Bare Metal Server (BMS)

Superb Performance

No performance loss from virtualization overhead; cloud storage and networks with high bandwidth and low latency

Efficient Deployment

A BMS can be provisioned within minutes. You can easily manage them throughout their lifecycle.

Secure and Reliable

Each BMS offers dedicated compute. The servers can be isolated for network security, protected by security services, and restored from disk backups.

Fast Integration

BMS can be quickly integrated into cloud solutions for faster cloud transformation.

Diverse Specifications for Your Choice

Diverse Specifications for Your Choice

选择产品
  • General-purpose BMS
  • Disk-intensive BMS

Why Huawei Cloud BMS?

More Services at Lower Costs

More Services at Lower Costs

  • BMS:

    You only pay for the BMSs. There is no need to purchase or install any software or hardware. This simplifies O&M and makes it easier to use BMS together with other cloud services.

  • Traditional physical servers:

    You invest heavily in servers, OS, equipment hosting, and complex O&M.

Fast Deployment and Easy O&M

Fast Deployment and Easy O&M

  • You can buy a BMS with specifications matched to your needs. After you place an order, the BMS can be provisioned within minutes.

  • A unified console, open APIs, and open SDKs make it easy to manage and maintain your own BMSs.

Enterprise-Class Data Security on Cloud

Enterprise-Class Data Security on Cloud

  • Anti-DDoS, Trojan scans, anti-brute-force cracking, and other cloud security services are used to control access to servers.

  • BMS provides dedicated compute and allows you to use virtual private networks and security groups for network isolation. Host security services on the cloud can protect your servers.

Reliable Storage on Cloud

Reliable Storage on Cloud

  • Elastic Volume Service (EVS) disks are used to persistently store data and quickly rebuild a faulty BMS. So, a BMS has higher availability and reliability than a traditional physical server.

  • You can attach and detach EVS disks to and from a BMS anytime online. The unique shared EVS disks can serve applications that run in a cluster better.

  • EVS disks can be used as both system and data disks. The disks can be backed up for quick service restoration.

Continuous Evolution with the Latest Cloud Technologies

Continuous Evolution with the Latest Cloud Technologies

  • QingTian architecture-based zero performance loss, 100% of your compute unleashed

  • Custom hardware specifications

  • Shared cloud storage to serve cluster-based applications better

  • Globally available resources for fast service expansion around the world

QingTian, an Industry-Leading Architecture

QingTian, an Industry-Leading Architecture

Terms & Conditions

Application Scenarios

Application Scenarios

AI
Advantages
Large-Scale Cluster

Up to 26,000 GPUs/NPUs per cluster

Over 10 EFLOPS aggregate compute

Non-Blocking RDMA Network

GPUs and NPUs communicate with each other through a non-blocking RDMA network (typical bandwidth: 100 Gbit/s, 200 Gbit/s, or 400 Gbit/s).

High-Performance Computing
Scenario

BMSs with the latest Intel CPUs plus 100 Gbit/s NICs provide you with high compute performance and low network latency.

Advantages
Low Latency

100 Gbit/s networks are provided for high-speed communications. The network latency is within microseconds.

High Performance

The latest Intel CPUs deliver extraordinary performance.

Scalability

Open APIs make it easy to integrate with the ecosystem.

Core Database
Scenario

BMS provides the performance and security required for running core databases. A variety of BMS flavors are available, and shared EVS disks allow you to scale up storage whenever needed.

Advantages
Cluster Deployment

Shared EVS disks allow concurrent reads and writes by multiple BMSs. This makes BMSs capable of running cluster-based applications.

Large Capacity

A BMS can run with multiple EVS disks attached, each one scalable to up to 32 TB, but you only pay for the storage you use.

High Reliability

Three copies of data ensure 99.9999999% data durability.

Big Data
Scenario

BMSs can be used to run typical services such as big data storage and analysis. BMS uses a decoupled storage and compute solution that combines local storage and Object Storage Service (OBS).

Advantages
Big Data-Specific

BMSs with local disks or Kunpeng processors are specific to big data services.

Cost-Effective

Compute and storage resources can be scaled separately.

Easy to Use

BMSs can be provisioned within minutes for faster service deployment and rollout.

Start an

Efficient and Secure

Cloud Journey
Try Now

More Services

More Services