Thank You RAPIDS Community

RAPIDS has had one heck of a year. I’m thankful for all the developers contributing to RAPIDS and all the users, who are just as important, who filed issues, blogged about RAPIDS, and shared their success stories. I also want to thank the PyData community, all the libraries that came before RAPIDS and work with RAPIDS to make this possible. Without Apache Arrow, Pandas, Scikit-Learn, NetworkX, Dask, Numba, GoAi, Chainer, PyTorch, Anaconda, BlazingSQL, and so many more, this wouldn’t be possible. In addition, I want to thank NumFocus, the Linux Foundation, the Apache Foundation, and all other non-profit organizations that help provide resources to the open-source community. Without these governing bodies, their organization, and their ability to collect financial and non-financial funding for projects, we wouldn’t have such a vibrant and thriving open source ecosystem.

The result of all the groups above, allows us to receive great joy for some user to tell us that things that were broken or not performant in prior releases have been fixed, and now exceeds their expectation. All of your hard work allows users to warm our hearts with stories of workflows that use to take weeks to run end to end, can now run in hours or even minutes.

Data Science is a hard, repetitive, time-consuming practice. Data Science is part science, part art, and those two forces are fused by the community. Again, thank you!

The best way I could thank the growing RAPIDS community for doing their part is to share a small portion of blogs, articles, and YouTube videos from over the year that highlight RAPIDS, the community, and the ecosystem of libraries we work with. Without further ado, here is a list of amazing examples of RAPIDS from the community:

Community Blogs, Articles, Podcast, and Videos:

Dask and RAPIDS: The Next Big Things for Data Science and Machine Learning at Capital One (Capital One blog from Mike McCarty, 3-minute read)

Quick Install Guide: Nvidia RAPIDS + BlazingSQL on AWS SageMaker (Towards Data Science blog from Iván Venzor, 4-minute read)

4 Graph Algorithms on Steroids for data Scientists with cuGraph (Towards Data Science blog from Rahul Agarwal, 12-minute read)

Where should I walk? (Towards Data Science blog from Tom Drabas, 6-minute read)

Here’s how you can speedup Pandas with cuDF and GPUs (Towards Data Science blog from George Seif, 3-minute read)

Accelerated with RAPIDS: Walmart uses RAPIDS to help develop better retail forecasting (Youtube video from Jon Bowman of Walmart Labs, 2:03 length video)

How Walmart Uses Nvidia GPUs for Better Demand Forecasting (Datanami blog from Alex Woodie, 3-minute read)

UMAP on RAPIDS: 15x Speedup: Exploring the new GPU acceleration library from NVIDIA (Medium blog from Ceshine Lee, 3-minute read)

Nvidia Rapids cuGraph: Making graph analysis ubiquitous (ZDNet blog from George Anadiotis, 5-minute read)

cuDF, cuML & RAPIDS: GPU Accelerated Data Science with Paul Mahler — TWiML Talk #254 (The TWIML AI Podcast with Sam Charrington and Paul Mahler, 38:13 length video)

BlazingSQL is Now Open Source (Medium blog from Rodrigo Aramburu, 4-minute read)

BlazingSQL Querying 2.6 Billion Rows (Medium blog from Rodrigo Aramburu, 2-minute read)

Python Pandas at Extreme Performance (Medium blog from yaron haviv, 6-minute read)

Dask/Numba Blogs, Articles, Podcast and Videos:

High Performance Python Processing Pipeline with Dask (YouTube tutorial video from Matthew Rocklin, 15:04 length video)

Interview with Dask’s creator: Scale your Python from one computer to a thousand (Medium blog from unbalancedparentheses , 7-minute read)

Episode #207: Parallelizing computation with Dask (Talk Python podcast interviewing Matthew Rocklin, 57:53 length video)

Dask Deployment Updates (Dask Working Notes)

JIT fast! Supercharge tensor processing in Python with JIT compilation (Medium blog from Chris von Csefalvay, 11-minute read)

Python, Performance, and GPUs (Towards Data Science Blog from Matthew Rocklin, 7-minute read)

Composing Dask Array with Numba Stencils (Dask Working Note from Matthew Rocklin)

RAPIDS Blogs on Finance:

RAPID Fractional Differencing to Minimize Memory Loss While Making a Time Series Stationary (Medium Blog from Ritchie Ng, 6-minute read)

Fast Fractional Differencing on GPUs using Numba and RAPIDS (Part 1) (Medium Blog from Yi Dong, 4-minute read)

A small sample of popular RAPIDS Blogs (more at RAPIDS Medium):

Accelerating TSNE with GPUs: From hours to seconds (Medium Blog from Daniel Han-Chen, 10-minute read)

Accelerating Random Forests up to 45x using cuML (Medium Blog from Vishal Mehta, 11-minute read)

GPU-Accelerated Signal Processing with cuSignal (Medium Blog from Adam Thompson, 8-minute read)

Cyber Log Accelerators (CLX) (Medium Blog from Bartley Richardson, 8-minute read)

RAPIDS cuGraph: multi-GPU PageRank (Medium Blog from Alex Fender, 7-minute read)

RAPIDS 0.10: Things We Love Inspired By Decades of Data Science (Medium Blog from Josh Patterson, 11-minute read)

Combining Speed & Scale to Accelerate K-Means in RAPIDS cuML (Medium Blog from Corey Nolet, 7-minute read)

Accelerating Deep Learning Recommender Systems by Over 15x Using RAPIDS, PyTorch and fast.ai (Medium Blog from Even Oldridge, 11-minute read)

cuSpatial Accelerates Geospatial and Spatiotemporal Processing (Medium Blog from Naphade, 8-minute read)

RAPIDS Forest Inference Library: Prediction at 100 million rows per second (Medium Blog from Andrey Adinets and John Zedlewski, 13-minute read)