Can I Cook a 5 o'clock Compiler Cake and Eat It at 2? - Optimizing Compilers for High-Performance Computing
Save Big on Coursera Plus. 7,000+ courses at $160 off. Limited Time Only! Grab it Explore the challenges and potential solutions in compiler construction for high-performance computing in this thought-provoking talk by Albert Cohen. Delve into the question of whether it's possible to create a compiler that reduces performance engineering efforts while still delivering competitive results. Examine the limitations of current compiler frameworks, the impact of ML-enhanced compilers, and the role of domain-specific code generators in numerical libraries. Consider the ongoing struggle between programmability and performance, and discuss the scalability of ML-based heuristics for large-scale code compilation. Gain insights into recent successes and partial achievements in academia and industry, and contemplate future research directions to address these challenges in compiler development and software engineering.