CoreWeave

CoreWeave is a specialized cloud provider, offering a range of products such as GPU clouds, clusters, servers, and workstations designed to accelerate deep learning and AI processes. Initially founded in 2017 as an Ethereum mining venture, the company has transitioned into a general-purpose cloud computing platform, focusing on GPU computing for applications such as AI, machine learning, visual effects, rendering, batch processing, and pixel streaming. It delivers GPU-accelerated compute resources on demand, catering to various industries with high-performance computing solutions. It claims that its infrastructure, tailored for Kubernetes, is designed specifically for compute-demanding tasks such as machine learning, VFX rendering, pixel streaming, and batch processing, with up to 35x faster speeds and 80% less costly than traditional cloud providers.

CoreWeave delivers GPU-accelerated compute resources on demand, using NVIDIA GPUs like H100s, A100s, A40s, and RTX A6000s. CoreWeave's virtual servers are built on top of Kubernetes and offer a range of features tailored to enhance artist workflows and productivity. These virtual servers provide dedicated GPUs through PCI pass-through without shared resources. Users can choose from a wide range of NVIDIA GPUs, and both Linux and Windows environments are supported. The company offers blockchain and Web3 infrastructure services, providing data and digital identity solutions as a service. 

In January 2023, CoreWeave acquired Conductor Technologies to expand its offerings for VFX and animation studios to enable the easy integration of workloads to the cloud with ease. 

Key customers and partnerships

In December 2023, Coreweave partnered with Dell Technologies to use Dell PowerEdge XE9860 servers with NVIDIA GPUs as the key infrastructure powering its cloud solutions. In June 2023 , Coreweave collaborated with Microsoft to provide cloud computing infrastructure for AI. In November 2022, CoreWeave was announced as NVIDIA's first Elite Cloud Solutions Provider for Compute in the NVIDIA Partner Network (NPN) program.