The Open Renderfarm Project

General Forums Blender and CG Discussions
1

I opened a poll on this page which I believe everyone should look into: http://blenderartists.org/forum/showthread.php?t=185164

Hello everybody.

I have an idea for a project which might be of interest to the entire Blender community and that is the Open Renderfarm Project. The basic idea behind the project is to create a community-funded render-farm which at all times would be available to the blender community for rendering of 3d-scenes and animations created in Blender.
I personally think this would be good for the community and the creative works of fellow blender-heads to have access to such a render-farm.
I would personally like to work a lot on this project, running the rendering service where all of you can render your work, but before I start I want to hear from you. Are the users interested in such a render-farm? Do you think it would be possible to get funded by the community?

If the community is positive to such a project and I get constructive feedback I will of course start looking for sponsors in the hardware-market(please tell me if you have ideas) and create a project web-page for fundraising and updating the community on the progress and life of the render-farm, I look forward to hearing from you.

Sincerely,
Bjørn Ihler

P.S. I will update this first post with ideas and information to keep you updated on the developments of this project.

edit: I am aware of services such as http://www.vswarm.com and renderfarm.fi. What I am thinking of is a traditional render-farm rather than clients running on peoples home-computers, providing a stable service for professionals and amateurs using blender for their daily work but without using peoples personal computer-power to provide as good a service as possible.

edit II: as said by Comeinandburn later in the thread

It would be great if it was an all web interface with NO installs. Just simply upload your packed blend and download the frames when finished. This way there are no java installs or OS client problems.

I agree with him and the plan is to make this renderfarm so that the only things you need locally on your computer is blender and a web-browser-

edit II: as you can see further down on this page the question of if it is possible to get this funded or not has come up, I am of course trying the best I can to push prices down by getting funding(if any of you want to help with this please PM me so that we can coordinate our efforts to find sponsors etc). I want the price to be as low as possible for a service as good as possible to the end-users which are you. So a question is, how much are you willing to pay? I quote myself from a later post in the thread:

One possibility would be to limit the number of free renders for each member to 5-full frames a month, so you would have to pay for rendering of animations etc, and then apply the system of those paying 10$ getting unlimited single frames, but lower priority than those paying 20$ etc. For those rendering animations I think I might have to base the price on the amount of render-time. But as I mentioned earlier, I am trying to push prices down in addition to getting substantial funding from multiple sources to keep this service as cheap as possible for the end-user.
The question is really, how much is the end-users willing to pay for a Blender-Specialized render-service, and do I have a market?

2

I just love the approach by renderfarm.fi to have it built in the GUI to hit render animation, you bundle and send project onto the renderfarm and get back the seq/mov animation later.

and can continue to work!

Im thinking of trying to go commercial with 3d done in blender and motiongraphics, I would rather see a farm with a really good web interface, and schedule to book.

I’d rather see a farm that’s not just a playground, and that you can bounce out peoples “tests” with commercial projects that has a deadline and you don’t want to put in jeopardy.

otherwise than that I think people can book time to just, render heavy scenes or frames on it.

Another aspect would to get sponsors, have a great idea. I want the farm I use to be as green as possible and maybe make some noise on the internet before setting up the farm. Maybe a service provider or HW manufactor will help out if there’s some PR to gain.

3

There are a number of these set up already in more of a distributed rending environment.

One is http://www.renderfarm.fi/

These are not really a traditional renderfarm, but it accomplishes the same tasks I believe.

4

What aemartin said. Also http://www.vswarm.com is another option, still waiting for the 2.5 Linux client.

5

I saw renderfarm.fi and it looks like it depends a lot on the computer of the voulenteers, what I am thinking of is a more traditional renderfarm as you speak of LiquidApe, where all the hardware is at one place and ran as a professional render-service rather than a cluster of home-computers which are not really optimized for rendering, please tell me if I have understood the concept of renderfarm.fi wrong.

@aermartin, the renderfarm will most likely be run on energy made by water-falls which is a renewable, clean and environment-friendly power-source. If I go through with this project I will definitely try to make more noise on the web etc. to get sponsors. I will try to create as good an interface as possible and would like your advice on what you think of as a good interface so that this can be used in the development of the farm.

6

yes I will love this since I will soon will be doing some shorts of my own and that will be really good. Also I really like the idea of booking it or it could be special days that are available to the public and other days are for people who pay small funds and book it.

7

@retro world; yes, that is a possibility but before doing that I would like to know the public opinion on the renderfarm-project and if anyone has any opinions on your idea I would have to take them into consideration before we in the spirit of OpenSource somehow reach a decision.

8

yes I understand, I really do think it will be useful to all. for example the large scenes that I couldn’t render on my 512MB RAM pc could be rendered there.

9

@retro world; yes that is the point of the project :slight_smile:

If you want to help the project move forwards one of the best ways to do so is currently to spread the word of this thread and get people to participate in the debate:D

10

how about something like what blendercookie.com does. You pay a citizen membership that gives you access to the render farm and jump the line when you have pressing renders. Still have the farm available to users for free but you could limit the number of frames rendered and time per frame to keep it from being bogged down.

Maybe even try to make a partnership with the blendercookie guys that would get you exposure and also give people access to their project files (through the citizen program). That would be real added value even if you don’t use the farm all the time.

good luck!

p.s. just keep it affordable like $10 a month. Something like this isn’t hard for people to justify for occasional use.

11

I would totally download a Seti@home type screensaver renderer. But I have no idea how or if it would work XD

Maybe it’s an earned thing? Hours you’ve logged equals hours you get? Sure, you might want your own comp rendering once you’ve hit postproduction, but you could bank on your early overnight contributions to a cloud meaning saved up points for your own rendering.

12

@comeinandburn; that is actually quite an appealing idea, what would common users think about this? Another thing is that I would need some money to get this started, I could create a deal which is something like, if you contribute by donating money (or hardware) you will get prioritized render-time(amount based on how much you donate), this deal could go on forever, money earned will of course be used for improving and running the farm.

@ogbog; I think you are thinking about farms running like renderfarm.fi using the users computers, as I have mentioned earlier I imagine this as a service running on a purpose-made farm rather than on the user’s home computers thereby giving a more stable running render-farm and letting the users use their computer-power on other things like modeling or whatever they wish.

13

Have you given any idea to how users would access this farm?

It would be great if it was an all web interface with NO installs. Just simply upload your packed blend and download the frames when finished. This way there are no java installs or OS client problems.

The University of WestmInster had a wonderful free rendering cluster in the past that did this…unfortunately it’s no longer running. Here’s the link the blendernation article http://www.blendernation.com/free-rendering-on-250-cpu-grid/

As an end user I think it’s important to see something, even if you have a modest renderfarm in a beta form (a few inexpensive dedicated servers) before you could expect people to start donating. Just to see that this will work and that you’re serious. You can always add to the farm as it develops.

hope this end user input helps.

14

@Comeinandburn; thank you, my plan was to not require any installs just as you describe. Thank you for the link. Before being able to build a farm I at least need some support in the community to build agreements with sponsors, but I agree, I can not expect people to start donating just yet(but I can wish them to). I will start looking today for sponsors and monetary support so that I can build some nodes as soon as possible and then later expand the service based on donations and new deals with sponsors.
This may be a great way for future sponsors to show off their hardware to a great and large community, I hope they recognize this and indeed that I am able to get support from HW companies, but I am also interested in monetary support from other companies.

15

@Polarbear if you want to create a good service, I think you have to make it fully commerical. Render farms are expensive to set up and keep running. Equipment cost, cooling cost, electricity cost will add up. Now if you are going to be charging and allowing some people to use the service for free at times and others to pay for better serivce. You will be robbing Paul so Peter can have his free render.

I will be starting work on a short towards the end of this year and will happily pay for rendering services, I figure it will be cheaper than building my own mini render farm but if I am paying for something I damned quality.

The are lots of commerical services that can be built around blender and this could be one of them.

16

@Tyrant Monkey I do indeed want to create a good service and I am currently researching funding-opportunities. I am aware of the costs of equipment, cooling, electricity et cetera. I do however not think that letting people who donate/pay get prioritazion(a better service) so that their renders get finished faster and first is the same as robbing Paul so that Peter can have his free render as you say. I think Paul is willing to pay for a better service, which he could have had for free, but then at a slower speed, especially if Paul is creating an animated short-film which requires huge render-capacity within reasonable time.

another solution as I see it is to create different membership-groups with different access to the service. Say those who want to render for free gets the lowest prioritazion (as suggested earlier I coulld have one day pr week for free rendering or something) if you pay 10USD a month your render can be rendered every day of the week, but members paying 20USD a month will get prioritized, and so on, the more you pay the higher prioritazion and better service you will get.

The reason for keeping it free is because this opens the possibility of geting funds from the Norwegian Goverment, and will increase the interest of sponsors. With a deal as described above the render still willl be “free” thus have the same opportunities for funding as keeping it entirely free.

please comment and enlighten me if I have gotten anything wrong or you have other ideas for funding :smiley:

17

Do governments pay for such things?

If you give away freebie renderings rest assured that the freebie guys will out number the payers by about (5000:1). Everybody loves a freebie, so unless advertising or government funding makes up the short fall you will have to make cuts some where. The easiest place to make cuts will be in equipment meaning the quality of service you will provide will suffer due to your altruism.

Unless you are sure you can subsidise the freebie user with advertising and government funds I would be wary about been so giving. As the say there is no such thing as a free lunch. Somebody out there is going to have to pay for those that make free use of your service now will it be your paying users, governement funding or advertisers I don’t know. I just think relying on the latter two is kind of wishful thinking.

18

@tyrant monkey. the freebie guys will probably out-number the payers by about (5000:1) yes. I am working to get it as cheap as possible to run with the best equipment possible for low costs, but it seems like I will be depending on funding from advertising and government etc. to be able to pay for it, which ofc. is unrealistic, but I think it at least is worth trying, if not to get the farm free for the users at least to lower the costs.

One possibility would be to limit the number of free renders for each member to 5-full frames a month, so you would have to pay for rendering of animations etc, and then apply the system of those paying 10$ getting unlimited single frames, but lower priority than those paying 20$ etc. For those rendering animations I think I might have to base the price on the amount of render-time. But as I mentioned earlier, I am trying to push prices down in addition to getting substantial funding from multiple sources to keep this service as cheap as possible for the end-user.
The question is really, how much is the end-users willing to pay for a Blender-Specialized render-service, and do I have a market?

19

I think this kind of sounds do able. I would warn you though that cheap is not always the friend of the small businesses. By providing a cheap service you margins are smaller and you make an opening where a far larger company can come and price things lower than you. Bigger companies(the Wal-marts of the world) do cheap far better than the little guy.

Offering fast service, techincal support etc could be a better way to go. You need to find something other than pricing alone that would attract people to your service.

The blender.org site has links to professionals that use Blender, most are small studios where this kind of service if done well would be welcomed. You would need to do some research.

20

Thank you for your advice tyrant monkey; I think I based on your input will launch a survey, poll based on this forum and e-mail-based in getting in touch with end users such as the companies mentioned on blender.org. I have started trying to get in touch with some potential sponsors, but have not yet received any replies. My final decision will be based on the replies from the potential sponsors and from the survey.

I am aware of the wall-mart phenomena and yes I recognize the fact that other factors than the price may be the way to go when it comes to marketing, what my research/survey result in will as I mentioned earlier define what kind of a product I will be able to launch with and then we can take it from there.