Skip to main content

February 21, 2025

What's new at CoreWeave?

CKS API

Deploy and manage CKS resources programmatically

CoreWeave is excited to announce the General Availability launch of the CoreWeave Kubernetes Service (CKS) API, a new way to programmatically deploy, manage, and scale HPC applications using Kubernetes on CoreWeave's high-performance infrastructure.

With the CKS API, CKS clusters and VPCs can be deployed and managed programmatically.

Learn more

Explore the API!

REST API reference

Protobuf source reference

Or, learn more about CKS.

CoreWeave Terraform provider

Manage CoreWeave resources as code

With the new CoreWeave Terraform provider, customers can deploy and manage VPCs and CKS clusters as code.

Welcome to the new Cloud Console!

Enhanced Console design and user experience

Create clusters in just a few clicks

The refreshed CoreWeave Cloud Console simplifies the process of viewing, creating, and managing CKS resources from your browser. Improved usability and creation flows mean faster times to get high performance clusters up and running, and a better experience managing deployed resources.

The new enhancements to the Console includes a revised cluster creation flow that enables customers to get a new standard CKS cluster up and running in just four clicks.

Node Pools

Enhanced control over compute instances with Node Pools

Node Pools now allow for easier management of multiple compute instances by bundling them as a single entity, allowing for better top-level control of compute resources.

Note

Node Pools are managed as Kubernetes resources, and can be deployed using the Kubernetes API or the Cloud Console.

Enhanced storage for HPC workloads

GPU-Local storage and Distributed File Storage

CKS reframes some of our storage offerings to better expose them to customers.

Distributed File Storage delivers high write speeds, high IOPS for datapath-related operations, as well as higher read speeds in cases where block size and queue depth are low. Distributed File Storage is ideal for operations that require a lot of synchronization between Pods, and performance bottlenecks caused by contention are minimized.

CoreWeave instances are equipped with SSD or NVMe-backed GPU-Local Storage for non-persistent local storage, scratch space, caching, and logs.