A typical converged infrastructure system found in most virtualization products today

The primary difference between a converged (CI) and a hyperconverged infrastructure (HCI) is that the former is hardware dependent on its building blocks. Whereas a full hyperconverged infrastructure system is software defined.

By design, HCI is inherently more scalable, flexible and maneuverable than CI. HCI empowers open concepts and is usually deployed with commodity server hardware and it provides simplified scale out architecture with simple components.

A full hyperconverged design. Hard drive units are located at the nodes & not in a separate JBOD or SAN hardware

Another main distinction between CI and HCI involves the rack system. A CI system normally consist of large, rack-scale platform that merges compute, storage, and networking into a turnkey product. While HCI consists of a 1U or 2U (rack-unit) systems that consolidate one or more multi-core servers with a local storage array direct in its nodes without incorporation of large and expensive SAN hardware.

The architecture varies between a CI and a HCI system. The converged architecture storage is attached directly to the physical server and the HCI architecture has a storage controller function that runs as a service on every node in the cluster. HCI shares storage to all compute and virtual machines (VMs), whereas CI does not.

Within a fully hyperconverged infrastructure system, hardware management is done with causing downtime to operating virtual machines or containers

With a HCI platform, engineers working at the data center to repair, maintain or service server hardware, will not cause scheduled downtime to running services at the front end. Because the HCI nodes itself acts like a giant pool of storage similar to the of storage pools in RAID controllers, nodes and hard drives can be hot-plugged removed and reinserted, while virtual machines and containers remain operational.