H2O.ai, the open source leader in AI and Machine Learning, announced that H2O Driverless AI, the award-winning automatic machine learning platform, and H2O4GPU, the open source GPU-accelerated machine learning package, are now optimized for the new NVIDIA-powered Data Science Workstations with NVIDIA Quadro RTX GPUs and NVIDIA CUDA-X AI acceleration libraries. H2O.ai also announced it will integrate NVIDIA RAPIDS to take advantage of its GPU-optimized machine learning algorithms.
When combined with H2O Driverless AI, NVIDIA-powered Data Science Workstations allow data scientists to work on extremely complex models and pipelines easily with highly optimized infrastructure that reduces time to results. Models and pipelines can be easily re-trained on the Data Science Workstations to support larger datasets or even include an H2O Sparkling Water cluster that runs on massive datasets.
“H2O.ai is excited for our partnership with NVIDIA towards accelerating machine learning for everyone. With H2O Driverless AI on NVIDIA-powered Data Science Workstations it is faster, cheaper and easier to build and deploy machine learning models. Recipes from our open source community and some of the world’s greatest data science grandmasters make solving a whole class of data problems fast, explainable and easy,” said Sri Ambati, CEO and founder of H2O.ai. “Our mission to democratize AI for the enterprise is now within reach with automatic machine learning from H2O.ai on GPUs by NVIDIA.”
NVIDIA RAPIDS builds on popular open-source projects — including Apache Arrow, Pandas and Scikit-learn. With the addition of GPU acceleration to the most popular Python data science toolchain, it becomes a natural choice to improve workflows inside Driverless AI by taking advantage of GPU-accelerated machine learning algorithms.
The combined offerings deliver the fastest and most accurate iterations of machine learning models. Driverless AI uses the following GPU-accelerated algorithms giving data scientists flexibility:
- LightGBM for a high performance and scalable implementation of Gradient Boosted Machines optimized for large datasets
- TensorFlow-based models and network for NLP use cases
- Truncated SVD (singular value decomposition) and PCA (principal component analysis) for dimensionality reduction and feature engineering
- Improved K-means and XGBoost performance on multi-GPU workloads for faster performance and ability to handle larger datasets
In addition, H2O.ai continues to democratize AI by making H2O Driverless AI and H2O4GPU available from the NVIDIA NGC software hub.
“The NVIDIA-powered Data Science Workstation combined with H2O Driverless AI and H2O4GPU provide a comprehensive solution for enterprises looking to transform their businesses with automatic machine learning,” said Bob Pette, Vice President of Professional Visualization at NVIDIA. “Our financial, healthcare and insurance customers are able to see real growth from the collaboration by gaining tremendous speed to insights and interpretability.”
Automating Data Science Workflows
H2O Driverless AI empowers data scientists to work on projects faster and more efficiently by using automation and state-of-the-art computing power to accomplish tasks that typically takes months in just minutes or hours. By delivering automatic feature engineering, model validation, model tuning, model selection and deployment, machine learning interpretability, time-series, NLP and automatic pipeline generation for model scoring, H2O Driverless AI provides companies with a data science platform that addresses the needs of a variety of use cases for every enterprise in every industry.