pic

Hi,

one could and should experiment with Interprocedural optimization (IPO) also known as link-time optimization (LTO), especially on the host side for smaller binaries and potentially faster code. It's supported by GCC, Clang, and ICC, among others, which are our typical go-to compilers in HPC.

It's very easy to implement as well

Need for a reduced diagnostics addition: general 2D particle histogram (think: "np.histogram2d").

• Applications: phase space, energy-emittance plots, etc.
  • Would be useful for WarpX, ImpactX, and HiPACE++
• Size: fit into local GPU memory, e.g., default 1024^2
• 3 parser functions: axis1, axis2, filter
  • All three into the kernel
  • No RNG needed
• Can be loosely on the exiting red