Hey,
Octane will support using multiple GPU's soon (in a few weeks),
without using or needing SLI.
So a double card like a GTX295 could be good.
you could start using it currently, use the 2nd card (sub-card of the dual card in one) for rendering,
and the first card for display, then you get a responsive display while rendering with octane...
then later on, you can use both when octane supports it.
Radiance
Nvidia FX 9800
Forum rules
Please add your OS and Hardware Configuration in your signature, it makes it easier for us to help you analyze problems. Example: Win 7 64 | Geforce GTX680 | i7 3770 | 16GB
Please add your OS and Hardware Configuration in your signature, it makes it easier for us to help you analyze problems. Example: Win 7 64 | Geforce GTX680 | i7 3770 | 16GB
THATS SO GREAT!
i have a question, how is it going to handle the Vram ? i mean is it goingo to create the scene twice (in each card´s memory) or is it gong to be distributed in both?
if not , could be this possible? (hope so
)
thanks
i have a question, how is it going to handle the Vram ? i mean is it goingo to create the scene twice (in each card´s memory) or is it gong to be distributed in both?
if not , could be this possible? (hope so

thanks
Portfolio: http://www.tres31.com/site/
http://www.flickr.com/photos/tres31/
sr2 mobo / dual Xeon E5620 3.2Ghz / geforce gtx 480 + geforce 580 3gb msi twin frozr III / 20gb RAM
http://www.flickr.com/photos/tres31/
sr2 mobo / dual Xeon E5620 3.2Ghz / geforce gtx 480 + geforce 580 3gb msi twin frozr III / 20gb RAM
I hope so too, but as far as I know, each GPU must have same scene and data.juan wrote:THATS SO GREAT!
i have a question, how is it going to handle the Vram ? i mean is it goingo to create the scene twice (in each card´s memory) or is it gong to be distributed in both?
if not , could be this possible? (hope so)
thanks
If not, data traffics between GPUs will make performance seriously down,
since band width of PCIe(2GB/sec(r/w)) is much lower than that of VRAM(90GB/sec).
Code: Select all
c:\> bandwidthTest.exe
Running on......
device 0:GeForce GTX 295
Quick Mode
Host to Device Bandwidth for Pageable memory
.
Transfer Size (Bytes) Bandwidth(MB/s)
33554432 4908.7
Quick Mode
Device to Host Bandwidth for Pageable memory
.
Transfer Size (Bytes) Bandwidth(MB/s)
33554432 4860.2
Quick Mode
Device to Device Bandwidth
.
Transfer Size (Bytes) Bandwidth(MB/s)
33554432 93794.1
&&&& Test PASSED
Press ENTER to exit...
Last edited by YoonKyung on Mon Mar 15, 2010 7:49 pm, edited 1 time in total.
Win7 64 | 2x XEON | 12GB | 8x GTX580
Win7 64 | 2x XEON | 12GB | 8x GTX580
Win7 64 | 2x XEON | 12GB | 8x GTX580
Win7 64 | 2x XEON | 12GB | 8x GTX580
Win7 64 | 2x XEON | 12GB | 8x GTX580
Win7 64 | 2x XEON | 12GB | 8x GTX580
Win7 64 | 2x XEON | 12GB | 8x GTX580
- MadMinstrel
- Posts: 63
- Joined: Wed Feb 10, 2010 8:20 am
As I understand it, both the cards need only the scene data. It would be nice to split the render target between them, which can get pretty large on a 12MP image.
Win 7 x64 | Geforce GTX295 | Quad 2.4GHz | 4GB
Yes, each GPU don't have to keep entire render target, I agree with you.MadMinstrel wrote:As I understand it, both the cards need only the scene data. It would be nice to split the render target between them, which can get pretty large on a 12MP image.
Win7 64 | 2x XEON | 12GB | 8x GTX580
Win7 64 | 2x XEON | 12GB | 8x GTX580
Win7 64 | 2x XEON | 12GB | 8x GTX580
Win7 64 | 2x XEON | 12GB | 8x GTX580
Win7 64 | 2x XEON | 12GB | 8x GTX580
Win7 64 | 2x XEON | 12GB | 8x GTX580
Win7 64 | 2x XEON | 12GB | 8x GTX580
no, cuda does not use SLI.reberts2 wrote:So SLI will not be needed, utilized or valuable in an existing or future Octane setup?
Thanks.
SLI is only for directX and OpenGL games/apps.
Cuda uses the GPUs as separate processors much like multiple CPU cores.
So it can function with a mix of different GPUs and no SLI is needed.
Radiance
Win 7 x64 & ubuntu | 2x GTX480 | Quad 2.66GHz | 8GB