| Author |
Message |
|
|
|
http://www.dvhardware.net/article57525.html
____________
|
|
|
|
|
|
http://www.pcauthority.com.au/News/329929,nvidia-allegedly-set-to-launch-8216geforce-titan8217-gk-110-based-supercard.aspx
____________
|
|
|
|
|
|
http://www.pcauthority.com.au/News/327809,gtx-700-vs-hd-8000-8211-the-future-of-graphics-cards-in-2013.aspx
____________
|
|
|
|
|
|
http://www.dvhardware.net/article57601.html
____________
|
|
|
|
|
|
http://www.brightsideofnews.com/news/2013/2/13/nvidia-geforce-titan-launches-february-18th2c-2013-loses-to-gtx-6902c-amd-hd-7990.aspx
____________
|
|
|
|
|
|
Here's the preview from Guru3d -- the performance review is scheduled on Thursday.
http://www.guru3d.com/articles_pages/geforce_gtx_titan_preview_reference,1.html
____________
I don't buy computers, I build them!! |
|
|
|
|
|
oh i'm very curious how many unit can crunch simultaneously
____________
|
|
|
|
|
oh i'm very curious how many unit can crunch simultaneously
If anyone has a GTX690 that is doing multi tasks, sure would like to know how many. You can probably double the count.
____________
I don't buy computers, I build them!! |
|
|
|
|
|
The number of units is irrelevant. The points is to not let empty cycles - thats why some ppl run more than one at a time. It can be done in most cases with just two units and anything more just increases the overhead of running several contexts and things get worse.
____________
|
|
|
|
|
|
http://www.anandtech.com/show/6763/highend-meets-small-form-factor-geforce-titan-in-falcon-northwests-tiki
Yup, I want that Falcon Northwest Tiki! :D
____________
Windows: "Where do you want to go today?"
Linux: "Where do you want to go tomorrow?"
FreeBSD: "Are you guys coming, or what?" |
|
|
|
|
|
http://www.dvhardware.net/article57639.html
____________
|
|
|
|
|
|
http://www.dvhardware.net/article57673.html
some initial comments around performance
____________
|
|
|
|
|
|
I ignored this thread at first, because of the obviously incorrect title, and also because these articles are all rather light on information.
But for those who haven't already seen them, there are some nice detailed reviews at Anandtech and Tom's Hardware, both of which are split into two parts because of Nvidia's embargo on the benchmark testing. I'm sure there are many other sites besides who have also run thorough reviews.
Anandtech Part 1
Anandtech Part 2
Tom's Hardware Part 1
Tom's Hardware Part 2
In summary, what I see is that the Titan sits in between the GTX 680 and GTX 690 in terms of game performance. Its reference card's physical dimensions also sits in between the GTX 680 and GTX 690, with a blower-style fan that lends itself to small form factor designs. The power consumption, too, is in between the GTX 680 and GTX 690.
GPGPU comparisons are harder, since different applications are favoured by different architectures - in some, AMD's Radeon HD 7970 outperforms the Titan (and the GTX 690), in others, the GeForces perform better. But universally, Titan appears to outperform the GTX 690 because of the obvious advantages the underlying GK110 GPU has over the game-oriented GK104 used in the GTX 670/680/690 cards.
For our purposes, however, in terms of SETI@home performance, CUDA performance on MB seems to be far better than ATI OpenCL. I don't have any experience on the HD 7000 series, but AstroPulse seems to be served fairly well by both Radeons and GeForces in that generation. Overall, I'm inclined to believe that the Titan will be a very nice GPU for SETI@home work above the GTX 690... assuming you have the money for it and are able to squeeze enough WUs from Berkeley's constrained bandwidth.
____________
Soli Deo Gloria |
|
|
|
|
|
if someone buy one please share with us the test results, i'm considering of buying one or GTX 690. Thanks
____________
|
|
|
|
|
|
They will be avaiable only the next month for us normal "mortals" at the same price??? of a 690.
____________
|
|
|
|
|
|
Yes, the GTX 690 is available for €950 - €1.000 here.
The GTX Titan for €950.
At one german shop I saw the GIGABYTE- & ZOTAC- GTX Titan is available already (1 - 3 days shipping).
BTW. The x41zc_cuda50 app, it's already optimized for the GTX Titan?
Or the app would need to be overwork for to get the max. performance?
Oh... - I'm very nervous & undecided...
2x GTX Titan for to replace the 4x GTX 260 in my 940 BE cruncher.
But - who give me €1.900? ;-D
* Best regards! :-) * Philip J. Fry (formerly Sutaru Tsureku), team seti.international founder. * Optimize your PC for higher RAC. * SETI@home needs your help. *
____________
>Das Deutsche Cafe. The German Cafe.< |
|
|
|
|
|
>techpowerup.com<
|
Colorful Announces Factory-Overclocked GeForce GTX Titan Graphics Card
Colorful announced a factory-overclocked GeForce GTX Titan graphics card, which sticks to NVIDIA-reference board design. The card ships with a maximum GPU clock speed of 983 MHz (an offset of 100 MHz over NVIDIA reference specifications), with a memory clock speed of 6.40 GHz (roughly 400 MHz overclock). The board and cooling assembly, as mentioned earlier, are NVIDIA reference designs. Colorful's package includes a pair of jacks that work to minimize PCB bending on the graphics card, in addition to the usual paraphernalia. Colorful will begin sales in the Greater China region.
* Best regards! :-) * Philip J. Fry (formerly Sutaru Tsureku), team seti.international founder. * Optimize your PC for higher RAC. * SETI@home needs your help. *
____________
>Das Deutsche Cafe. The German Cafe.< |
|
|
|
|
BTW. The x41zc_cuda50 app, it's already optimized for the GTX Titan?
Or the app would need to be overwork for to get the max. performance?
Look the Jason post on this thread, he explain that.
____________
|
|
|
jason_gee Volunteer developer Volunteer tester
 Send message
Joined: 24 Nov 06 Posts: 4048 Credit: 60,428,014 RAC: 57,700

|
BTW. The x41zc_cuda50 app, it's already optimized for the GTX Titan?
Or the app would need to be overwork for to get the max. performance?
Look the Jason post on this thread, he explain that.
That was in a GPUUG thread, so I'll summarise here:
Looks like a tasty refresh. There's plenty of improvements, and added instructions etc that could provide very great performance improvements. The Cuda 5 app should work and scale on it quite well (or a quick rebuild in case of some unforeseen issue). As with any new major architecture there are new things that could potentially be used to extract more performnance. Those take time even if the initial build works well. I'd consider the apps currently about half optimised for Kepler, so less than that for Kepler 2 ;).
____________
"It is not the strongest of the species that survives, nor the most intelligent that survives. It is the one that is the most adaptable to change."
Charles Darwin
|
|
|
|
|
|
angst
____________
|
|
|