Are those times using the blend I uploaded? If so, thats not hugely surprising, as nVidia have put more effort into CUDA support (unlike AMD’s OpenCL :P) Still I wouldn’t push SLG/Lux as a GPU alternative to Cycles, the broad range of features, and better support CUDA has is unbeatable if thats what your looking for (I’m not)
Thats just sad!
I’ve been thinking about getting a 760 4GB to get some more Vram for the large scenes. But with the numbers you have put up here, it just shows that the 760 will be a huge step down in speed, compared to my GTX480 and that wasnt really what I was hoping for 
I really really wish that my 480 had more Vram…
Hi Wega, render my benchmark file (first post) and compare with the sheet, than you know. 
I am also interested how the 480 perform.
Cheers, mib.
I already know it, because my GTX480 is faster than my GTX570 thats already in the sheet. But i will run your test file when it’s done with the current render.
Blender 2.68a
Windows 7
GTX 480
01:53.16
Blender 2.68a
Windows 7
GTX 480 + GTX 570
01:02.95
Blender 59377
OS = Win 7 x64
GPU = GTX650TI
Render Time = 3:27 (stock clocks)
Render Time = 3:03 (over clocked)
CPU i7-870 - Render Time = 10:21
Blender 2.68a
Windows 7/64
GTX 550Ti
04:53:23
Thank you Wega, aljo and babareilly, new spreadsheet comes next week.
Still waiting for GTX 770 and GTX 780 user to jump in.
Cheers, mib.
Blender 2.68a
Windows 7 / 32bit
Nvidia GTX 560 1 Gb GDDR5
02:52:31 (1000 samples)
01:26:02 (500 samples)
Blender 2.68a
Windows 7 / 64bit
GTX 660ti x2
01:27.75 (1000 samples)
Blender 2.68a
Windows 7 / 64bit
GTX 570
01:54.35 (1000 samples)
Titan soon !!
Blender 2.68a
Windows 7 /64bit
GTX 680 2GB (EVGA brand)
02:07.04 (1000 samples)
Re-tested (after closing photoshop, and browsers):
02:05.78
Question… how are you guys running 2 different video cards. Like I’m seeing a guy with a Titan AND a 560? I thought to SLI, you had to have 2 of the same cards.
I have a 550ti currrently… could I add a Titan or would I need to take the 550 out? Feels like such a waste of a good card but I guess I could always sell it second-hand.
Also, does anyone have a 780 to try out yet?
Hey bud, you can connect two different nvidia cards… but the problem is that it can actually cause a loss in performance in better card, primarily if the memory is different.
This from Nvidia’s FAQ:
Can I mix and match graphics cards with different sizes of memory? When purchasing a second graphics card, you should try to match the memory size so that you are ensured full value and performance from your purchase. However, while it is not recommended, NVIDIA does offer the flexibility to run graphics cards with different sized memory by using CoolBits. Using CoolBits (value set to 18), you can force both of the cards to use the lower of the two memory sizes and operate them together in SLI mode. When dissimilar memory sizes are enabled to work together using CoolBits, the effective memory size for each card becomes the smaller of the memory sizes.
@Tommywright
It is all up to how many pci-e slots your motherboard has. You can have multiple video cards in your computer of different types and cycles will use one or all of them depending on your settings.
When rendering with a single card the UI can become very slow so you can use one card for your display and the second one for rendering and then use both cards to render the final image.
For rendering with cycles it is actually better not to have SLI enabled.
@tommywright
I have two NVidia cards that both happen to be 660ti 2GB, but I don’t run them in SLI anyway. The NVidia driver “Control Panel” comes with a way to disable SLI. Blender recognizes them each individually and runs them independently, and I’ve heard it’s faster with cycles this way anyway. I haven’t noticed any real difference in game performance, but then again, Call of Duty Modern Warfare 3 isn’t exactly stressing one card, let alone two.
Oh wow! Thanks for the info! I had no idea Cycles could use multiple cards like that.
One last question. I’m noticing that all the new graphics cards are PCIE 3.0 and I have a PCIE 2.0 slot. I’ve read that it is backward compatible but what kind of performance loss do you think I will get with something like a Titan?
You should check out the Cycles guide http://www.systemagnostic.com/faqs/ as you will find all this info there.
Regarding the PCIE speed it says the following in the guide:
Short answer: No, this matters only for gaming, you can ignore it for rendering.
Long answer: The advantage of a faster bus is that you can shuffle textures (and other stuff) in and out of the GPU much faster, but when you’re rendering, the textures are all copied to the card before you start, and are not changed until you render the next frame (if you are rendering an animation). This means that the bus speed is mostly irrelevant for rendering. It only impacts the time Blender spends setting up the work at the GPU. If you render a lot of frames with short render times (less than a minute for example, with 1000+ frames) it will matter, but for most people it won’t matter at all. If you have a motherboard where you can set the PICe speeds like on mine, you can try for yourself.
I was worried about this, as my motherboard reduced the PCIe speed of all cards when adding two (down from 16x to 8x on both cards) and even worse for three cards. But I did some benchmarking and it turns out there was no performance drop at all for Cycles. For gaming it’s a different story though, but I’m sure there are plenty of other FAQs on that.
The same applies to PCIe16 ports that are “electrically” only PCIe8 ports. This you can, for rendering purposes, completely ignore.
Minor correction for the spreadsheet, Mib - I’m using Linux Mint 13, and not 14