CoreWeave partners with Google Cloud to enable cross-cloud AI workloads via CoreWeave Interconnect and LOTA Cross-Cloud
CoreWeave and Google Cloud announced a collaboration to enable customers to run AI workloads across both cloud infrastructures through CoreWeave Interconnect, a fiber-based network interconnect linking CoreWeave's cloud infrastructure to Google Cloud. The interconnect offering minimizes latency and improves reliability and security via built-in MACsec encryption, utilizing Google's Partner Cross-Cloud Interconnect to facilitate training and inference workloads across both AI infrastructures. CoreWeave also launched LOTA Cross-Cloud, its patented Local Object Transport Accelerator technology, which enables workloads to run across cloud environments regardless of data storage location by placing data directly onto GPU nodes. The LOTA Cross-Cloud solution offers near-local throughput of up to seven GBps per GPU and removes egress fees, currently available in limited availability. The CoreWeave Interconnect will soon be available in private preview, with plans to roll out the solution in select regions, addressing a key challenge in cross-cloud AI workload distribution.
Counterparts (2)
Register free to access full counterpart details, deal analysis, and timeline.
Register free →Timeline
Get the full picture — timeline, source intelligence, and counterpart analysis.
Register free →
Global Infrastructure Sherpa