Axiom 3: New Sparse Simulations & GPU Performance Boost

Axiom 3 released
Axiom 3

Matt Puchala has unveiled Axiom 3 for Houdini, a significant update to his volumetric fluid simulator. This latest version enables true sparse simulations, leveraging the GPU for rapid creation of captivating fire, smoke, and other pyrotechnic effects. Here’s a look at the key features and enhancements:

  • 16-bit float fields, reducing memory footprint by half.
  • Native Cuda support for Nvidia GPUs.
  • Native Metal support on macOS for compatible devices.
  • Native Arm support for Apple Silicon devices.
  • Enhanced device and system memory usage, allowing the solver to utilize system memory in addition to GPU memory, crucial for large simulations. This is available when using Metal or Cuda.
  • Larger neighbor tile cache sizes, resulting in faster performance for high-resolution simulations.
  • Faster advection tracing.
  • Improved project non-divergent, leading to greater stability and reduced artifacting.
  • Overall solver performance improvements, ranging from 8% to 32% faster.
  • Introduction of a 3rd disturbance and turbulence force.
  • A disturbance seed parameter.
  • Velocity drag functionality.
  • Source shape transform handles.
  • Source shapes can now inherit the orientation of their input.
  • Plane type source shape, affecting anything below the plane.
  • Source shapes can composite values when overlapping, enabling multiple influences with different forces.
  • Updated to the latest NanoVDB version.
  • Numerous small performance optimizations.

For further details about Axiom, explore the resources available from the developer.