Kyrylo Perelygin

Kyrylo Perelygin has worked on features such as CUDA's Multi-Process Service and the new Cooperative Groups since joining NVIDIA, where he first interned in 2013. Kyrylo graduated with a bachelor's from EPITECH and obtained a master's from CSULB.
Avatar photo

Posts by Kyrylo Perelygin

Data Center / Cloud

Improving GPU Utilization in Kubernetes

For scalable data center performance, NVIDIA GPUs have become a must-have.  NVIDIA GPU parallel processing capabilities, supported by thousands of... 15 MIN READ
Simulation / Modeling / Design

Cooperative Groups: Flexible CUDA Thread Programming

In efficient parallel algorithms, threads cooperate and share data to perform collective computations. To share data, the threads must synchronize. The... 16 MIN READ