New and Improved Watercooling Sticky - Post Your Systems Here

SonDa5

Supreme [H]ardness
Joined
Aug 20, 2008
Messages
7,415
What are your temps like with this slim rad?

Not bad. Prime95 after about an hour it's in mid 50's at 4.6GHZ with 1.33v on i5-2500k.

I also have the same model on top 120mm. Works fine.
 

Zoson

Supreme [H]ardness
Joined
Aug 4, 2001
Messages
6,115
Not bad. Prime95 after about an hour it's in mid 50's at 4.6GHZ with 1.33v on i5-2500k.

I also have the same model on top 120mm. Works fine.

Run linx all memory with win7 sp1 and avx enabled binaries. Then get back to us with your temps. Prime is no longer a real stress tool. 99% of all overclocks 'tested' with prime will fail linx.
 

ekuest

Supreme [H]ardness
Joined
Feb 23, 2009
Messages
6,091
Run linx all memory with win7 sp1 and avx enabled binaries. Then get back to us with your temps. Prime is no longer a real stress tool. 99% of all overclocks 'tested' with prime will fail linx.

99% of all overclocks "tested" with prime that fail linx will never give the user a single problem.
 

Patriot

[H]ard|DCer of the Month - March 2011/June 2013/De
Joined
Dec 15, 2010
Messages
2,496
99% of all overclocks "tested" with prime that fail linx will never give the user a single problem.

Possibly... But an hour is hardly testing... Eitherway... Just run fah for a few smp or bigger units..
 

Zoson

Supreme [H]ardness
Joined
Aug 4, 2001
Messages
6,115
Possibly... But an hour is hardly testing... Eitherway... Just run fah for a few smp or bigger units..

An hour of prime is like testing less than one linx loop. An hour of linx is like 48+ hours of prime, or 24 hours of FAH. Nothing stress tests like linx.

And I beg to differ on the '99% of overclocks tested with prime that fail linx will never give people an issue.' Those people either don't really stress their machines in general, or they are liars and are pretending they're stable. There's so much rampant lying that goes on in the overclocking community, it's disgusting.

Prime simply doesn't prove anything any more and is a worthless piece of software to the overclocking community now. It's only used by people that either don't know better, or are fakes.

If you think 'i don't need anything more' then you're a fake and should never discuss stability like you have any idea what you're talking about. If you say your pc doesn't crash - you don't use your machine fully, or you're a liar.
 

ekuest

Supreme [H]ardness
Joined
Feb 23, 2009
Messages
6,091
An hour of prime is like testing less than one linx loop. An hour of linx is like 48+ hours of prime, or 24 hours of FAH. Nothing stress tests like linx.

And I beg to differ on the '99% of overclocks tested with prime that fail linx will never give people an issue.' Those people either don't really stress their machines in general, or they are liars and are pretending they're stable. There's so much rampant lying that goes on in the overclocking community, it's disgusting.

Prime simply doesn't prove anything any more and is a worthless piece of software to the overclocking community now. It's only used by people that either don't know better, or are fakes.

If you think 'i don't need anything more' then you're a fake and should never discuss stability like you have any idea what you're talking about. If you say your pc doesn't crash - you don't use your machine fully, or you're a liar.

then why are you talking about 99%? you extreme overclockers are the 1%, not the 99%. the vast majority (99%) will never stress their cpu to this extent. what kind of gaming/browsing/torrents/whatever could possibly stress a 2600 or even a 2500 to the point where its more stressed than running prime95? there are VERY FEW instances where us normal users will exceed the stress put on the cpu by prime95. it is absolutely not the most intense stressing tool, but it is still good enough for 99% of users. hell, ive seen a lot of people who just play crysis on all max settings at the most intense parts of the game for half an hour to stress test their cpus. sure its not a great way to test, but its good enough for them. 99% of users would be happier with 4.5GHz thats prime95 stable than 4.2GHz thats linx stable.

tl;dr linx is a more powerful stress testing tool, but prime95 is still more demanding on their cpu than 99% of users will ever encounter using their machines. imo its a perfectly acceptable stress tool for an average user. anyone who might actually put more stress on their cpu than prime95 better be knowledgeable enough that they should know to use linx.
 

Tsumi

[H]F Junkie
Joined
Mar 18, 2010
Messages
13,386
I've had situations where LinX returned no errors after ~10 runs with 24gb memory while P95 resulted in a BSOD within ~30 minutes.
 

Zoson

Supreme [H]ardness
Joined
Aug 4, 2001
Messages
6,115
then why are you talking about 99%? you extreme overclockers are the 1%, not the 99%. the vast majority (99%) will never stress their cpu to this extent. what kind of gaming/browsing/torrents/whatever could possibly stress a 2600 or even a 2500 to the point where its more stressed than running prime95? there are VERY FEW instances where us normal users will exceed the stress put on the cpu by prime95. it is absolutely not the most intense stressing tool, but it is still good enough for 99% of users. hell, ive seen a lot of people who just play crysis on all max settings at the most intense parts of the game for half an hour to stress test their cpus. sure its not a great way to test, but its good enough for them. 99% of users would be happier with 4.5GHz thats prime95 stable than 4.2GHz thats linx stable.

tl;dr linx is a more powerful stress testing tool, but prime95 is still more demanding on their cpu than 99% of users will ever encounter using their machines. imo its a perfectly acceptable stress tool for an average user. anyone who might actually put more stress on their cpu than prime95 better be knowledgeable enough that they should know to use linx.
The above is just not true. The vast majority of games are now more stressful than prime95. They may not load up all your cores, but they'll stress the cores they do load more than prime can. That's why we see messages all over the forums whining about how they're prime stable but keep getting crashes playing xy game.

I've had situations where LinX returned no errors after ~10 runs with 24gb memory while P95 resulted in a BSOD within ~30 minutes.
That's because there's a bug in boomfield/lynnfield/gulftown over 9216MB memory size. To test more ram than 9216MB, you need to disable hyperthreading. First you stress your ram with HT off, make sure it's stable, then you stress with 9216mb problem size and HT on to verify full stability.

This was an issue with the actual CPU HT scheduler itself. It has been fixed in SB and SBE.
 

SonDa5

Supreme [H]ardness
Joined
Aug 20, 2008
Messages
7,415
I'll try the linx.


The radiators do work well though. I have an over clocked GTX470 on the same loop. Played UT3 for hours yesterday at 5ghz without any problems.

Can't play stability testers but I do understand how important stability and temperature testing is.
 

Zoson

Supreme [H]ardness
Joined
Aug 4, 2001
Messages
6,115
I'll try the linx.


The radiators do work well though. I have an over clocked GTX470 on the same loop. Played UT3 for hours yesterday at 5ghz without any problems.

Can't play stability testers but I do understand how important stability and temperature testing is.

When you do, follow this guide:
http://www.overclock.net/t/645392/h...inx-ibt-properly-an-explanation-maybe-a-guide

Also keep in mind the things I noted above, as not all of them are mentioned in the guide.
 

Nickbaldwin86

2[H]4U
Joined
Nov 8, 2009
Messages
2,135
UT3 is the best stress test EVAR!!!!!! I am sure it is using 25% CPU lol

you guys should move your annoying off-topicness to another thread and post more pictures of water cooling
 

Zoson

Supreme [H]ardness
Joined
Aug 4, 2001
Messages
6,115
Don't have time right now to read all of that. Can linpack be ran from a bootable USB key in a pure environment?
Nope. Easiest way to do what you want is to boot diagnostic mode.

UT3 is the best stress test EVAR!!!!!! I am sure it is using 25% CPU lol

you guys should move your annoying off-topicness to another thread and post more pictures of water cooling

K, my current project: Watercool EVGA 560 Ti 448 Classified:


 

CMadki4

Supreme [H]ardness
Joined
Feb 3, 2011
Messages
4,438
Shameless plug, but still WC related: I'm selling my loop that I posted on here a few pages back here. Mods, please don't eat my face off! :D
 

LoBoxJa

Limp Gawd
Joined
Aug 5, 2008
Messages
153
just re-did the loop and changed the liquid, not as fancy as the rest but it works

 

Zoson

Supreme [H]ardness
Joined
Aug 4, 2001
Messages
6,115
Got my gpu's cooled!




2x EVGA 560Ti 448 Core Classified
2x Swiftech MCW82
2x Swiftech GTX480-HS Unisinks
1x Swiftech 3way SLI Water Bridge
4x 30mm Bitspower C63 1/2" IDextenders
4x Bitspower 1/2" Quick Disconnects
2x Hiper UV Red 1800RPM(they're quiet too!) 80x15mm fans

Idling at about 35c, full load about 55 with furmark!
920/1840/2340 core/shader/mem! Mem slider is maxed out in evga precision!
 

SticKx911

2[H]4U
Joined
Mar 14, 2004
Messages
2,450
A friend told me that those classifieds use the same pcb layout as a gtx 480 (you can just use 480 waterblocks), is that true?
 

Zoson

Supreme [H]ardness
Joined
Aug 4, 2001
Messages
6,115
A friend told me that those classifieds use the same pcb layout as a gtx 480 (you can just use 480 waterblocks), is that true?

That's exactly right.

The thing is the memory and power components are completely UNCOOLED using the stock heatsink. So adding a unisink with a fan is really all that's needed for those parts, and then you watercool the core... I'm able to max out the memory frequency in evga precision - just a measure so you can see how unnecessary FC blocks are.

I was pleasantly surprised how quiet these Hiper 80x15mm fans are!
 

ryuji

2[H]4U
Joined
Jun 11, 2004
Messages
3,115
That's exactly right.

The thing is the memory and power components are completely UNCOOLED using the stock heatsink. So adding a unisink with a fan is really all that's needed for those parts, and then you watercool the core... I'm able to max out the memory frequency in evga precision - just a measure so you can see how unnecessary FC blocks are.

I was pleasantly surprised how quiet these Hiper 80x15mm fans are!

Actually, you would be surprised how many cards you can cool with generic gpu water blocks and a 80mm fan, i been doing it for almost 10 years now. you just cant do those multi gpu cards without serious modifications. Its a lot easier if you can locate gpus that have segmented/missing vreg cooler from the gpu core. Im still going strong with my DD Maze4 gpu. 5950 ultra, 6800gt, x1800xt, hd2900xt, and now my 5870. couldnt water cool my 3870 x2
 

Zoson

Supreme [H]ardness
Joined
Aug 4, 2001
Messages
6,115
Actually, you would be surprised how many cards you can cool with generic gpu water blocks and a 80mm fan, i been doing it for almost 10 years now. you just cant do those multi gpu cards without serious modifications. Its a lot easier if you can locate gpus that have segmented/missing vreg cooler from the gpu core. Im still going strong with my DD Maze4 gpu. 5950 ultra, 6800gt, x1800xt, hd2900xt, and now my 5870. couldnt water cool my 3870 x2

i had a pair of mcw60 v2's, got them for my gtx275's....and sold them both... ended up regretting it when i got these cards. Now have some mcw82's. On the fence about the 'improvement'... The decrease in flow for the minor increase in thermal performance doesn't really matter much to me, since my cards never get over 55c. I wouldn't mind 60c at all.

I have to say: buy a unisink. They're great! I benched 3dmark last night at 966/1932/2340 and got 12789. highest score for gtx 560ti 448's in sli! There are no 3 card submissions yet, and I'm interested in seeing how close they can come to some 580's. I wonder how i'm stacking up against good 2 card 570 and 580 setups?

http://3dmark.com/3dm11/2559368
 

SticKx911

2[H]4U
Joined
Mar 14, 2004
Messages
2,450
I used a universal thermaltake for some time. it was a pos, but it did an amazing job on at least 3 different cards. So I totally agree. Not as awesome looking as a full block, but just as silent.
 

FireBean

Gawd
Joined
Apr 12, 2010
Messages
994
Made a few changes and added some lighting. Had the videocard block hooked up backwards. Whoops!

 

Wingless92

Limp Gawd
Joined
Jan 22, 2011
Messages
333
My Main Rig that I use for work and play. Thought about doing WC for a long time, so I finally decided to drop that hammer. The fans are stupid quiet but they are not pushing enough air through the radiator, so, new, more CFM fans are on the way. Hopefully it will drop my temps even more. We will see, thank god for fan controllers.











 

R-Type

[H]ard|DCer of the Month - October 2011
Joined
Mar 6, 2006
Messages
2,623
My Main Rig that I use for work and play. Thought about doing WC for a long time, so I finally decided to drop that hammer. The fans are stupid quiet but they are not pushing enough air through the radiator, so, new, more CFM fans are on the way. Hopefully it will drop my temps even more. We will see, thank god for fan controllers.












Beautiful, what case is this?
 

Wingless92

Limp Gawd
Joined
Jan 22, 2011
Messages
333
Beautiful, what case is this?

It's a CaseLabs M10 with the 64mm Extended top, which really came in handy with the WC setup.

Thinking about doing the cards next but I think i'll have to add another loop and that's something that I don't want to do.

I am pretty happy with it right now. I'm running the Intel Burn Test right now seeing what my temps are with a i7 920 at 4Ghz. It's staying right around 60C which seems to be about normal with these older i7 chips.

Overall though i'm stupid happy with the case and the WC setup.
 

gs274

[H]ard|Gawd
Joined
Jul 5, 2005
Messages
1,313
Oh how my wife hates you...I must now go shopping at caselabs.
 

Wingless92

Limp Gawd
Joined
Jan 22, 2011
Messages
333
It is a great case. I came from a Corsair 800D but I wanted something bigger. Was a snap installing everything but they are a bit pricey. As for not having USB/Audio jacks on the front panel, I actually like it. Makes for a cleaner look IMO. Kinda sucks that they make you pay for it if you want it but I would say that is the least of your worries with this case.

The removable Mobo tray is a godsend. Makes for installing everything that much easier. I am looking at the extended base but I am not sure if I would use it or not. Time will tell I guess.

Thanks everyone for the nice comments. I am been lurking on these forums for awhile and have used some ideas from other pics that I have seen.
 

jdk

Limp Gawd
Joined
Nov 23, 2006
Messages
437
My Main Rig that I use for work and play. Thought about doing WC for a long time, so I finally decided to drop that hammer. The fans are stupid quiet but they are not pushing enough air through the radiator, so, new, more CFM fans are on the way. Hopefully it will drop my temps even more. We will see, thank god for fan controllers.













Nice setup! Which fan controller are you using? How do you like it?
 

traderj

Limp Gawd
Joined
Jan 13, 2005
Messages
321
Is that the Bitspower res? How are you not getting air bubbles from the top inlet? Mine should be coming today but I thought it would create air bubbles if you use the top

Looks to be just a fill tube. His inlet is coming from the video card. He's expelling behind the tray and around into the 2x120 rad.
 
Joined
Oct 6, 2010
Messages
572
Is that the Bitspower res? How are you not getting air bubbles from the top inlet? Mine should be coming today but I thought it would create air bubbles if you use the top

Looks to be just a fill tube. His inlet is coming from the video card. He's expelling behind the tray and around into the 2x120 rad.


Yes, the top tube is just a filling tube. I have it capped off with a 1/2" barb/plug.
The return line is running from the graphics card, then I'm pumping water out through a 90 degree fitting into the lower radiator, and from there the loop is pretty self explanatory.

 

SonDa5

Supreme [H]ardness
Joined
Aug 20, 2008
Messages
7,415
New MSI GTX 560ti 448 TFiii PE/OC waiting for somebody to build a high quality full VGA block. :(

 
Top