In a significant step forward in enabling the adoption of AI at scale, NVIDIA today announced their latest GPU-powered innovation for data science, the RAPIDS suite of open-source libraries.
RAPIDS will allow data scientists to move workflows for model training and iteration onto the GPU. Previously, enterprises had to make major investments in CPU infrastructure to run analytics and machine learning. Not only was it resource intensive, the data orchestration needed to make it happen was also very complex. That’s where the RAPIDS libraries come in.
Speed is of the essence for organizations looking to operationalize their investments in data science and gain a competitive edge. With RAPIDS, enterprises can leverage the power of the GPU across the model development toolchain to dramatically simplify and speed up the data science pipeline, reducing the time needed to train models from days to hours. Data scientists can optimize training with more iterations for better model accuracy and increase their productivity by accelerating data preparation and training.
We’re looking forward to working with NVIDIA to continue to make AI accessible to all organizations. With RAPIDS, you can train and iterate your model on NVIDIA GPUs to develop and optimize your models. Then you can bring your models to Kinetica to operationalize and put them into production in the real world and gain business value from AI. It’s an end-to-end GPU-accelerated strategy for model training, iteration, and deployment. Small improvements in model training and deployment can have a major impact on efficiency and the bottom line. We’re proud to partner with NVIDIA to help organizations make AI a reality.
You can download RAPIDS and the Kinetica engine from NVIDIA GPU Cloud (GPU).
Read the RAPIDS press release to learn more.
To learn more about how Kinetica’s GPU-accelerated engine can help you to apply AI in the enterprise visit our Solutions page.
Ready to augment BI with AI? Download our white paper.
Ken Wattana is Sr. Partner Marketing Manager at Kinetica. You can follow him on Twitter @KenWattana.