Zesty introduces an automated GPU fractioning tool built on NVIDIA's Run:ai, allowing companies to maximize hardware utilization through dynamic memory and workload management.
Zesty is launching a new autonomous capability that leverages NVIDIA's Run:ai platform to dynamically adapt GPU memory allocation and workload placement. This product enhancement is designed to increase the number of workloads per GPU, directly improving the value proposition of Zesty's infrastructure management suite.
It continuously adapts GPU memory allocation and workload placement to enable more workloads to run on each individual GPU.
NVIDIA's open-source Run:ai platform serves as the underlying technological foundation for Zesty's new autonomous GPU management capabilities. This integration highlights the growing ecosystem adoption of NVIDIA's software stack, which enables third-party developers to optimize GPU utilization and efficiency.
It provides an open-source foundation that allows companies like Zesty to build advanced GPU memory and compute sharing capabilities.
Zesty has unveiled an autonomous GPU fractioning capability designed to optimize resource allocation and compute sharing. By leveraging NVIDIA's open-source Run:ai platform, the solution enables enterprises to run significantly more workloads on existing hardware, driving higher efficiency in cloud infrastructure.
It continuously adapts GPU memory allocation and workload placement in real-time to allow multiple workloads to run on a single GPU.
Zesty's new product is built upon the foundation of NVIDIA's open-source Run:ai platform, which provides the necessary framework for managing GPU resources.