Abstract: As computer architectures and application paradigms continue to evolve, compiler technology plays an increasingly important role in hardware–software co-design. Its research scope and technical approaches have also become more diverse. This paper systematically reviews modern compiler technologies from several perspectives, including compiler infrastructures and intermediate representations, program analysis and classical optimizations, backend code generation, memory systems and data layouts, domain-specific compilation, autotuning and learning-based methods, and runtime coordination. Based on this review, the paper summarizes the development trends of these techniques across different computing platforms, methodological approaches, and application scenarios. It further analyzes how emerging hardware architectures, increasingly complex application requirements, and data-driven methods are shaping the evolution of compiler technology. In addition, the paper discusses major challenges in scalability, optimization effectiveness, and system-level coordination. This study provides a systematic reference for understanding the current research landscape and future directions of modern compiler technology.