VMware presented the new version 8 of its virtualization software vSphere and new management tools for the multicloud at its VMware Explore in-house exhibition. VMware CEO Raghu Raghuram described the automation and shifting of more and more workloads to the cloud as a major challenge for IT operators.

The most important innovation in vSphere 8 is the support of data processing units (DPUs), intelligent network cards that increase data throughput in the network, reduce latencies and relieve the CPU at the same time. VMware is thus implementing the Monterey project announced two years ago. In a test with the in-memory database Redis, VMware claims to have measured a 36 percent increase in throughput and a 27 percent reduction in latency through the use of DPUs. Dell presented a suitable HCI platform for the VMware Explore, HPE a server designed for vSphere 8.

VMware Tanzu ships Kubernetes Grid 2.0 with vSphere 8. It is designed to make it easier to set up and manage Kubernetes clusters. For ML workloads, vSphere 8 supports up to 32 Nvidia GPUs. The add-on service VMware Cloud Disaster Recovery is designed to protect mission-critical workloads from failures, while VMware vRealize Operations helps with planning and optimizing infrastructure capacity.

The multicloud under control

VMware Aria is designed to help tame the complexity of multicloud environments. At its core is the graph database Aria Graph, which stores the applications, resources, roles and accounts and their relationships in a multicloud environment. The tools Aria Guardrails (security and compliance), Aria Migration (migration of workloads in the multicloud) and Aria Business Insights (analysis) are based on this.

VMware promises high scalability for the Aria portfolio to hundreds of millions of nodes, the collection of all relevant events even at high change rates and the integration of data from tools from other providers. The information can be integrated into other observability and management tools via a GraphQL API.