Daz 3D is part of 
Connect
DAZ Productions, Inc.
7533 S Center View Ct #4664
West Jordan, UT 84084
Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2025 Daz Productions Inc. All Rights Reserved.
Comments
On one hand, I totally agree with this. On the other hand, if it isn't hard or time-consuming for PAs to do it, then it isn't hard or time-consuming for the end user to do it either.
This is a big oadvantage of Octane in my opinion.
Wow really ? I just got my Titan X SC Hybrid running at 1400Mhz for $1099 usd direct from EVGA , it is almost 40% faster than Titan X standard , but it is power hungry beast
just looked at their site most titan variants around same price but standard version on sale for $1395, think read you talked about m6000 cards price for m6000 is $7059 the 1080/70s are listed with no price ones that did list got into a little trouble think cause weren't supposed to say yet Nvidia Australia did get asked by news sites here on pricing but they said couldn't say didn't know what the price be and that it was really up the the disrubitors and retailers what they wanted to charge and so far few are saying $1299 been seeing a few new hands on reviews but the thing is it's starting to look scripted way toooo much they're all virtually saying almost exactly the same things sometimes almost word for word
You see if you buy the cards in USA you will pay import tax so not much less in total . I think the reason why it is so expensive over seas . The best solution to have good friend that ship for you as gift lol
Yes, we who are under a 'free' trade regime, pay import tax.
infos leaked: The news around nvidia pascal talk about GP102 based graphic cards, supposedly named 1080Ti and Titan. More than 3500CUDA (3840?) and ton of ram (16+) but not sure about GDR5+ or HBM2 (probably the first one)
The cards will be better than 2 1080 in SLI...
So, Im looking forward to change my MSI 980Ti with those....
Mec4D:
What Housing and What Motherboard are you currently using?
sad... but it will be $50 import tax per 1 Titan X + shipping cost lol
btw my new business card
.. and no , not 3 times faster AT ALL ! at last not with iray it is almost 10000 cudas at 1375 Mhz , as predicted
I saw it , you will be better with 1 or max 2 for the money and good enough!
I am using LGA 2011 v3 i75820K 6 cores 12 threads , MB from November last year : GA-X99P-SLI , it is made for graphic work and games and to use with more than one card , no internal video , Thunderbolt v3 so I can plug in to 6 external devices and use all the GPU power with just one cable , USB 3.1 and other new stuff . It run 1 card at 16X , the other 2 at 8x but it make no big difference here 16x or 8x so having better processor is just waste of money , modern motherboards working different than before , so your CPU can have only 26 lane and still the card will run at 32 total but what I had to do is turn off the Sata ports under Bios that I don't use and enough for all PCI-e , also installed the USB CPU water cooler outside the case and CPU run at 22C on idle inside 35C
Zbrush and other apps are so fast with this new rig.. love it
Thanks! And What housing do you use to support the 4 Radiators (CPU & 3x GPU?) Thats my main problem with switching to Water Cooling....
I use closed system for each one Corair H80i v2 for CPU and EVGA Hybrit Kit for each Titan X all 120 mm double fans , the CPU radiator is double that why installed it outside as my case have opening for that and for custom water cooling but I am not as confident enough to do it so I use closed systems , easy to replace if needed
any word on when these cards will support CUDA? At last glance Nvidia forums point to a "not yet"
https://forum.nvidia-arc.com/showthread.php?14632-Will-the-Geforce-GTX-1080-be-supported
That is a good warning. I wasn't planning on upgrading anytime soon anyway. I don't have a need to, but I keep my eyes out sometimes just incase something blows up so I know what is out there.
Marble posted that earlier, and it's where I came across it, but something to keep an eye on.
they gey you another way then here via postage and anything else excuses here getting cunning sneaky plus postage and handling is horrible they don't take good care of things delivery wise and now they've implimented their what I call ransom on your mail where they can charge you extra fees if you don't collect your mail by a certain time see they've cut down how much mail is delivered and days it's delivered and if it's a parcel they may or may not leave it sometimes jammed into letterbox or left on side road other times if your not home they'll leave a note to pick it up but sometimes they can't be bothered to see if your home so leave a note anyways but because they have cut down delivery times and days they could have hold of something for few days before letting you know or delivering and if it's over a certain number of days they can then charge a holding fee hasn't happened to me yet but I've had things delivered very late including bills
so that's one of the reasons been eyeing of one custom build pc place that has their products delivered by courior and not only packed in lots of styro but also big wooden crates speaking of which the gtx 1080 is out and one build with either two titans or two 1080s the price difference is $1037 so that is a bit of a saving if the 1080s are actually worth it for high quality rendering or least really good rendering of large scenes and detail otherwise they're still expensive as in wasting money instead of spending bit extra that's worth the money like the titans though that saving maybe enough to buy zbrush hopefully hopefully the reviews will come in fast now
The release 1080's seem like they will only enable an 8gb scene. The idea of 16gb+ on a 1080ti or titan makes me think I might be prepared to wait quite a while for this. My new rig should be able to render 3delight like this one is standing still and the idea of being able to plug in up to two 16gb 1080's at some point... . Maybe by then it'll be clear as to what will or won't be supported, and when. Can't wait to get stuck into Iray.
This isn't a 'leak'....the only site listing info makes it quite clear it is only CONJECTURE (based on a lot of previous behavior from nVidia), but nothing more.
NO fab or indications of a GP102 yet. No sightings on shipping manifests.
Nothing.
Furthermore, I doubt seriously a 1080Ti card would have that many cores. It would put it on the same level as the P100 (which has 4096) and at a considerably lower (read consumer-level) price. If I had to make a bet (and note, I'm not....at least not until we have something more concrete) I'd estimate that IF they make a 'GP102', it will clock similarly to the GP104, but probably with a higher base clock. I'd estimate it would have either 2816 or 3072 total CUDA cores. No idea if they'd keep the same ratio of FP32 to FP64 cores as in the 1080, though. And, I'd expect it to be using GDDR5X just like the 1080. And I'd expect 8GB to 12GB of that GDDR5X. And I'd expect the price to be in the $800 USD range.
As for the Iray support, I guarantee they'll get it working pretty quickly. With the 1080 GTX performance specs, they don't want to limit their market any more than they have to. OpenCL support is another matter, though......
At least holding out means I'll be most likely to end up with best value for money, and something that will do what I need it for.
Need....more....info.....
I don't know...980 ti has 2816 CUDA cores and the M6000 has only 3072. That's closer than the conjectured 3560 (3840?)is to the P100's 4096. I think that's a reasonable estimate of CUDA cores. The thing normally gimped on the GTX cards is the double precision floating point performance. I predict that the Titan equivalent of P102 will go back to having this enabled, as did the original Titan and Titan Black, but the 1080ti will not, which will finally give both cards a reason to exist. Maybe the titan-class will use HBM2 but the ti class with GDDR5X?
...not holding my breath on this. I remember a couple years ago when information was "leaked" on the upgraded Maxwell 970 and 980s which said they would both have 8 GB of memory. Didn't turn out that way in the end. At only 350$ more than the 980 TI, the Maxwell Titan X was a much better deal if you needed video memory and the most CUDA cores.
I remember reports last year were touting that the new Pascal HBM cards would have between 5,000 - 6,000 CUDA cores. 3,840 seems a bit more realistic.
...that's my idea. Again after the letdown I had when the Maxwell 970 and 980 TI were released, I'm content to wait and see what comes out in the wash. Who knows, maybe I can score a couple Maxwell Titan-Xs for a decent price in the interim.
read in one interview that they considered the current versions of 1070/80s weren't really being aimed at those that recently bought/upgraded to 980ti or titans but more aimed at those with lower end cards as being a cheaper bettter upgrade for them which thought was interesting but there are a few reports next versions of 1080s with more cooling options
...for myself, I'd like to see benchmarks for CG production/rendering and not just games. As I don't do games, the results of the reviews and tests i see are pretty meaningless.
well, several things are messed up out there...
What we know for sure:
P100 -> 3840 core with all FP32&FP64 die 600mm2+
P104 is half this size
P102 has been referenced (in drivers if i remember well); putting (partially) away the FP64 of P100 will considerably reduce size of the new die, making it around 480mm2 with all/almost all core enabled: ence the Ti/Titan at a consumer market compatible price/size/cooling
P1xx are compatible with GDR5 & GDR5+ & HMB2, but manufacturers of HMB2 are net yet ready (Q3/Q4 even later)
Having a 1080Ti/Titan with 3840 (3500+) CUDA cores gives a similar ratio to the one of 980/980Ti . Knowing that 1080GTX has already 8GB GDR5+, and older titan/980ti with 12/6GB, they can't make 1080Ti less than that! (12/16 is a good guess). More of this, I've read Q3/Q4 release. The P100 is done, the P104 is done, so releasing P102 (and upcoming P106) is a ligher process now.
So, for me, this sounds like more than reasonable speculation.
The reason I did not got it was that it would affect the GPU scaling in combination with my cards , but the deal will be ok for people that run on CPU only and want upgrade for faster rendering , but if you want to use any other card with? forget !!! even using more than 2 x 1080 will affect already the GPU scaling and slow down the cards and rendering so for me not a deal.
GTX 1080 is not supported yet for rendering with iray so hold your horses until the compatible version is released and DAZ Studio updated .
Reading what little info is out there, I am not sure what needs to be updated to get 1080 running on iRay. It is either the 1080 driver, the iRay software itself, or both. If it is the latter (or both), then as MEC4D says, we could be waiting some time to use the 1080 in DS, if DAZ first needs to wait for another iRay update, and then we have to wait for another DS update. If the issue is solely with the 1080 driver, then the wait could be shorter.
.
It isn't the only thing to consider; they may not be as good for rendering as the specs suggest as NVidia considers them a consumer card (games and general stuff); rendering is more their commercial variants.
What they will be able to do, will be down to drivers. So am not in any hurry to part with my cash.
@ sources of Pascal information
Maybe it helps someone who can speak german or uses the google auto translator:
PC Games Germany has a page that lists all the latest news about the Pascal Generation.
http://www.pcgameshardware.de/Nvidia-Pascal-Hardware-261713/
Some of the german news also link to the english speaking source.
- - -
@ GP 102
This one
http://www.pcgameshardware.de/Nvidia-Pascal-Hardware-261713/News/AIDA64-Entwickler-bestaetigen-angeblich-GP102-1196519/
links to that one
http://videocardz.com/60289/breaking-news-aida64-developers-confirm-pascal-gp102-gp106-gp107-and-gp108
- - -
@ how many cards to buy for each generation?
In any case I will stop buying three or four of the same GPU but instead place the maximum of the same GPU at two.
This means I plan to use the following setup:
The 2 new GPU are used for rendering.
1 old GPU is used for the display.
(optional) 1 additional old GPU is toggled on for rendering IF the scene also fits in the VRAM of the older GPU
Maybe I also give away the 4th card and keep it at 2 new 1 old. depends on available power and space of the mainboard
-> The reasoning behind this is that I expect a lot more GPU improvements in 2016, 2017, 2018 then there were in the last three years. I am not going to upgrade 3 to 4 cards each year but rather just get 2 of each new generation that offers larger VRAM.
- - -
2nd gen I'll consider, 1st gen GPU's are for folks with disposable income who will beta test while I wait.