hipDF#
hipDF enables GPU accelerated DataFrames and DataFrame operations. This library, built on top of ROCm™, enables large scale data processing on AMD Instinct™ GPUs, allowing you to perform many data manipulation operations at breakneck speeds. Built on the familiar Apache Arrow memory format and using APIs similar to the well known Python Pandas library, hipDF allows you to not only build accelerated data processing workloads, but accelerate your existing Pandas applications with minimal effort.
By adopting the well-known cuDF API on AMD hardware, hipDF ensures compatibility and ease of use across various computing environments. This API compatibility enables existing cuDF workloads to be effortlessly transitioned to run on supported AMD devices, allowing you to use the ROCm platform for your data processing tasks.
hipDF is currently in an Early Access state and is intended as a preview to the production ready release planned for later this year.
