With Run:AI’s “Run:it your way”, data science teams can use Kubeflow, MLflow, a new ResearcherUI or other tools to manage GPU allocation directly inside their workflows
September 2, 2021, Tel Aviv, Israel. Run:AI, the leading compute management platform for orchestration and acceleration of AI, announced the launch of a new ResearcherUI, as well as integration with machine learning tools including Kubeflow, MLflow and Apache Airflow. The new UI option is a part of Run:AI’s “Run:it your way” initiative, enabling data scientists to choose their preferred ML tools that manage modeling and other data science processes on top of Run:AI’s compute orchestration platform.
“Some data scientists like Kubeflow; some prefer MLFlow; some would rather use YAML files. We even heard of a Fortune 500 company that uses 50 different data science tools. With Run:AI, there’s no need to force all data science teams to use a specific ML tool in order to take advantage of the Run:AI GPU orchestration platform,” said Omri Geller, CEO of Run:AI. Instead, each team can “Run:it their way”, sharing pooled, dynamic GPU resources while using the best ML tools to match the company’s data science workflow.”
There are dozens of data science tools used to run experiments, and naturally some data scientists are more comfortable with one tool or another. Run:AI dynamically allocates GPU to data science jobs across a whole organization, regardless of the ML tools they use to build and manage models. Teams can have guaranteed quotas, but their workloads can use any available idle GPU resources, creating logical fractions of GPUs, stretching jobs across multiple GPUs and multiple GPU nodes for distributed training, and maximizing hardware value for money.
With “Run:it your way”, Run:AI supports all popular machine learning platforms including, but not limited to, Kubeflow, Apache Airflow, MLflow, API support (including for air-gapped data science environments), YAML, Command Line, and Run:AI’s new ResearcherUI.
Run:AI is a cloud-native compute management platform for the AI era. Run:AI gives data scientists access to all of the pooled compute power they need to accelerate AI development and deployment – whether on-premises or in the cloud. The platform provides IT and MLOps with real-time visibility and control over scheduling and dynamic provisioning of GPUs to deliver more than 2X gains in utilization of existing infrastructure. Built on Kubernetes, Run:AI enables seamless integration with existing IT and data science workflows. Learn more at www.run.ai.