In this role, you are responsible to build a highly scalable compiler toolchain for our future chip generations. You will participate in many fundamental architectural decisions and have the opportunity to contribute on upstream open-source projects.
WHAT YOU WILL DO :
building / maintaining ML compiling toolchain for our accelerator
design & implement various compiler operations as well as lowering and optimization passes
collaborate with ML & quantization engineers / researchers to realize compilation of generic NNs for hardware
WHAT YOU SHOULD BRING IN :
in-depth knowledge of C++
hands-on experience in compiler technologies like LLVM
HELPFUL BUT NOT REQUIRED :
experience with machine learning compiler infrastructure like MLIR / IREE or TVM
experience in polyhedral compilation techniques