Swapping TCP/IP for RDMA in GVirtuS didn’t just improve speed—it transformed GPGPU remoting. With Mellanox Infiniband, we slashed SAXPY execution time by 82% and Matrix Multiplication by 55%. The secret? Eliminating system calls and leveraging pre-registered memory.
🌟 Follow us on LinkedIn! https://www.linkedin.com/company/clever-project/?viewAsMember=true
🌟 Check the updates from the website: www.cleverproject.eu
🌟 Full paper in: https://zenodo.org/records/14717622