Model Overview

Last Updated At: 2025-10-21 09:10:00

A CDH is a physical server equipped with a virtualized environment. Each model corresponds to different physical server configurations, including the physical CPU model, number of CPU cores, memory size, local disk type, disk size, and other hardware resource configuration information. You can choose the appropriate CDH model that suits your business characteristics and scale.
At present, CDH models include:

CDH Type Standard HS20 Standard HS10 High IO HI20 High IO HI10 Memory-Optimized HM20 Compute HC20
Physical CPU Model Intel Xeon E5-2680 Broadwell(v4) Intel Xeon CPU Intel Xeon E5-2680 Broadwell(v4) Intel Xeon CPU Intel Xeon E5-2680 Broadwell(v4) Intel Xeon® E5-2667v4
Number of CPU Logical Cores 56 48 56 48 56 32
Memory Size (GB) 224 96 224 220 480 96
Local Disk Type Local disk Local disk Local SSD Local SSD Local disk Local SSD
Local Disk Size (GB) 2452 2452 7052 7104 2452 1000

Standard Family

The standard type is a balance of computing, memory, and network resources, which can meet the application resource requirements in most scenarios.

Standard HS20

The Standard HS20 is the latest generation of standard CDH instances, utilizing the Intel® Xeon® Broadwell processor. It offers a 40% improvement in integer and floating-point operation performance compared to previous generations. Paired with DDR4 memory, it achieves a 30% performance increase. This CDH type provides balanced computing, memory, and network resources, making it an excellent choice for many applications.

Model Characteristics

  • Intel Xeon E5-2680 Broadwell (v4) processor with a base clock rate of 2.4 GHz, DDR4 memory;
  • CPU performance is 20% higher than Series 1 Standard HS10;
  • Balance of computing, memory, and network resources.

Standard HS10

The Standard HS10 meets users' needs for flexible configuration options, featuring moderate pricing and configurable flexibility.

Model Characteristics

  • Covering low-core to high-core, users can flexibly choose configurations.
  • Intel Xeon CPU, with DDR3 memory.
  • Balance of computing, memory, and network resources.

High IO Family

The High IO I2 instances are optimized to provide applications with tens of thousands of low-delay random I/O operations per second (IOPS), making them the best choice for high IOPS use cases.

High IO HI20

The High IO HI20 instance is optimized to provide applications with tens of thousands of low-delay random I/O operations (OLTP) per second, making it the best choice for high-IOPS application scenarios. The High IO HI20 instance uses an Intel Broadwell processor and is equipped with DDR4 memory.

Model Characteristics

  • 2.4 GHz Intel Xeon E5-2680 Broadwell (v4) processor, DDR4 memory.
  • Using SSD instance storage, system disks are all local SSD disks, providing high IOPS and high throughput performance to meet the needs of users with high disk read and write requirements.
  • Low delay: In typical scenarios, it provides sub-millisecond access delay (blocksize = 4 k, iodepth = 1).

Applicable Scenario

  • High-performance databases, NoSQL databases (such as MongoDB), and clustered databases.
  • I/O-intensive applications that require low delay, such as online transaction processing (OLTP) systems and Elastic Search.

High IO HI10

The High IO HI10 is equipped with high-performance local SSD, which can meet the needs of users with high requirements for disk reading, writing and latency.

Model Characteristics

  • Intel Xeon CPU, covering low-core to high-core, users can flexibly choose the configurations.
  • Equipped with high-performance local SSD, it can meet the needs of users with high requirements for disk reading, writing and latency.
  • Low delay, providing sub-millisecond access delay.

Applicable Scenario

  • High-performance databases, NoSQL databases (such as MongoDB), and clustered databases.
  • I/O-intensive applications that require low latency, such as online transaction processing (OLTP) systems and Elastic Search.

Memory-Optimized Family

The Memory-Optimized family, characterized by its large memory capacity, is suitable for high-performance databases and distributed memory caching, and other applications that require extensive memory operations, lookups, and calculations.

Memory-Optimized HM20

The Memory-Optimized HM20 is designed to deliver fast performance for workloads that process large data sets in memory. Its characteristic is large memory, which makes it the best choice for high-memory computing applications.

Model Characteristics

  • 2.4 GHz Intel Xeon® E5-2680v4 processor, DDR4 memory.
  • The processor to memory ratio is 1:8.

Applicable Scenario

  • High-performance databases, distributed memory caching, and other applications that require extensive memory operations, lookups, and calculations.
  • Gene computation and other scenarios with self-built Hadoop clusters.

Compute Family

The Compute family is equipped with a base frequency of 3.2 GHz and has the highest single-core computing performance. Suitable for computationally intensive applications such as batch processing, high-performance computing, and large-scale game servers.

Compute HC20

The Compute HC20, with its high-performance processor and excellent cost-performance ratio, is an ideal choice for applications that require high computational performance and high concurrent read/write capabilities, among other computationally intensive applications.

Model Characteristics

3.2GHz Intel Xeon® E5-2667v4 processor, turbo frequency up to 3.6 GHz, DDR4 memory.

Applicable Scenario

The Compute C2 type is highly suitable for the following scenarios:

  • Processing workloads in batch.
  • High-traffic web servers, massively multiplayer online (MMO) game servers.
  • High Performance Computing (HPC) and other compute-intensive applications.
    For best performance, we recommend that you use current generation instance types when you create new instances.