Teaching SIMD Instructions Using Intel Intrinsics in Computer Organization Course
Event Type
Workshop
Big Data
Education and Training and Outreach
HPC Training and Education
W
TimeSunday, 14 November 202111:41am - 11:51am CST
Location241
DescriptionThis talk focuses on integrating high performance computing topics in Computer Organization and Design class at Marquette University. The instructor is a TCPP early adopter. SIMD parallelization was covered in the class using CPU and GPU. On Intel CPU, Advanced Vector Extension (AVX) was used for teaching vectorization at the assembly programming level. The content was covered in two fifty minutes lectures, one lab, and one homework. On GPU, CUDA was used with image manipulation example. We describe the topics covered along with programming exercises and homework assignments. The course materials, lecture slides and lecture video have been shared online.