Open & Lightweight
Enterprise Cloud Platform
Full-stack HCI with 11M+ IOPS, native virtualization, distributed storage, SDN, and AI-ready infrastructure.
4K random read • Sequential throughput • Sub-100μs latency
11 consecutive quarters — IDC
SmartX Enterprise Cloud Platform
Founded in 2014, SmartX delivers open, lightweight, full-stack Hyperconverged Infrastructure (HCI) that empowers enterprises with full control, high performance, and simplicity — a true VMware alternative.
SmartX ECP combines native virtualization (ELF), distributed block storage (ZBS), software-defined networking (Everoute), Kubernetes services, data protection, and centralized management (CloudTower) into a single platform. Complete installation within 2 hours, resources online within 1 day, and zero service disruption during scaling and upgrades.
Full-Stack Components
- 01SMTX OS
Native ELF hypervisor + ZBS distributed storage (VMware ESXi compatible) - 02CloudTower
Centralized management, intelligent O&M, and multi-cluster control - 03Everoute
SDN, micro-segmentation, load balancing, VPC, and zero-trust networking - 04Backup, DR & Kubernetes
Sync/async replication, active-active DR (RPO=0), and native container platform
Surpassing high-end all-flash arrays
Deep architectural evolution — Intel DSA hardware acceleration, IO_uring async framework, RDMA multi-link aggregation, and multi-storage-instance parallelism deliver multiple-fold improvements over v6.2.
Storage I/O
High Availability
Complete HA support for mission-critical workloads with zero-downtime guarantees.
- HA for RDMA storage network (cross-NIC bonding)
- HA for VMs using SR-IOV NIC
- HA for vGPU workloads (AI & VDI)
- NIC Teaming for high-speed network resilience
Native synchronous replication & compliance
First in the domestic HCI market to deliver synchronous replication (RPO=0) — reducing reliance on costly external storage while providing real-time data protection.
Synchronous Replication
VM-level synchronous replication with RPO=0. No need to protect entire clusters — auto-tolerates DR network fluctuations with failover and DR drill support.
- RPO=0 for critical workloads
- VM-granular protection
- CloudTower HA failover
Security & Compliance
Native security architecture with agentless antivirus, national cryptographic algorithm encryption, and built-in KMS.
- AES-256 data-at-rest encryption
- Agentless antivirus
- Live migration encryption
Active-Active DR
Cross-site stretched clusters with minimum 4-node deployment, placement group policies, and automatic management-plane failover.
- Async + sync replication
- 4-node minimum (down from 6)
- ERSPAN network auditing
Essential vs. Standard vs. Advanced
All editions run on the same SMTX OS foundation. Essential provides core HCI, Standard adds Kubernetes and full DR, and Advanced unlocks enterprise SDN, VPC, and active-active stretched clusters.
| Feature | Essential HCI Core | Standard Standard HCI | Advanced Full Enterprise Cloud |
|---|---|---|---|
| Virtualization (SMTX OS - ELF) | |||
| Hypervisor support (ELF + VMware ESXi) | |||
| HA, DRS & availability domains | |||
| vGPU and GPU pass-through | |||
| VM pinning in performance layer | |||
| VM-centric storage QoS | |||
| Max hosts per cluster | 255 | 255 | 255 |
| Distributed Storage (ZBS) | |||
| Block storage with erasure coding | |||
| Max raw capacity per node | 256 TB | 256 TB | 256 TB |
| Storage protocol support (NFS, iSCSI) | |||
| File storage for external hosts | — | ||
| Data-at-rest encryption (AES-256 + KMS) | |||
| Data Protection & Disaster Recovery | |||
| Agentless VM backup & recovery | |||
| Async & sync replication | Add-on | ||
| Active-active stretched cluster | — | Add-on | |
| Kubernetes & Containers (SKS) | |||
| SMTX Kubernetes Service (SKS) | — | ||
| Kubernetes on bare metal | — | — | |
| Pod-VM integrated networking | — | ||
| Networking & Security (Everoute) | |||
| Distributed firewall & micro-segmentation | — | Add-on | |
| Load balancing | — | Add-on | |
| Virtual Private Cloud (VPC) | — | — | Add-on |
| Network observability | — | ||
| Management & Analytics (CloudTower) | |||
| Single console & multi-cluster management | |||
| Health monitoring & alerting | |||
| CloudTower HA (management-plane failover) | — | — | |
ECP Essential
Core HCI with full virtualization (ELF), distributed block storage with erasure coding, encryption at rest, agentless VM backup, and CloudTower management. Replication is available as an add-on for organizations that need DR capabilities.
ECP Standard
Everything in Essential plus: SMTX Kubernetes Service (SKS) for VM-container convergence, file storage for external hosts, built-in async & sync replication, Pod-VM integrated networking, and network observability. Firewall and load balancing available as add-ons.
ECP Advanced
Full enterprise cloud infrastructure. Everything in Standard plus: built-in Everoute SDN with micro-segmentation and load balancing, active-active stretched clusters, Kubernetes on bare metal, CloudTower HA, and VPC networking available as add-on.
Full specifications: SmartX ECP Edition Comparison
SMTX AI Platform
Enterprise Edition — secure on-premise generative AI infrastructure built natively on ECP. Unified model resource delivery for MaaS (Model-as-a-Service) production environments.
Comprehensive Model Management
- Native support for text generation, embedding, and reranking models
- Hugging Face import + custom/proprietary model upload
- Model catalog with pre-defined inference engines, resource specs, and runtime configs
Unified Compute Resource Pool
- Heterogeneous GPU pooling across NVIDIA and AMD, on bare metal, VMs, and K8s
- GPU sharing with intelligent partitioning and isolation for higher utilization
High-Performance Inference
- KVCache-aware multi-replica scheduling for optimized request routing
- High-concurrency inference with intelligent load balancing and cache hit optimization
Multi-Tenant Governance
- Tenant-level isolation with separate resource and model spaces
- Fine-grained RBAC across model management, inference, and scheduling
- API key management with token quota, usage statistics, and billing readiness
Full-Stack Monitoring
End-to-end observability from low-level GPU status to high-level inference services — API token usage, request throughput, and model performance in one dashboard.
Deployment Options
Deploy directly on SmartX ECP with shared virtualization and storage
Dedicated GPU infrastructure for maximum inference performance
Mix ECP and bare-metal resources in a unified compute pool
Ecosystem Partners
Joint solution validations covering the full AI workflow — from model training and fine-tuning to application orchestration.
Technical Specifications
Scalability & Limits
Storage & Data Services
Networking & Security
Market leadership in China
#1 HCI Software for 11 consecutive quarters (Q1 2023 - Q3 2025) according to IDC, with recognition by Gartner three times.
Why SmartX for Hong Kong Enterprises
Ready to deploy SmartX in Hong Kong?
Contact Zens Distribution — your official SmartX partner for enterprise cloud, HCI, and AI infrastructure.
Request Demo / QuotePerformance data based on SmartX internal tests with specific hardware configurations (Intel, Hygon, Kunpeng platforms, 3-node clusters, 100Gb/200Gb storage networks, SMTX OS 6.3). Actual results may vary. Visit smartx.com for complete details.