Yi Xu

> Can this be done by passing a Python AST object to Taichi? I can convert the Julia kernel function to Python AST at the Julia side. Yes. You can...

cc: @re-xyr

I think this is definitely some optimization opportunity, which may or may not be taken care of by the backend compilers. However, is this the solution to the current problem...

> Partially yes, if this optimization is possible, Vulkan backend could benefit. (CUDA backend does not need it as the runtime opt is too powerful). Then feel free to hack...

It seems that my PR last month hasn't been merged yet (https://github.com/taichi-dev/quantaichi/pull/3) :-) cc: @k-ye @yuanming-hu

I assume the problem has been resolved so I'm closing this issue. Feel free to reopen if you meet further problems.