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
© 2024 Daz Productions Inc. All Rights Reserved.
Comments
still it would be nice if the 1080's did end being really good cheap alternatives to titan x ok still maybe not as powerful core wise but still pretty good 2x 1080s instead of 2x titan xs would save me $1000 and one computer I looked at now offers intel core i7 broadwell-e processor with up to 10 cores and 20 threads, thats'spretty good isn't? btw power supply wise what's best for either 2x 1080s or 2 titans? on one computer with max 2 cards the max power supply is 850w and on the 64gb computer you can have up to 4 cards max power supply is 1500 w-can have lots of drives in that one btw also offers 650, 850 and two different 1200 w power
It is , the biggest event of the year! cream de la cream of computer and softw world ... how possible you missed it ? are you newbie? LOL
Oh it will be for sure better as any cards below 980Ti for sure , and if you use 2 of them then you better not mean 2 times better as 2 GPU don;t mean double power in iray .
Also make sure your system is ready for that as that is important , good fast processor will speed up work and if you run it on slow processor then the render will go fast anyway but the viewport will suck big times .even with 3 Titan X .. that why I upgraded my processor to run 3 Titan X SC in viewport other way I could use only 2 , i74790K and above are recomended for more than 2 cards
my current most powerful computer v17 Nitro laptop has a single gtx 850M so 2x 1080s desktop would be a huge boost again if they turn out really good for daz iray and programs. Actually the place I got this from via a 3 year try buy contract told me about new computer they're getting trying to tempt me with it the Predator said it was a 64Gb looked it up sounds like the Predator G6 model looks cool at first read about it reviewers said kind a mess inside with cables and extremely tight virtually impossible to upgrade cause so tight and can only have one card which is a basic gtx 980 still better than what I have yes but still, computers offered here for the most part either store bought or custom build best we can get on offer 730 to 980 some 980ti not many though and computers or custom builders that offer them with titans are like unicorns, those rare ones that do most are very overpriced on their 1070, 1080 and titan builds would like at least a 2 card setup and 2 1080s would save me a fair bit on either the 32gb desk or 64gb desk it'd make the 64 one more in my reach
I'm thinking the 8 core to be considerably better value; unless you really need those two extra cores.
Siggraph North America 2016 is scheduled for July 24. Not too long of a wait, unless you already own a 1080 card.
Siggraph has been a big convention/trade show/learning experience for many years.
SIGGRAPH is the ACM (Association for Computing Machinery, the big professional association for computing software/hardware) special interest group (SIG) on Graphics. There are a lot of these. SIGGRAPH is a pretty big SIG within the ACM.
There are OTHER SIGs within ACM. And most also have a big yearly convention. But SIGGRAPH has gone beyond any of the others. It has been a huge spectacle and massive gathering for a couple of decades. I went to one (back in 1994 or 1996, I believe) when it was in Orlando, Florida. It was mind-blowing. It was expensive (though I got by a lot cheaper since I was a student back then) and a blast to go to. If you had connections with some of the big companies so you could get invited to their big parties at SIGGRAPH, you were in for some major fun. Even just walking through the exhibition hall was freaking amazing. The giveaways and contests were boggling....I think HP was giving away a high-end Alpha workstation EACH day, by raffle. Intel was raffling off CPUs, one smaller company was having a raffle for a 64 cpu i860 renderman rendering appliance (though only one, not every day...)
The workshops, classes, presentations, and animation competition were an absolute blast and very educational.
IF you can afford it, I highly recommend going.....especially if you do any actual development or high-level animation stuff. Or just love a good computer-graphics-geek party that lasts all week long!
(ETA: Just checked, I was there in 1994. Total registered attendees was around 25,000.)
1994, the year I bought my very own first computer (a few grand) and Animation:Master software which cost $695. I heard about the convention back then.
1994 was a good year moved to Holland , got 1 year early win 3.1 .. 22 years run so fast , but my big graphic adventure started in early 1998 I messed up with Maya, 3D Studio Max and other very expensive soft on this time , I have to work 2 weeks to pay for Poser 4 lol
Yeah, I don't think cuda is their first priority. Nvidia is doing VR and games and driver enhancements.
I can beat you both by over a decade. I got my first computer in 1982. It was an IBM System 36 and it didn't cost me anything. On the downside, I had to share it with about thirty other people and it was spectacularly bad at anything remotely resembling graphics.
Cheers,
Alex.
was a long time before I got my first computer was an acer can't remember what type come to think of it think all my computers have been except for my gateway desktop but thats probably an acer off shoot too. Anyway remember back in highschool few had the old spectrums and ataris don't think anyone had an amiga but spectrum more common
iray will not even use 3 cores with 2 gpu, you will need 4 card for 4 cores, go with the 8 cores if u can , I have it and it works better than i75820 plus u get 40 lanes needed for other stuff like M2, USB 3 SSD etc I will recommend not going below 1000W with 2GPU for 3 1300W is optimal , rendering use less power than games anyway .
My first computer was a Commadore 64 with a casette drive for data storage and cartrages for games and it hooked up to my small color tube tv. It had basic programming and it was fun.
I've got a question about using a 1080 and a 960...
I was lucky enough to order a 3rd party gtx 1080. I've heard conflicting claims about how a card with lower memory (960 has 2GB, 1080 has 8GB) can limit the memory DAZ can use. My question is should / can I use both cards to power DAZ? Or would the 960's 2GB actually hurt overall performance compared to just using the 1080? Also I love IRAY and am a bit dissapointed that the 1080 is not yet suppported. I was thinking I might need the 960 for IRAY renders for a month or so. But again I don't know if a dual 960 - 1080 setup would even work with IRAY?
Thanks for reading!
Iray treats all your gpu cards separately, loading the entire scene into each. So if the scene will fit in 2 GB of vram, both cards will be used. If the scene takes more than 2 GB, the 960 will be dropped and the 1080 will be running solo. Your best bet, once Iray supports the 1080, would be to use the 960 to drive your monitor(s) and let the 1080 handle the rendering.
I believe even running the 960 alongside the 1080 will impede performance since the renderer needs to service both the slower and faster card simultaneously.
As for SIGGRAPH, I was going to attend this year out of my own pocket (not work related unfortunately) but being a Canadian meant that the trip quickly turned unaffordable :(
Actually what will happen is the 960 will run at 100% and the 1080 will only maybe 40 % or less slowing down everything , that why I had to adjust my cards to the same clock speed for best perfomance, only if you use GTX card a side with Telsa or Quadro you will profit from thanks to the special iray driver they have .
Are you sure about that, I have a Titan Z that runs at about 900mhz and a 980ti overclocked running at about 1300mhz the cards are at 100% load when rendering in iray, the 980ti is not slowimg down to match the titan Z speeds. I've confirmed these numbers with GPUz. I've also compared my results with octane bench and my cards add up singe scroes versus multiple cards with a slight drop in multiple only due to heat.
When i run tests in iray i always keep the aux window open in iray but size it very small, 50x50 pixels, this keeps the scene in memory, so when you render there is no overhead for loading, same goes when switching the main view from texture shaded to iray. If you render this way adding multiple cards scales very well with almost no loss.
I am sure .. That what you see in your control software has nothing to do with what is happening in iray , the cards may run at full load and clock but the all resource is not used as rendering power with multiple GTX GPUs so only one of your card is actually used for rendering at 100% resource and the other just partially , only memory are used at the same level on all GPUs
And actually not even 100% resource but 80% with basic rendering setting for the first card, 100% only if you use Optimization like Caustic and Architectural on . So 80% for the first card , 45% for the second etc.. etc..
with iray you can''t stack the GTX GPU power or memory .. So how more GTX you add how less it is going to be used . And the best rig for iray rendering you can do for your money is having max 2 fast GTX GPUs in your system and one smaller for monitor and the best performance if they are from the same series . Unless you go pro and switch to Telsa or Quadro then things change as the cards run on own special iray driver but don''t waste your money yet since you will need 16 Quadros to have instant real time render in iray . But if you go just for one cards, GTX 1080 will be faster than Quadro , as your 980ti is faster too , but with combination of other GTX card you losing the resource and energy .
For the same reason you need SLI when when playing games for optimal result , you don't mix Titan Z and 980ti in SLI , you can only use SLI with two, three or four or more Identical GPUs. And if your CPU is not fast enough may result in even lower than expected GPU usage . Just because you don;t render with CPU don't mean you don;t need better and faster one when using multiple GPU's , Yes you do but in most case overclocking slightly the CPU will do the trick , even with my powerful i75960 X and 3 x Titan X SC cards I need to run the CPU at last 1Ghz faster to gain the max performance to feed the 3 cards or everything slowing down even more , so to save the energy and resource I work preparing the scene with just 2 GPU's and CPU at standard clock and use my full system power and 3x GPUs when rendering the final image . Other way I am only killing slowly my GPU and CPU for nothing .
So just because you stack all cards you have under hand for Iray means nothing , the statement that Iray will use all GPUs it finds in your system is only half of the story , and it will find them all but it does not tell you how much resource it will use . If that was the case nobody will spend 4K on workstation cards
and since we can't connect SLI bridge for the super GPU power to use with Iray , it is how it is ..
so if you have 2 cards and one is faster than the other , iray can't render faster on one GPU and slower with the other , for that reason the slower card will slow down the rendering performance
The same situation is for example rendering with Adobe after effects , 1 card will run at 100% and if you add second only 50% faster so 150% on 2x GPUs , The same situation is with Blender Cycles little less good than that
But then you switch to Octane rendering and add second card you gain 100% speed so total 200% with 2x GPUs but the best results are always the best when the GPUs are from the same series
and all the thing about stacking GPUs and gain more rendering power has nothing to do with the cards in general , but the software that run them with own limitations due to the coding or set this way by the owner on purpose. You will say they need new driver ? no because Octane can run the same cards with the same driver at full speed using all the card resource . I would better say you need an unlock key hidden in the special driver but it will cost you 4 x times more
So if someone plan to make new rig, think about everything, not only GPU but Also the system that have to be efficient to allow you run it optimal .
I made my new rig in December last year , well with 2 x Titan X SC and i74790K and 32 GB it was perfect fine set , then in May I added one more GPU and everything started to drag down , I had to upgrade the memory and CPU and at the end also motherboard so plan a head and stop focusing on stacking up the GPU only as it will not make anything faster .
I just ran a few test this morning, and things scale very nicely. I'm sorry I do not see any slow down.
It can vary from scene to scene; I've had it both slower and faster when adding Optix. If it's a complex scene then it is worth texting at a lower resolution before getting ready for a final render. Slightly more often than not, slower though.
You see everyone focus on rendering time and second but forget how much money they invested where the card is used bellow 45% when in combination
and your 980TI got slow down ( it is not about the rendering time in second but the card performance ) you missed the point , by adding 980ti you speed up the rendering only by 13 second what is poor performance for the money you spent on the card, other way you should have at last below 20 sec with Optix ON so you are running on one leg with 980ti and the rest go into the air unused and iray used your fastest card as additional source and not as main source what was one of the first Titan Black card only and not 980ti
you forgot to make the most important test in this conversation what is 1 980ti + 1 Titan Black however due to closer numbers of Cudas it will be not so dramatic, as test with 980ti +760 would show it better
The performance for 980ti with Optix On is very poor compared to Titan Black it should be much faster
with the Optix Off the ratio between the 2 standalone cards is more pleasant for the money you spend on it .
"The performance for 980ti with Optix On is very poor compared to Titan Black it should be much faster"
I ran this test again and it wenrt from 30 seconds to 28 seconds... a little faster.
1 titan black + 980ti 39 seconds vs 2 titan black 43 seconds. so the 980ti is not slowing down.
The times change when you sometimes re-render 2 times so the timing in DS it not accurate in my DS it jump from 5 to 10 sec in 2 second of real time so I would not count on that
what you got exactly is what you see in the graph below .. based on % you gain adding each card , you got faster rendering but wasted resource
Here is one more with another scenario what is not exactly what happening for real (the top one )
In my case turning Optix Off make 2 Titan X as fast as 3 xTitan X with optix On so huge difference in rendering but not in all scenarios as having Optix on increase the speed in my viewport than when using it without
it also depends of what you render I mean what kind of materials is used . But I guess you profit from that more in Interactive Direct light mode than Photoreal
The tests from below are in Photoreal mode without Optimization samplers , and the results will be always different depends of the system and other factors .
any chance any of you post some example renders so we can see how much content - figures, backdrop, props extra are in your scenes and quality and then say what your systen setup is cards extra maybe have some basic scenes, medium scenes as well as more complex scenes including lots of characters doesn't have to be anything special looking just so we can judge the performance quality of the cards and rest of your computers specs probably help us more those planning on upgrading or buying new computers on what we want/need to do the type of rendering 3d creation more than reading just statistics
That is not a good idea , you need to focus on statistics not on renders , the quality of a render is based on the time you spend to render it so you can have cheap GPU and render it as long you want to get the sane quality as with GPU that is 10 times more expensive
I am building my rigs for 18 years , a quality cards means nothing if your system is outdated , you need to focus on what processor you have or want to have, how much lanes it have then choice the right motherboard that support it and the numbers if GPUs you want, how much memory and how much power Supply you have and then based on that you can choice the cards to match your system for best performance and when you cheat on some parts then you are at risk to kill your system sooner or later or you will not even boot .
We still don;t know how well 1080 will perform with iray so if you about to upgrade just wait and then you can make best choice based on real iray benchmarks and not by counting seconds in DS or look at pretty renders that is how things get bad and you waste your money vs performance , You can still buy cheaper cards for the same money you spend on high end GPUs and still have faster rendering power but the only bad side of this is that the cheaper GPUs don''t have as much video memory . In my last rig I skipped the 760 2 GB card for less cudas card, since I use it just for 2 monitors and the 4gb will be enough for the OpenGL in DS so I have my GPUs just for rendering power
from all rigs I build in the last 1 and half years for iray the best for the money was cheaper 2 GPU system with 4 core processor and 32GB memory .
here are my 3 rigs I build from the past 18 months
---------
_____________________________________________________________
____________________________________________________________
future upgrade to full open water cooling system for All GPUs and CPU