Jiqun Tu

Jiqun Tu received a physics PhD on lattice QCD from Columbia University before joining NVIDIA as a senior compute developer technology engineer. Currently, he is working on using the Tensor Cores, available on the latest NVIDIA GPUs, to accelerate a broad range of HPC applications.

Posts by Jiqun Tu

AI / Deep Learning

Running Python UDFs in Native NVIDIA CUDA Kernels with the RAPIDS cuDF

In this post, I introduce a design and implementation of a framework within RAPIDS cuDF that enables compiling Python user-defined functions (UDF) and inlining… 12 MIN READ