At Rebellions Inc, I develop compilers for Neural Processing Units (NPU), focusing on performance optimization and hardware efficiency. Currently exploring efficient methods to deploy Large Language Models on NPUs, while developing generalizable optimization techniques. My work combines deep learning optimization expertise with compiler development to maximize AI acceleration capabilities.
Skills: MLIR/c++/Pytorch
- ๐ญ Working as a Compiler Engineer at Rebellions Inc
- ๐ซ How to reach me: jaehunryu@icloud.com