이재진 연구실

Parallelization and optimization of deep learning frameworks to overcome the programming wall and accelerate large-scale deep learning applications using compiler, runtime, and architecture techniques.

Parallelization and optimization techniques for large language models (LLMs) to efficiently exploit heterogeneous processor resources in GPU-based systems.

Development of programming systems and tools for heterogeneous machines that use GPUs and FPGAs as accelerators alongside general-purpose CPUs for efficient parallel computing.

Design and development of programming and simulation environments for quantum computing platforms to enable efficient quantum software development and execution.