Save Big on Coursera Plus. 7,000+ courses at $160 off. Limited Time Only! Grab it Explore a comprehensive 33-minute conference talk that delves into optimizing data processing performance through various Python libraries. Learn about switching BLAS versions in NumPy, leveraging GPU acceleration with CuPy for Linear Algebra, upgrading from NetworkX to RetworkX and cuGraph, and transitioning from Pandas to alternatives like Modin, cuDF, or Dask. Discover the advantages, pitfalls, and hidden inefficiencies of each tool while gaining practical insights into selecting the most appropriate solution for specific data science tasks. Computer Science and AI researcher Ashot Vardanian, who specializes in High-Performance Computing and Systems Design, shares expertise in GPGPU Algorithms, Compilers, SIMD Assembly, and Linux kernel technologies, drawing from his experience as both a startup founder and open-source contributor.