Sign in
Categories
Migration Mapping Assistant Your Saved List Partners Sell in AWS Marketplace Amazon Web Services Home Help

GPU Data Science Workstation

By: BlazingDB Latest Version: v0.15
Linux/Unix
Linux/Unix

Product Overview

A secure GPU data science workstation ready in minutes at the click of a button.

No drivers to install, no packages to setup, simply login to the automatically launched JupyterLab and start coding on GPUs with RAPIDS.

RAPIDS is an open-source data science framework designed to have a familiar look and feel for data scientists working in Python. Built on a GPU-accelerated version of Apache Arrow, RAPIDS enables end-to-end data science workloads to run on GPUs, yielding dramatic performance improvements at a fraction of the cost.

BlazingSQL is a distributed SQL engine built on the RAPIDS Apache Arrow dataframe. With BlazingSQL and RAPIDS, users can seamlessly build and scale Python data science workloads to terabytes and beyond.

BlazingSQL helped build and is a large contributor to the RAPIDS ecosystem to support our vision of creating an open-source SQL engine that provides an ETL bridge to GPU-accelerated data science.

FOR DATA SCIENTISTS
- Secure Jupyterlab deployment for quick interactive analytics within seconds.
- Code in familiar APIs such as SQL, Pandas, SciKit-Learn, NetworkX, and more, but in their GPU-accelerated counterparts.
- Dozens of examples notebooks to building end-to-end data science workloads on GPUs.

FOR DATA ENGINEERS
- Create complex data pipelines at scale leveraging BlazingSQL, RAPIDS, and Dask.
- Go from one to many GPUs for larger pipeline requirements.
- Robust SQL support simplifies accelerating old workloads onto GPUs.

Version

v0.15

Operating System

Linux/Unix, Ubuntu 18.04

Delivery Methods

  • Amazon Machine Image

Pricing Information

Usage Information

Support Information

Customer Reviews