Ailing Zhang

Hey @helloworldstone ! Thanks a lot for trying it out! 1. taichi's generated spv file is no different from standard spv file, but it does has assumptions about how to...

To give a bit context on why we moved examples from top level into python dir: When we had examples in top level, everytime you build taichi it copies the...

+1 on implementing #2 as a start! Btw I feel like based on the deployment need, not modifying computation code might not be as hard requirement as we thought. IMHO...

@k-ye Yup that wrapper is mainly used to solve the floating point atomics problem we've seen. Is it correct understanding that to achieve much larger scale simulation on mobile, we...

minor nit: for subtask b.2, I wonder if `ti.types.vector(n, dtype)` where `ti.types.quant.int/float/fixed` are added to the whitelist of `dtype` makes it simpler for users?

+1 on `quant.int`!

+1 that we should try to delegate to LLVM for this optimization if possible. Thanks for the detailed proposal! nit: the llvm code in optimized implementation should be ``` [llvm]...

Hey @helloworldstone , wow it's pretty impressive to see you get it up running! We haven't done any benchmarks comparing to opencl yet, would you mind sharing your benchmark repo...

FYI @PGZXB #5889 introduced an extra dependency from `gfx_runtime` to `spirv_codegen` which is unexpected. You can reproduce this by `TAICHI_CMAKE_ARGS="-DTI_WITH_VULKAN:BOOL=ON -DTI_WITH_OPENGL:BOOL=OFF " python setup.py develop` and then `import taichi`. Specifically...

@helloworldstone Our implicit_fem demo is based on vulkan runtime and opengl support is on our todo list but we're still working on it. Is [vulkan backend](https://github.com/taichi-dev/taichi-aot-demo/blob/master/implicit_fem/python/implicit_fem.py#L15) sufficient for your android...