This page will help you to estimate the cost of a Cooperative Project. First, let's calculate the Sampling Level (SL) gain you can expect from the RANCH, compared to your system. Download the 8BALL scene and run it on your system with the default parameters (5 minutes limit, threads=0, 1024 x 576). When this is done, enter the SL reached in the 'Your SL' field, then click on 'Compute'. The SL difference between your system and the RANCH - let's call it DELTA - will be displayed. You can apply this DELTA to all your projects. For instance, if your system takes one hour to reach a given SL, you know that in the same amount of time, the RANCH will reach this SL + DELTA. The table below shows a few typical DELTAs. The RANCH reaches SL = 22.2 in 5 minutes (Maxwell 1.7.1 x64) on the 8BALL.scene. When your system (below) reaches SL----------->The RANCH reaches... (in the same amount of time) Pentium 4C @ 3.4 GHz (1-core + HT)---------------->SL+15.8 Athlon 64 X2 @ 2.0 GHz (2-core)---------------------->SL+14.5 iMac Core 2 Duo @ 2.0 GHz (2-core)----------------->SL+14.2 Core 2 Duo E6600 @ 2.4 GHz (2-core)-------------->SL+13.9 Dual Opteron 285 @ 2.6 GHz (4-core)--------------->SL+12.4 Core 2 Quad Q6600 @ 2.4 GHz (4-core)------------>SL+12.2 Core 2 Extreme QX6850 @ 3.0 GHz (4-core)------->SL+11.8 MacPro - 2 x Xeon 5365 @ 3.0 GHz (8-core)-------->SL +10.3 A concrete example: Your SL on the 8ball test is 10. You want to reach a final SL of around 18 on your project: how much RANCH time do you need? Simple: in your case, DELTA=12.2. The only thing you have to do is render your project on your system for a few minutes, with a target sampling level of 6 (18-12.2 = 5.8, rounded to 6 as you can't ask Maxwell to render to a non-integer SL). After a short while, Maxwell will tell you how much time your computer will need to reach SL 6. That's all the information you need: the RANCH will take the same amount of time to reach SL = 18.2 (6+12.2). If it sounds good to you, you can then enter this render time in the cost estimator at the bottom of this page. Your SL (8ball, 5 mn) :  DELTA = :  (+- 0.2)   How to use the cost estimator below:

In the first field, enter an estimation of the preparation time of your project (this includes the voxelization phase and more generally the time needed for various calculations before the rendering itself begins). Two minutes are automatically added to account for the distribution time of the scene across the whole network. In the second field, enter the render time limit you will specify when you submit your project on the RANCH. In the third field, enter the size of the MXI file generated by your project, in megabytes (MB). To do this, simply launch a render at full resolution on your computer for a few minutes, until Maxwell has written on disk the first MXI. Then enter its size in the 'MXI size' field. On a large renderfarm, gathering and merging big MXIs from a lot of nodes can take a while. Please be very cautious with the Multilight option, as it can easily multiply the size of your MXIs by a factor of 4 to 6. Gathering and merging MXIs on the entire RANCH takes 3 minutes with 256 MB MXIs, 10 minutes with 1 GB MXIs and around 20 minutes with 2 GB MXIs. The individual MXI size limit that can be processed with our parallel multi-merging technique is around 2.5 GB. For an animation, you can take an average / representative frame of your animation and do the estimation, then multiply by the number of frames of your animation. But in that case do not count the gathering+merging time: there is none in animation projects, as each entire frame is rendered on a different node. Again, please be aware that this is a reasonable estimation, not a guaranteed result. If the render time of your cooperative project is >= 60 mn, the MXI gathering+merging phase is FREE!     Preparation time (mn):  (integer value > 0 and <= 180) Render time (mn):  (integer value > 0 and <= 720) MXI size (MB):  (integer value > 0 and <= 2500)