Compiling FLIP Fluids for Blender 2.81 (Windows x64)

Screenshots and notes as l compiled this add-on, for Blender 2.81

Compiling FLIP Fluids for Blender (Windows x64)


Master flip fluids file downloaded from Github – click here    samples click here

Visual Studio 2017 Community Edition 15.9.17 – click here

CMake-gui  3.16.0-rc04 – click here

Windows 10 x64

Nvidia Cuda 10.1 + OpenCL 1.2 installed














































After a lot of issues with the process above, l discovered below and successfully got the addon compiled


5 days, and still baking a example scene at 400 resolution!!!

Cascading Water Feature scene





Gooseberry Benchmark 1.0 – Results – Blender 2/79b – Blender 2.8 beta

This is a CPU only render test.  Click here for the download source.

Recommended a minimum of 12 GB RAM
Creative Commons Attribution 4.0

PC 1:

  • Intel i7-2600K CPU@3.4Ghz    32Gb RAM  Windows 10  GTX1080 8Gb
  • Blender 2.79b
  • Time:  1:38:20.61         1 hr 38mins


PC2 – A17R5:

  • Intel i7-8760H CPU@2.20Ghz  16Gb  GTX1060 6Gb Windows 10
  • Blender 2.79b
  • Time:  58 mins      58:34:08

using the latest [as of today] Blender: 2.8 beta, the CPU rendering engine time is greatly improved to:  34 minutes ! – an improvement of 24 mins !



Lighthouse – Flip Fluid Bender Simulation – Using Cuda w/Issues Found

Using Blender and FlipFluids Addon

This is still baking, so far 60Gb bake files and climbing at 334 out of 501 frames. Domain resolution used – 500 – ie High Quality.

Bake files grew from 55Mb [frame 1] to currently 277Mb [frame 333] !

Will update on the progress and final render.

and finally… after almost 2 days baking non stop…..

Below shows me trying to render one single frame and yet getting out of CUDA memory error.

I’m using a GTX1080 with 8Gb VRAM, on a i7 Quad core PC with 32Gb RAM

Switching to a CPU render and the results are below.

Time taken 1 hour for frame 348

Until l can find a quicker way to render each frame, by cuda which is usually faster especially since l have access to about 3000+ cuda cores collectively on my core workstation, l will give this project a rest for now.

update 9/28/2018

I tried command line blender rendering of frames 16, 348 with my 8gb gtx1080 / l  get the cuda error as the gpu ram used reached about 7gb out of the 8gb,

Currently rendering the first 200 frames as shown below.



Blender Open Data

I’ve used several benchmark software to compare computer hardware especially nvidia GPU cards used for cuda programming.  One good free benchmark software is Maxon CineBench    , another is Unigine Haven


Here are surprising results generated from the new Blender Benchmark under its Blender Open data website program.

Quick summary/my findings:

  • less render time running Linux [LinumMint] than under Windows 10 [all things turned off], same hardware
  • SSD’s help

Linux run: Workstation with GTX1080 8Gb, GTX970 4Gb, 32Gb RAM, Quad core Intel processor. No SSD

Same hardware as above but run under Windows 10

Ran on same hardware, only using the GTX1080 8Gb , under Linuxmint – impressive resuly showing in my view and from cuda/nvidia research l did that the 4Gb of the 2nd graphics card limits the performance.



Home PC #2  Results below, newer PC, no SSD, 2 x GTX560Ti

Below results from Alienware 17 [4 years old] with Samsung Evo 500Gb SSD, GTX860M



Trying out Blender      Flip Fluids      – currently baking below lighthouse.blend sample file, 2-3 days so far, about 10 hours go go and 89Gb cache space used up!

Final space cost and rendered frame one, as of 5/29 6am [frame 0 rendered in 2.5 mins]



After 60 frames l got….


Updates 5/30

Some Log entries:

28-May-2018 23h12m33s COMPLETE Update Marker Particle Velocities
28-May-2018 23h12m33s BEGIN Delete Saved Velocity Field
28-May-2018 23h12m33s COMPLETE Delete Saved Velocity Field
28-May-2018 23h12m33s BEGIN Advect Marker Particles
28-May-2018 23h12m55s COMPLETE Advect Marker Particles
28-May-2018 23h12m55s BEGIN Update Fluid Objects
28-May-2018 23h12m55s COMPLETE Update Fluid Objects

*** Time Step Stats ***

Fluid Particles: 48044598
Fluid Cells: 7162540

Diffuse Particles: 11995192
Foam: 1968108
Bubble: 9991435
Spray: 35649

Pressure Solver Iterations: 185
Estimated Error: 3.84183e-007

Step Update Time: 161.546

*** Frame Timing Stats ***

Update Obstacle Objects 0.000s 0.0% |
Update Liquid Level Set 38.922s 5.4% ||||
Advect Velocity Field 84.487s 11.6% |||||||
Save Velocity Field 2.262s 0.3% |
Calculate Surface Curvature 99.262s 13.6% |||||||||
Apply Body Forces 0.573s 0.1% |
Apply Viscosity 0.000s 0.0% |
Solve Pressure System 200.169s 27.5% |||||||||||||||||
Constrain Velocity Fields 4.435s 0.6% |
Simulate Diffuse Material 73.749s 10.1% |||||||
Update Marker Particle Velocities 38.413s 5.3% ||||
Delete Saved Velocity Field 0.088s 0.0% |
Advance Marker Particles 63.663s 8.8% ||||||
Update Fluid Objects 0.000s 0.0% |
Output Simulation Data 1.887s 0.3% |
Generate Surface Mesh 119.589s 16.4% ||||||||||

Frame Time: 727.5
Total Time: 161269



So far l have only been able to render up to 149 frames, then l get a cuda out of memory crash. I use both a GTX970 with 4Gb and a GTX1080 with 8Gb, l’ve tried with both and with each cards. Will keep you updated, meanwhile heres a render of the 149 frames









OpenExr notes – Tears of Steel – Blender Production notes


OpenEXR half float files, in [16-bit] 4096 x 2160 pixels.
Pictures have been shot using the (4k native sensor) fantastic Sony F65 camera.
The raw files were converted with Sony software to OpenEXR, using ACES color.
We then converted these with OpenColorIO to Rec709 “scene linear” which we further used for the movie pipeline.

Source:  Click here   OpenEXR



Tears of Steel Notes

Blender Foundation tears of Steel notes l found useful.

“80,000 frames, each in OpenEXR half float files, in [16-bit] 4096 x 2160 pixels. ”
Pictures shot using the (4k native sensor) Sony F65 camera.