Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
This conference talk explores the SIMD (Single Instruction Multiple Data) programming model, comparing ARM SIMD implementations (NEON & SVEx) with X86 SIMD variants (SSEx, AVXx) as critical software performance optimization techniques. Learn about the fundamental concepts of SIMD vectorization across both ARM and X86 platforms, examining their respective advantages and disadvantages. Discover practical code vectorization examples and potential applications across various IT industry sectors where performance is paramount. The presentation addresses challenges when working with NEON and SVEx on ARM systems (both edge computing and HPC platforms), discusses SIMD programming limitations, and highlights key design factors for optimizing workload performance. Examine before-and-after optimization results and gain insights into the pros and cons of SIMD approaches, with a look at potential future improvements in this important performance optimization domain.
