Striveworks has announced a new partnership with Neural Magic, aiming to integrate the latter’s core offerings into its Chariot MLOps platform’s training and model services. The partnership will provide Chariot users with a no-code/low-code way to deploy machine learning models to CPUs while getting GPU-like speeds.
Striveworks Partners with Neural Magic
Striveworks, a pioneer in responsible MLOps for national security and other highly regulated spaces, has collaborated with Neural Magic to offer its customers a new way to exclusively use CPUs and achieve necessary inference speeds. The integration of Neural Magic’s SparseML libraries and DeepSparse Inference Runtime with Chariot MLOps platform will result in machine learning models that are smaller in size, equally as accurate, and significantly more performant than even expensive hardware-accelerated compute platforms.
Integration Offers Machine Learning on CPUs
Chariot is an end-to-end MLOps software platform that provides a factory floor where data scientists, analysts, Subject Matter Experts (SMEs), and others can all work, develop, train, deploy, monitor, retrain, and redeploy custom models and custom workflows on various datasets and data sources. GPUs are typically necessary to deploy models with operationally relevant inference times. However, this integration will provide Chariot users with the ability to deploy these models to CPUs, breaking down the barriers of specialized hardware and providing great performance at a lower cost.
Chariot MLOps Platform to Benefit from Partnership
This partnership between Striveworks and Neural Magic will be beneficial to Chariot MLOps Platform users, as it offers a cost-effective, efficient, and seamless experience on readily available CPUs. Customers will have flexible deployment options with Chariot, where they can access its machine learning capabilities wherever they need it. Furthermore, cloud-deployed customers will see model serving costs decrease as well, introducing a valuable economic benefit.
In conclusion, Striveworks’ partnership with Neural Magic promises to bring cutting-edge performance to machine learning models on the Chariot MLOps platform with cost-effective and efficient CPU deployment.