avatar
Senior Member
 
Posts: 1,817
Last Seen: Today
Age: 52
From: NYC
iTrader: 1 / 100%
Given the many variables, like cpu, heatsink, etc, different methods work differently for many.
United States  Offline
    Register to Reply to This Post
Old 07-08-2010, 11:08 PM   #67
IcyLynx
Waiting for Diablo3
IcyLynx's Avatar
Senior Member
 
Posts: 970
Last Seen: 04-12-2012
Age: 34
From: in the corn
iTrader: 0 / 0%
What about a square shaped spot? If you start as a "pea", it circular, and as pressure is applied, it will expand more or less as a circle, and as chips are square... Just an idea to try, Im going to try it on my gpu
United States  Offline
    Register to Reply to This Post
Old 07-08-2010, 11:58 PM   #68
Odanez
I am Jean Luc Bacardi
Odanez's Avatar
Senior Member
 
Posts: 7,285
Last Seen: 07-23-2016
Age: 32
From: Bavaria
iTrader: 0 / 0%
I think if you make it a square, it will spread just about the same as a circle. Look at how the X-method spreads. After just a small amount of spreading it is already a square, and on full spread it's also not much different than how a blob spreads. The X-method is there so that more paste reaches the four corners of the heat spreader.
Germany  Offline
    Register to Reply to This Post
Old 07-10-2010, 10:27 AM   #69
IcyLynx
Waiting for Diablo3
IcyLynx's Avatar
Senior Member
 
Posts: 970
Last Seen: 04-12-2012
Age: 34
From: in the corn
iTrader: 0 / 0%
You would "draw" a square, just like "drawing" an "x", I didn't say it was a good way or not, just another idea to try. On another note, my gpu heatsink has heatpipes that make direct contact with the chip, what it also has is grooves between the heatpipes, like miniature canyons. Now the thermal paste is going to want to channel along these grooves, so the spread method I would say would be the best way in this situation? I'm prob thinking way too complicated, but am going to try spread, "x" and pea methods, see how it goes
United States  Offline
    Register to Reply to This Post
Old 07-10-2010, 03:32 PM   #70
BigE4u
Yessss!
BigE4u's Avatar
Senior Member
 
Posts: 5,492
Last Seen: 01-09-2017
From: Texas
iTrader: 50 / 100%
"X" marks the spot... need i say more?
United States  Offline
    Register to Reply to This Post
Old 07-11-2010, 01:34 AM   #71
Odanez
I am Jean Luc Bacardi
Odanez's Avatar
Senior Member
 
Posts: 7,285
Last Seen: 07-23-2016
Age: 32
From: Bavaria
iTrader: 0 / 0%
Quote:
Originally Posted by IcyLynx View Post
You would "draw" a square, just like "drawing" an "x", I didn't say it was a good way or not, just another idea to try. On another note, my gpu heatsink has heatpipes that make direct contact with the chip, what it also has is grooves between the heatpipes, like miniature canyons. Now the thermal paste is going to want to channel along these grooves, so the spread method I would say would be the best way in this situation? I'm prob thinking way too complicated, but am going to try spread, "x" and pea methods, see how it goes
Drawing a square would mean there is a large gap of air in the center. If you force the heatsink down, the air in the middle would most likely be trapped and you would have a large air bubble right in the middle, where you'd need contact the most. So tbh, I doubt this method will be any good.
Germany  Offline
    Register to Reply to This Post
Old 07-11-2010, 08:02 AM   #72
IcyLynx
Waiting for Diablo3
IcyLynx's Avatar
Senior Member
 
Posts: 970
Last Seen: 04-12-2012
Age: 34
From: in the corn
iTrader: 0 / 0%
I should have been more specific, "Draw" a solid square in the middle of the chip. Well got the new paste applied, the "x" method did not cover the entire chip, left north, south, east and west ends of the chip uncovered, but the gpu chip is alot bigger than a cpu.
Click image for larger version

Name:	P7100206.JPG
Views:	432
Size:	407.0 KB
ID:	150393
Seems pea and spread covered the chip best, i did try the "x" twice. Just fyi for anyone reapplying to their Fermi
United States  Offline
    Register to Reply to This Post
Old 07-12-2010, 05:50 PM   #73
ShrimpBrime
7GHZ CLUB
ShrimpBrime's Avatar
Senior Member
 
Posts: 7,310
Last Seen: Yesterday
From: nowhere
iTrader: 63 / 100%
Quote:
drdeathx: Too much will insulate the chip. The cores do not extend to the end of the chip....
It's an IHS (Integrated Heat Spreader). You can cover the entire top it won't hurt a darn thing.

The IHS plate is necessary in order to remove heat from the cores in a timely fashion. They used Copper, the #1 cheapest yet fastest heat removing metal.

If 2 surfaces are lapped near perfectly flat, you can use the spread method or dot or square method all you wan't. Use the least amount possible, there's no bowl in the IHS plate ir scratches nicks and dings to fill. Just be sure to have a good thermal paste.

You'd have better results just removing the IHS plate and mounting a water block directly to the cores. And you would Need to use the spread method because there's no IHS plate lol.

I already know I can't test this. I haven't a Cpu with an IHS plate..... Other wise I'd be up to it all... Perhaps drdeathx could be kind enough? Just please don't mount the Plexiglas with your fingers and hope to convince me lol.

Additional Comment:

Quote:
Originally Posted by IcyLynx View Post
I should have been more specific, "Draw" a solid square in the middle of the chip. Well got the new paste applied, the "x" method did not cover the entire chip, left north, south, east and west ends of the chip uncovered, but the gpu chip is alot bigger than a cpu.
Attachment 150393
Seems pea and spread covered the chip best, i did try the "x" twice. Just fyi for anyone reapplying to their Fermi
But how can you tell where air bubbles are?


965BE getting TIM'ed
Attached Thumbnails
Click image for larger version

Name:	DSC02456.jpg
Views:	192
Size:	102.8 KB
ID:	150412  

Last edited by ShrimpBrime; 07-12-2010 at 05:56 PM. Reason: To show proper TIM spread proceedures IHS'less Cpu.
United States  Offline
    Register to Reply to This Post
Old 07-13-2010, 02:56 PM   #74
ShrimpBrime
7GHZ CLUB
ShrimpBrime's Avatar
Senior Member
 
Posts: 7,310
Last Seen: Yesterday
From: nowhere
iTrader: 63 / 100%
I've never tried any pea method because I've spread like I was taught to do.

The reason why I may get that 1-2c better temps is simple.

Use your drawing in the previous post with the Dot and 4 corners showing. Those four corners probably do not contact the surface area of the cooling devise.

Using the actual picture taken by IcyLinx Attachment 150393 The TIM spread similar to the circle just in different directions. There is also spots where there is no heat being transferred. Where the TIM wasn't at.

I also try to mention that this may pertain to much higher wattage Cpu's like the 6000+ from 939-AM2, Dual Core Opteron 940pin (not AM2 I also pwn a couple of Opteron 270's also liquid cooled) those get pretty hot.

Basically anything 89w and up including getting over clocked IMO should be spread Especially if the surface area of the cooling apparatus covers the entire top of the IHS plate.

If the water block or HSF does not cover the entire surface area of the IHS plate, then any method will work fine because it's probably a 45w chip. Which I might add can be under clocked an cooled with out any heat sink at all around 1000mhz and under 1.0v.
United States  Offline
    Register to Reply to This Post
Old 07-13-2010, 03:40 PM   #75
Sephiroth
Apple Elitist
Sephiroth's Avatar
Senior Member
 
Posts: 6,924
Last Seen: 03-20-2017
Age: 29
iTrader: 17 / 100%
Quote:
Which I might add can be under clocked an cooled with out any heat sink at all around 1000mhz and under 1.0v.
ummm... I rather doubt that.. You might be able to get away with no fan, but if my 720QM (which *is* a 45W cpu) runs @ 44-49C @ 900mhz and .9V then I'm sure no heatsink at all will force a shutdown fairly quickly.
Otherwise you're saying that we were wasting our time putting heatsinks on everything up to the tualatin pentium 3's which were all sub 35W cpus.

for what it's worth, I thought I had used the spread method correctly, and it it appeared that I had not covered the entire core die and did have some bubbles.

Last edited by Sephiroth; 07-13-2010 at 04:09 PM.
United States  Offline
    Register to Reply to This Post
Old 07-13-2010, 05:29 PM   #76
ShrimpBrime
7GHZ CLUB
ShrimpBrime's Avatar
Senior Member
 
Posts: 7,310
Last Seen: Yesterday
From: nowhere
iTrader: 63 / 100%
Quote:
Originally Posted by Sephiroth View Post
ummm... I rather doubt that.. You might be able to get away with no fan, but if my 720QM (which *is* a 45W cpu) runs @ 44-49C @ 900mhz and .9V then I'm sure no heatsink at all will force a shutdown fairly quickly.
Otherwise you're saying that we were wasting our time putting heatsinks on everything up to the tualatin pentium 3's which were all sub 35W cpus.

for what it's worth, I thought I had used the spread method correctly, and it it appeared that I had not covered the entire core die and did have some bubbles.
No it needs the fan, just no heat sink. I have a sempron here I could boot up under clocked and then pull the water block. I still have the fan mod I made for it right here too.

I don't know about what P3's do. I do have some P4's, but I've never tried it. The die size is rather larger than todays cpu's......

I'll see about some pics with this sempron I got. Gimmie a bit....
United States  Offline
    Register to Reply to This Post
Old 07-13-2010, 06:11 PM   #77
ShrimpBrime
7GHZ CLUB
ShrimpBrime's Avatar
Senior Member
 
Posts: 7,310
Last Seen: Yesterday
From: nowhere
iTrader: 63 / 100%
That it may be touching, how well does it touch is the question. Being well experienced with tool and die measurements, I can tell you for a fact that those IHS plates and almost ALL HSF's are not as "flat" as people think. Even a mirror used for lapping isn't perfect for example.

Now, on to the NO HSF on a low end cpu under clocked about 800mhz but still running 1000mhz and under 1v a well.

Cpu on tonights experiment, Details as follows.
http://www.cpu-world.com/CPUs/K8/AMD...400CNBOX).html

Single Core Manila @ 1000mhz on 0.944v HT @ 800mhz stock and Ram DDR2 800mhz stock.

Loaded and running WPrime 1024m under 50c, read it and weep boyz

I've done plenty of testing, but when it comes down to it, I just need to lower my ambient temps to get my cooler to work a little better.....BUT it needs full thermal conduction. Does any one here follow yet? I am not even thinking of benching unless It's been spread.

Again all this is just My Opinion, like I said months ago. I show you, you take it however. Test all you want...

I'm on the PC right now no HSF.... Just a fan
Attached Thumbnails
Click image for larger version

Name:	DSC03165.jpg
Views:	183
Size:	208.0 KB
ID:	150430   Click image for larger version

Name:	Sempron 3400+ No HSF.JPG
Views:	182
Size:	144.9 KB
ID:	150431   Click image for larger version

Name:	Sempron 3400+ No HSF WPrime 1024 5%.JPG
Views:	165
Size:	152.7 KB
ID:	150432   Click image for larger version

Name:	DSC03163.jpg
Views:	178
Size:	135.5 KB
ID:	150433   Click image for larger version

Name:	DSC03166.jpg
Views:	185
Size:	256.0 KB
ID:	150434  

Click image for larger version

Name:	DSC03159.jpg
Views:	176
Size:	244.0 KB
ID:	150435   Click image for larger version

Name:	DSC03162.jpg
Views:	178
Size:	152.0 KB
ID:	150436  
United States  Offline
    Register to Reply to This Post
Old 07-13-2010, 06:26 PM   #78
Sephiroth
Apple Elitist
Sephiroth's Avatar
Senior Member
 
Posts: 6,924
Last Seen: 03-20-2017
Age: 29
iTrader: 17 / 100%
That's cheating, it has an Integrated heat Sink

also, a uni-core sempron is not a 4core (with another 4 exe cores) cpu *with* a MCH, so I call shenanigans.
United States  Offline
    Register to Reply to This Post
Old 07-13-2010, 06:50 PM   #79
ShrimpBrime
7GHZ CLUB
ShrimpBrime's Avatar
Senior Member
 
Posts: 7,310
Last Seen: Yesterday
From: nowhere
iTrader: 63 / 100%
And that entire IHS plate get's warm then cooled by... well in this case air off a fan. The IHS plate underneath has an entire spread of thermal compound across the core to ensure total heat dissipation.

So after watching manufacturers solder the IHS plates on, it was determined in my own mind that the thermal paste doesn't cut it. So the solder creates a total contact point, just like we create one with the HSF on the IHS plate with TIM.

It's just a spreader.... not a heat sink. IHS the S is Spreader not Sink. The thing with the fins and fan on it is actually what your sinking with. Just Fyi.
United States  Offline
    Register to Reply to This Post
Old 07-13-2010, 07:16 PM   #80
Sephiroth
Apple Elitist
Sephiroth's Avatar
Senior Member
 
Posts: 6,924
Last Seen: 03-20-2017
Age: 29
iTrader: 17 / 100%
Quote:
Originally Posted by ShrimpBrime View Post
And that entire IHS plate get's warm then cooled by... well in this case air off a fan. The IHS plate underneath has an entire spread of thermal compound across the core to ensure total heat dissipation.

So after watching manufacturers solder the IHS plates on, it was determined in my own mind that the thermal paste doesn't cut it. So the solder creates a total contact point, just like we create one with the HSF on the IHS plate with TIM.

It's just a spreader.... not a heat sink. IHS the S is Spreader not Sink. The thing with the fins and fan on it is actually what your sinking with. Just Fyi.
by common name maybe, but functionally, you're wrong.

It's moving heat from the core to a fluid medium, in this case air.. so It is a heat sink. It would be a heat spreader if it were transferring the heat to a secondary heat exchanger, but it's not in this case. You can call it a spreader if you want, but functionally it's not spreading anything since it's open and the heat from the core is being radiated into the open air. It might not do the job as well as a finned heatsink, but I'm thinking the processor would run a lot more hot without it.

Last edited by Sephiroth; 07-13-2010 at 07:26 PM.
United States  Offline
    Register to Reply to This Post
Sponsored Links:
Please Register to Post a Reply

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 -6. The time now is 06:45 AM.

Copyright ©2000 - 2016, Jelsoft Enterprises Ltd.
Powered by vBulletin
Copyright ©2000 - 2016, EXTREME Overclocking