blazingdb/blazingsql repository overview

⁠BlazingSQL

BlazingSQL is a GPU accelerated SQL engine built on top of the RAPIDS data science framework. RAPIDS is a collection of open-source libraries for end-to-end data science pipelines entirely in the GPU. BlazingSQL extends RAPIDS and enables users to run SQL queries on Apache Arrow in GPU memory.

Prerequisites

• GPU support
  • Pascal or better
  • Compute Capability > 6.1
• CUDA support
  • 9.2 or better
• OS support
  • Ubuntu 16.04 LTS (tested and confirmed working)
• Docker support
  • Docker CE v18+⁠ - apt for Ubuntu 16.04 doesn't include v18+ by default
  • nvidia-docker v2+⁠

Usage

1. Pull the BlazingSQL docker image: 

docker pull blazingdb/blazingsql

1. Launch an interactive session:

docker run \
  --runtime=nvidia --rm \
  -p 8888:80 \
  -p 8787:8787 \
  -p 8786:8786 blazingdb/blazingsql:latest

1. Open your web browser, and navigate to:

   http://localhost:8888⁠
   or http://{IPADDR}:8888   (e.g.)  http://12.34.567.89:8888⁠

You will find an example notebook that will run a demo test. They are under /blazingdb/notebooks/

More information https://docs.blazingdb.com/docs/using-blazingsql⁠