DAZ Studio Cloud

StratDragon

since Nvidia is still going to manufacture Maxwell cards I would think the 1xxx series of cards wont see much of a price drop except for those who paid extra for the founders editions that are shipping as CUDA tools for those cards are still awaiting release. The card to watch right now might be the GP102

http://wccftech.com/nvidia-pascal-gp102-gpu-titan-graphics-card/

Drip

Silver Dolphin said:

I have 3 Nvidia 780 6gb editions and i render my Iray scenes with HDRI lights in under 3 min. How is Cloud going to help us hobbiest do still renders. I think this is geared toward Pros and people with commerical needs or animations where time = money. Hobby users unless they have deep pockets will not use this. Now here is a thought! If Daz had a way for us users to sell our render capabilties with our cuda cores to other Daz users so they could speed up their renders that would be great. We could sell render time for credit at Daz store if you could meet the render requirements for that render. It would come in as a encrypted package (not viewable by you) and be rendered by your Cuda cards and go back to Daz where you would be credited for the render once it was done. On a side note, I hope Nvidia gets off their duff and makes the 1070 cheap and Iray capable soon! A 8gig of video ram Nvidia card for under $400 with lower power consuption sounds great.

It really depends on the hobbyist. Some hobbyists render a lot and have a decent rig that can render a complex scene in just 2-3 hours. Other hobbyists make only a few renders a month on an old rig that takes more than 10 hours to converge to 99.5% on a simple scene with just one model.

To some, improving their rig isn't just a matter of spending a few hundred bucks on a pair of new graphic cards, but requires upgrading their entire system. A $30 render service would then make a useful stopgap while saving up for a better rig, allowing them to render more complex scenes in a matter of minutes, as long as at least their bandwidth permits. And then there's ofcourse the people who just use a simple laptop with some integrated videocard, who wouldn't want that any different.

So a render service could enable people to take up rendering as a hobby. Not every leisure pilot has an airplane, not every leisure sailor has his own boat. People actually rent those things to enjoy their hobby. I wouldn't see an accessible Iray render service for DS any different from that.

kyoto kid

...indeed.

I create vere "big" scenes in the terms of poly count and texture detail as well as render at high quality settings.  If Daz Really plans to open an Iray render service I'd be for it.  30$ a month I can scrape up.  Several thousand to build a new system with a good miulti core/thread CPU, a GPU card capable of handling the scenes I create, and enough memory and drive space to back that up, well that is another matter entirely, paticularly on my income.

nonesuch00

I use 10 hours doing one render, $30 for 10 hours is not economical for me but I understand it might be for Lucas Film or Marvel Comics or such places.

Kevin Sanderson

The idea is to do fast renders using all that cloud computing power in a small amount of time that would take 10 hours or longer on a weak personal computer.

Tobor

Drip said:

It really depends on the hobbyist.

I just wanted to start by saying thanks for spelling hobbyist correctly. The other spelling suggests something on the order of hobby, hobbier, hobbyest...

Anyway, two years ago I prepared a business proposal for some backers who do cloud computing for various industries, including entertainment. Daz had just introduced Iray to D|S, and I could see a cottage industry of render farms equipped with both CPU and GPU renderers for small-scale amateur and semi-pro productions. The main emphasis was directed at animation, since that's where the real clock cycles come in.

In the end they decided there wasn't enough money in this market segment, at least not yet. They wanted to see $1M to $2M per month in revenues, or else why bother. The same company built render farms for at least one major Hollywood studio, and they were aware of the potential, but could not see how to scale it down to meet their revenue requirements while providing benefit to the amateur and semi-pro (e.g. YouTube) markets.

They also saw the interface option in D|S for the cloud beta, and that scared them off, thinking users would prefer the built-in system for convenience. Two years later, it still says beta, and with very little public information about it...

nonesuch00

Kevin Sanderson said:

The idea is to do fast renders using all that cloud computing power in a small amount of time that would take 10 hours or longer on a weak personal computer.

I guess they'd give the full specifications of each node on the network and how many nodes would be used at once to avoid wasting too much money on the part of the customer so if they do that a one-time sign-up for a month for $30 and cancel at the end of the month would be a good trial period but then no businesses are offering these services for DAZ Studio / nVidia products so maybe one day I'll try a one month run of a Blender Render Farm.

kyoto kid

..so they need a large render farm or maybe just a couple servers and VCA (which is designed for networking and has eight Quadro M6000s)?

ebergerly

Tobor said:

In the end they decided there wasn't enough money in this market segment, at least not yet. They wanted to see $1M to $2M per month in revenues, or else why bother.

Can't say that I blame them. I'd think it would be real tough to get significant revenue from the D|S/Poser community for something like this, espcially considering most of us aren't under signficant time pressure to produce our renders, and therefore can't really justify any signficant expenditures to do them faster. 

Maybe if DAZ included a few bucks a month in its membership (I forget the name of it...prime or something?) to cover a cloud rendering service it might gain some traction. But even then it's tough to generate significant revenue to justify the investment. 

Although I'm shocked they were looking for $1M or $2M PER MONTH in revenues? Really? I can't imagine even the big studios could lay out money like that, rather than investing in their own render farms, which they already have I assume. 

Luv Lee

Khory said:

I can see plenty of people who render for a large print size thinking that using a render farm for a couple of minutes vs having Studio busy for a couple of hours is a good thing. Pretty much anyone who ends up using a batch render set up would think it was worth it much of the time.

Oh Agreed, I would do it in a heart beat and  I am currently researching this to see what my options are.