WATER COOLED 4 x GTX 690 - OCTANE benchmark test

[whersmy]

Incredible, can a 2 PSU system be stable?

[karlo]

You do know the fact that 690 is almost the same in speed as Titan? =) still I don't get why to put the same money in something that has 'limited' usability - lifespan of Your system is basically shorter /what if..You step on higher resolutions, bigger textures, more complex models.. - 3x vRAM, when You don't have to pay any extra..seems a big miss to me =)

but, It's Your decision here =) so it doesn't matter as long as You happy with it I'm just trying to understand =) maybe I have missed smth here, that's all.

cheers

I manage on my desk lots of reviews, opinions and internal discussion with my collegs. So, the final agreement was 690 Anyway, if Nvidia in future release some Titan beast in next few years, of course that we will do the change... World is moving so fast

[karlo]

Incredible, can a 2 PSU system be stable?

Very, very stable....

[mbetke]

Nice rig but not very "future save".
Titan has less noise, more textures, less heat, better overclockability, more ram, could also be water cooled.

I don't see any point why to pick up 690 cards. Maybe price is lower if you get them used.
New Titan card dropeed below 900€ here (GER) this week.

[karlo]

one of the reasons...

we need speed instand of vram... every sec. counts dude, time is money!

[smicha]

Overclocked TITAN is at least as fast or faster than 690, draw less power, has more vram, and so on.

BTW - those bars are false. If 1xTITAN=100%, then 2xTITANs=200% for Octane performance. Their performance should scale linearly, 4xTITANs=400%.

PS. GTX 780 is 10% slower than TITAN (3GB vram).

[glimpse]

Overclocked TITAN is at least as fast or faster than 690, draw less power, has more vram, and so on.

BTW - those bars are false. If 1xTITAN=100%, then 2xTITANs=200% for Octane performance. Their performance should scale linearly, 4xTITANs=400%.

PS. GTX 780 is 10% slower than TITAN (3GB vram).

Yeah, agree & best part is that You get all that for the same price /almost.

well, he has the information, but..I assume, he looked into gaming type review - that's all =)
.."relative performance" ? sorry, wtf is that? =DDD..based on what workloads? etc..

overall nice build and it will do good, helping to bring a lot of nice stuff =) but..

[mbetke]

This graph is totally bogus.
Octane scales 100% with every new card added.
Also if you need speed then Titan will overclock much better and let you squeeze out every single bit because of aboove mentioned features.

Better ask us guys here before you buy next rig.

[karlo]

Hello ppl,

i had update rig with new motherboard P9X79-E WS which supports 16x PCIe 3.0 speed on all four graphic cards and change to 2011 socket i7 4930K Extreme processor which architecture also support PCIe 3.0...

I had done octane render benchmark test, well... I'm pretty surprised with results on kernel direct lighting - almost 200 ms/sec with same GTX690 as used with asus extreme v mono and i7 3770... Probably was the problem in bottleneck of mobo and cpu.

The results of kernels PMC and Path Tracing are higher but small improvement if its compare to direct lighting kernel...dunno why ?!

Here you can see video and at the end is benchmark


Cheers,
Karlo

[FrankPooleFloating]

Hi Karlo - Long time no see.. where you been hiding?..

Again, extraordinarily high production values for a video of this nature. Your real work must be astonishing...

I really think that the benchmark nums are meaningless, if you do not keep your camera parked at the default location. This would probably be boring - in a high-energy time lapse video like this... but M/s is based on what is happening in the scene at that moment - and where the camera is pointing can make huge difference... er, right guys?

You really put a shit-load of thermal goop on you CPU! Yikes!.. Way too much imo. When they say pea-sized, they mean it. I hope your waterblocks on GPUs are not drenched.

See you at the disco bro!