Cloud Computing 12 min read

Infrastructure Convergence: Hardware Fusion and Hyper‑Converged Systems Overview

The article explains the evolution of enterprise IT infrastructure toward both custom, small‑scale distributed designs driven by cloud computing and integrated fusion/hyper‑converged architectures, detailing their design principles, differences, major vendor solutions, and the role of software‑defined storage.

Architects' Tech Alliance
Architects' Tech Alliance
Architects' Tech Alliance
Infrastructure Convergence: Hardware Fusion and Hyper‑Converged Systems Overview

In traditional enterprise IT architecture, the evolution from mainframes to minicomputers and x86 servers has moved from centralized to distributed models, and recent custom cabinets and converged systems have created two divergent trends: cloud‑driven, customized, small‑scale distributed infrastructure (e.g., Facebook, Google) and integrated fusion/hyper‑converged architectures for servers, storage, and networking.

Infrastructure Convergence

Although hardware redesign is led by internet companies, traditional IT also offers fusion and hyper‑converged solutions, which differ in several key aspects.

General design philosophy: Traditional IT fusion architectures distribute servers, storage, and networking evenly to support diverse application scenarios.

Scalability: Fusion architectures have limited scaling, typically up to around one hundred nodes.

Customization: Traditional fusion offers little hardware customization or extra functionality.

Virtualization platform: Fusion, especially hyper‑converged (HCA), is built on virtualization platforms, combining software‑defined technologies with hardware redesign.

Difference Between Hardware Fusion and Hyper‑Convergence

Hardware Fusion essentially stacks different devices to improve compatibility and efficiency, integrating existing products such as SAN arrays, switches, and compute resources. Examples include VCE’s Vblock and NetApp FlexPod, which combine components from EMC, Cisco, and VMware.

Hyper‑Convergence merges storage software with commercial hardware, typically using virtual storage appliances. It adopts a VM‑centric approach that integrates storage management, data protection, and networking into a single box, eliminating the need for external SANs. Hyper‑converged solutions can be purchased as pure software and deployed on certified commercial (e.g., x86) hardware platforms, allowing separate procurement of software and hardware. Notable vendors include VMware, Maxta, Nutanix, SimpliVity, Scale Computing, and VMware’s EVO:RAIL.

Mainstream Hyper‑Converged Systems

Since VMware released the EVO:RAIL reference architecture, Dell signed an OEM agreement with Nutanix, and HCA has become more visible. This architecture enables almost any server or storage vendor to participate in the stack and develop their own products. Below are some leading hyper‑converged systems.

EVO:RAIL:

VMware’s EVO:RAIL is a reference architecture based on proven VMware vSphere technologies and related suites, using VMware Virtual SAN for storage and providing a generic commercial server reference model.

Numerous partners, including EMC, HP, HDS, and others, have joined the ecosystem.

EMC VSPEX BLUE Hyper‑Converged Infrastructure: Launched in February, it combines VMware‑EVO:RAIL with EMC software to deliver compute, storage, networking, and management capabilities.

HP’s VMware EVO:RAIL Hyper‑Converged Device: This product integrates HP ProLiant‑SL servers with the VMware suite.

Supermicro EVO:RAIL Device: Provides a complete HCA solution by combining compute, networking, and storage within a single 2U, 4‑node server.

HDS’s EVO:RAIL Devices: Includes the UCP 1000 hyper‑converged platform and the UCP 2000 fusion platform, both using new rack servers designed for small‑to‑medium, remote, or branch office environments.

SANsymphony‑V: DataCore transforms standard x86 servers into enterprise‑grade SAN or hyper‑converged infrastructure devices, offering a choice of processors, memory, disks, or flash, and can be sourced from Dell, Fujitsu, Cisco, Lenovo, etc.

Dell’s XC Series: Dell adds Nutanix’s software stack to PowerEdge servers. The first model, XC730xd, is based on the PowerEdge R730xd rack platform; the second, XC630, uses the 1U PowerEdge R630 platform.

HP’s StoreVirtual: Acquired from LeftHand, this virtualized storage solution provides advanced data services, disaster recovery, and heterogeneous interoperability across physical and virtual workloads, integrating ESXi, HP StoreVirtual VSA, and OneView. HP ConvergedSystem 200‑HC‑StoreVirtual runs the Helion cloud version.

Based on commercial ESXi servers, the VSA provides network RAID functionality, distributing data across multiple server nodes.

Nimboxx AU‑110x: Launched in April, this system integrates compute, storage, networking, and security into a single appliance. Its MeshOS environment provides high availability, allowing two nodes to sustain critical workloads continuously.

Nutanix NX‑3060‑G4: Nutanix offers selectable VMware vSphere, Microsoft Hyper‑V, or KVM hypervisors, with server nodes suitable for 2U rack placement.

Pivot3 Enterprise HCI: Based in Austin, Texas, Pivot3 builds its enterprise‑grade HCI using patented vSTAC‑OS software, delivering VMware‑Horizon‑based virtual desktops and enabling business continuity and disaster recovery through shared compute and storage pools.

Additional vendors include Sphere3D’s V3 hyper‑converged series (San Jose, CA), SimpliVity’s OmniStack/OmniCube products (Massachusetts), and Riverbed’s SteelFusion 4.0 (San Francisco).

Hyper‑Converged Systems and Software‑Defined Storage (SDS)

Most HCA solutions aggregate storage from cluster nodes via a hypervisor to create a shared storage pool accessible to all virtual machines. SDS is a core component of HCA, eliminating dedicated storage and networking hardware, thereby reducing storage costs while providing essential services such as snapshots and cloning. For small enterprises, HCA may meet all needs, but large data centers face limitations:

HCA solutions rarely support existing external storage, leading to data silos and insufficient inter‑system data sharing.

Compute, memory, and network resources are shared among all VM workloads; a sudden spike in one application can exhaust resources.

Scaling often requires adding physical servers, which can cause storage capacity waste and mismatched growth of compute, performance, and capacity.

Please search “ICT_Architect” or scan the QR code below to follow the public account for more content.

Today's Recommendation

Best flagship store offers women's clothing, bags, and shoes; scan the QR code or click the original link to shop.

cloud computingVirtualizationInfrastructurehyper-convergedsoftware-defined storage
Architects' Tech Alliance
Written by

Architects' Tech Alliance

Sharing project experiences, insights into cutting-edge architectures, focusing on cloud computing, microservices, big data, hyper-convergence, storage, data protection, artificial intelligence, industry practices and solutions.

0 followers
Reader feedback

How this landed with the community

login Sign in to like

Rate this article

Was this worth your time?

Sign in to rate
Discussion

0 Comments

Thoughtful readers leave field notes, pushback, and hard-won operational detail here.