• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

GTX 280 TRI-SLI / Core i7 965 Watercooled!

Page 3 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Pfffft. I just overclocked the E2160 in my garage computer from 1.8GHz to 2.4GHz.


Your computer costs more than my car. And since it's a Hyundai your computer is probably faster too.
 
Originally posted by: Zenoth
I'd be curious to know what is the power consumption of the whole system.

Well..I have a Cooler Master UCP 1100 which is rated to handle up to 1200W

Based on some calculatons I have read.....with a full load on the system...

3 GTX 280's overclocked = Approx 660W

Core i7 overclocked = Approx 270W

Hard Drives/Optical Drives
Water Pumps / 14 rad fans
Misc Fans = Approx 120W

Total = Approx 1050W

-M
 
Originally posted by: balane
Pfffft. I just overclocked the E2160 in my garage computer from 1.8GHz to 2.4GHz.


Your computer costs more than my car. And since it's a Hyundai your computer is probably faster too.


LOL!!

-M
 
I just noticed you have a top plate on the bottom 280 but are using Zalman ram coolers on the other 280's. By design?
 
SteelSix....Good eye!

After installing all three cards with a backplate...I found that two of the cards had much higher temps.... after multiple attempts at reapplying TIM and testing I narrowed it down to the backplates were not allowing enough contact from the block to the GPU...so I attached the block directly to the cards and used the Ramsnks for the two cards....temps came way down...

-M
 
Originally posted by: managerman1985
SteelSix....Good eye!

After installing all three cards with a backplate...I found that two of the cards had much higher temps.... after multiple attempts at reapplying TIM and testing I narrowed it down to the backplates were not allowing enough contact from the block to the GPU...so I attached the block directly to the cards and used the Ramsnks for the two cards....temps came way down...

-M

That's odd. With the backplate screwing into the block, seems like it would only help with gpu contact. It's common for the primary card in an SLI setup to be warmer, in a Tri setup the first two, due to scaling I suppose.

I wonder if by "disconnecting" the components from each other, temps came down. With all connected, they're in effect sharing heat output. Think removing the back memory from the link helped lower gpu temps?

With the bottom card typically being the coolest running, perhaps the backplate on didn't make a difference? Not taking the water cooling link into consideration, no experience with that..
 
Originally posted by: managerman1985
I have had a chance to tweak the system some more. I am now running at 4.3 Ghz @ DDR3-1724, 172x25 GTX 280 SLI now at 740/1540/2460

I have attched some new benchmark runs.

3D Mark Vantage and 3D Mark06 plus some CPU benches

Crysis Warhead @ 1920x1200 No Anti ailiasing / 16xQ Anti-aliasing

More to Come...

-M

Man those Crysis settings and frame rates are completely :Q.........

Playing Crysis at 60+ on Enthusiast at 1920 must be insane. Curious, have you tested how much difference that 3rd card makes in frame rates? Trying to get an idea of how GT300 will play Crysis. 🙂
 
ya, i was looking at the thermaltake Mozart box the other day and i was thinking there is now way we ever need such a big box for a system. someone can live inside that thing.

and just before you know it......
 
ya, i was looking at the thermaltake Mozart box the other day and i was thinking there is now way we ever need such a big box for a system. someone can live inside that thing.

and just before you know it......
 
FarCry2 85 min, 100 avg 1920x1200 AAx8 maxed LOL

And yeah, 15 min for Crysis at 1920 with no AA sucks a bit...But if it happened like once or twice and wasn't a consistent slowdown, where the game just bogs down during something for several seconds, it's fine. Avg is still 60fps, not to mention Crysis DX10 V High/Enthusiast looks absolutely stunning.

Do you have GTA4? I read this game tanks on every rig today. I wonder how yours would handle it... 🙂

Cheers.

EDIT: And I think it was Anand that showed running Crysis from an SSD drive improved min FPS dramatically? Would you mind installing it on your SSD and trying the test again at 1920x120? Thanks!
 
FC2 DX 10 1920x Ultra settings with 8xAA enhanced to 16x CSAA from the nVidia CP.

Min FPS 66.13 Avg FPS 92.24 Max FPS 134.38
http://img208.imageshack.us/my...mage=fc216xnohtac4.jpg

FC2 DX 10 1920x Ultra settings with 8xAA enhanced to 16xQ AA from the nVidia CP.

Min FPS 46.13 Avg FPS 80.76 Max FPS 120.28
http://img78.imageshack.us/my....mage=fc216qnohtif8.jpg

No AA

FC2 DX10 1920x Ultra Settings Ranch Small

Min FPS 65.87 Avg FPS 126.50 Max FPS 174.21
http://img509.imageshack.us/my.php?image=fc2nohtfw4.jpg

If you want FC2 to run smoother with AA try setting AA in game and then "enhancing" it with nVidia 8x or 16x CSAA. Again, I'm using mixed GTX 260 Tri-SLi @ 711|2300. Also make sure you kill HT for this game. It will rob you of 6-14FPS on average. You take the largest FPS hit when using HT and 16xQ AA.
 
hey m or wateva lol, can u run some benchmarks on the ORIGINAL Crysis instead of crysis warhead please? U have the only pc ive ever seen that has been able to max out crysis warhead with 16Q! I want to see if its the same with the original. sick rig btw.
 
Back
Top