FLAMEGPU2
FLAMEGPU2 copied to clipboard
FLAMEGPU
FLAME GPU 2 is a GPU accelerated agent based modelling framework for CUDA C++ and Python
Minimise the impact of disk usage by buffering output, and/or asynchrenous disk output, so that the simulation can continue whilst disk IO occurs. This will probably be an extension to...
Example showcasing the use of network edge based communication. I.e how road network microsimulation can be implemented. Depends on #68
The ValidateIDCollisions method / kernels(s) run when agent populations are set. If the population was created at runtime (i.e. not loaded from disk) and theh population was empty originally then...
Currently we dynamically link against msvc standard library, this means users require visual studio 2019, or vs2019 redistributable installed. If/when we distribute binary releases, we will either need to switch...
Unlikely, but due to the windows dlls having such generic names e.g. `sdl2.dll`, `glew.dll`, it's possible a user will already have these on their system path but different versions. This...
Whilst discussing how we'll require multiple python packages with @ptheywood, came to recognise there would be value in highlighting to users when they are using a library build with `SEATBELTS`...
Would be useful to have something to indicate progress of runtime compilation. Due to where RTC occurs, perhaps difficult to do a combined progress console indicator. But adding a started/finished...
How are we going to generate api docs for the swig interface? SWIG does appear to copy across doc comments to the generated python, but we haven't actually tested or...
Once #379 is merged, there are still improvements that can be made to the use of streams to improve performance through the use of streams, within a single simulation and...
D2H / H2D transfers see increased bandwidth by transferring to/from pinned memory (2x+). The downside is that pinning too much memory on a given system can lead to excessive paging...