Waiting for 980?

Generic forum to discuss Octane Render, post ideas and suggest improvements.
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
User avatar
smicha
Licensed Customer
Posts: 3151
Joined: Wed Sep 21, 2011 4:13 pm
Location: Warsaw, Poland

http://videocardz.com/52321/nvidia-gefo ... 0-pictured

I hope I am mistaken but I feel 980 will be disappointment for Octane users (compared to 780, 780Ti and TITAN).
3090, Titan, Quadro, Xeon Scalable Supermicro, 768GB RAM; Sketchup Pro, Classical Architecture.
Custom alloy powder coated laser cut cases, Autodesk metal-sheet 3D modelling.
build-log http://render.otoy.com/forum/viewtopic.php?f=9&t=42540
User avatar
Seekerfinder
Licensed Customer
Posts: 1600
Joined: Tue Jan 04, 2011 11:34 am

Yup. Looks to be an extremely disappointing for Cuda rendering. Power efficiency seems great and hopefully the new chip will bring improvements. But only 2048 cuda cores? Agh! Why would they go backwards? It should be 3048. Calling all conspiracy theorists, here are some options:

1. The GTX line started treading on the pro line's toes. There were some internal disputes and one or two people got fired;
2. AMD are not playing the cuda game, so why bother?;
3. Like Intel, they decided to alternate release cycles between an 'evolutionary' and a 'revolutionary' release. This clearly being the former;
4. They're saving all the goodness for the 1000 (or whatever it'll be called) release, next year.

Benefits seem to be:
a. Lower power usage (less heat);
b. It 'should' push the prices of the 7xx series down (tricky one because some of the 7xx cards will certainly be more powerful. So who knows?)
c. POSSIBLY... performance increase per cuda core with the GM204 chip.

To me this seems like a gamer's release...

Seeker
Win 8(64) | P9X79-E WS | i7-3930K | 32GB | GTX Titan & GTX 780Ti | SketchUP | Revit | Beta tester for Revit & Sketchup plugins for Octane
User avatar
smicha
Licensed Customer
Posts: 3151
Joined: Wed Sep 21, 2011 4:13 pm
Location: Warsaw, Poland

There are also rumors about twice as powerful titan 2. We'll see.
3090, Titan, Quadro, Xeon Scalable Supermicro, 768GB RAM; Sketchup Pro, Classical Architecture.
Custom alloy powder coated laser cut cases, Autodesk metal-sheet 3D modelling.
build-log http://render.otoy.com/forum/viewtopic.php?f=9&t=42540
enigmasi
Licensed Customer
Posts: 146
Joined: Tue Apr 12, 2011 7:50 pm
Location: Warsaw
Contact:

c'mon! how you will be disappointed? it's Maxwell, so don't compare with Kepler. just like you can't compare Kepler with Fermi.
min 4gb vram. 180W!
look at "poor" gtx750Ti
User avatar
smicha
Licensed Customer
Posts: 3151
Joined: Wed Sep 21, 2011 4:13 pm
Location: Warsaw, Poland

980 will be as fast as 780Ti - this is a joke. I don't care about 170W, there could be 300W with twice performance as 780.
3090, Titan, Quadro, Xeon Scalable Supermicro, 768GB RAM; Sketchup Pro, Classical Architecture.
Custom alloy powder coated laser cut cases, Autodesk metal-sheet 3D modelling.
build-log http://render.otoy.com/forum/viewtopic.php?f=9&t=42540
enigmasi
Licensed Customer
Posts: 146
Joined: Tue Apr 12, 2011 7:50 pm
Location: Warsaw
Contact:

you would care if you want to use 8 gpus, or 4 in one box, about temp/money
User avatar
Seekerfinder
Licensed Customer
Posts: 1600
Joined: Tue Jan 04, 2011 11:34 am

enigmasi wrote:c'mon! how you will be disappointed? it's Maxwell, so don't compare with Kepler. just like you can't compare Kepler with Fermi.
Need we be reminded that Kepler was a significant 'downgrade', partially to achieve power efficiency? I don't think that will be the case again - that is why I said "possibly" a performance increase as a potential positive. But it's just a hunch.
enigmasi wrote:you would care if you want to use 8 gpus, or 4 in one box, about temp/money
Of course. But what is the point of installing 8 low powered cards if 4 current gen can do the job?

The big question is how effective Maxwell will be with cuda.

Seeker
Win 8(64) | P9X79-E WS | i7-3930K | 32GB | GTX Titan & GTX 780Ti | SketchUP | Revit | Beta tester for Revit & Sketchup plugins for Octane
User avatar
glimpse
Licensed Customer
Posts: 3740
Joined: Wed Jan 26, 2011 2:17 pm
Contact:

personally I don't expect performance bump more than..20-30%, certainly not looking for 2x..-maybe efficiency, but not performance..as we've seen from 750ti (70W) that performs on par with 460 (nearly 200w?).. =) don't think nVidia is going to shot itself in a leg as TitanZ is still sold for boutique builders for 2k$.. even if they manage to show faster card tun this beast - I just guess it will be out of reach for 99%...with Kepler they've managed to get the almost the same performance on gk104 (gtx680) that they had on gf110 (gtx 580), & then with Titan get twice the performance compared to 580 with the same amount of energy - that's a leap..& I think they made it quite nice.

Just because we don't like the price, doesn't mean the technology is bad. If You don't want to buy it, don't do it =) You're not forced & as upgrading CPU with every release, the same should be said with GPU - if You always want to have the "latest & greatest" then there's a price to paid..You choose =) if You need more vram, grab 780 (6gb) while You still can & enjoy for a while..(some models OC like crazy..& still runs stable on air..)& for those who don't need more vram, well 590 seems still a good deal =)
Last edited by glimpse on Tue Sep 16, 2014 5:36 pm, edited 1 time in total.
voon
Licensed Customer
Posts: 527
Joined: Tue Dec 17, 2013 6:37 pm

Maybe Nvidia wants to slowly let die CUDA on GTX cards and force us to buy expensive pro cards?
User avatar
smicha
Licensed Customer
Posts: 3151
Joined: Wed Sep 21, 2011 4:13 pm
Location: Warsaw, Poland

I was about to get 3x780 6gb. I'll wait patiently and see how 980 performs on Octane, and if they go with 8gb version I'll get them.

Some more interesting stuff
http://compubench.com/compare.jsp?bench ... +Ti&cols=2
3090, Titan, Quadro, Xeon Scalable Supermicro, 768GB RAM; Sketchup Pro, Classical Architecture.
Custom alloy powder coated laser cut cases, Autodesk metal-sheet 3D modelling.
build-log http://render.otoy.com/forum/viewtopic.php?f=9&t=42540
Post Reply

Return to “General Discussion”