Pak Markthub

Pak Markthub is a senior software engineer at NVIDIA. He leads the research and development of GPUDirect Async—Kernel Initiate and Kernel Submit technologies, in addition to assisting the development of other core GPUDirect technologies. Pak received a Ph.D. in mathematical and computing sciences from the Tokyo Institute of Technology, Japan. He has spent nearly a decade in the high-performance computing (HPC) field related to GPU communication technologies.
Avatar photo

Posts by Pak Markthub

Networking / Communications Oct 06, 2026

How DOCA GPUNetIO Unifies GPU-Initiated Networking Across the NVIDIA Software Stack

GPU applications increasingly need networking and data movement to behave like first-class GPU-controlled operations rather than host-driven services. When the... 19 MIN READ
Networking / Communications Nov 10, 2025

Fusing Communication and Compute with New Device API and Copy Engine Collectives in NVIDIA NCCL 2.28

The latest release of the NVIDIA Collective Communications Library (NCCL) introduces a groundbreaking fusion of communication and computation for higher... 9 MIN READ
Networking / Communications Sep 06, 2024

Enhancing Application Portability and Compatibility across New Platforms Using NVIDIA Magnum IO NVSHMEM 3.0

NVSHMEM is a parallel programming interface that provides efficient and scalable communication for NVIDIA GPU clusters. Part of NVIDIA Magnum IO and based on... 7 MIN READ
Simulation / Modeling / Design Nov 22, 2022

Improving Network Performance of HPC Systems Using NVIDIA Magnum IO NVSHMEM and GPUDirect Async

Today’s leading-edge high performance computing (HPC) systems contain tens of thousands of GPUs. In NVIDIA systems, GPUs are connected on nodes through the... 14 MIN READ