This topic describes available ECS instance families and their features, specifications, and application scenarios. An ECS instance is the smallest unit that can provide compute capabilities and services for your business.

ECS instances are categorized into different instance families based on the business scenarios to which they can be applied. Each instance family is divided into different instance types based on their CPU and memory specifications. ECS instance type defines the basic properties of an ECS instance: CPU (including CPU model and clock speed) and memory. In addition to the instance type, you must also configure the Block Storage devices, image, and network type when you create an ECS instance.
Note The available instance families and types vary from region to region. You can go to the ECS Instance Types Available for Each Region page to view the instance types available in each region.

Enterprise scenarios have high requirements for business stability. Alibaba Cloud ECS instance families are divided into enterprise-level and shared instance families based on whether the instance families are suitable for enterprise scenarios. Enterprise-level instance families offer consistent performance and dedicated resources. Each vCPU in enterprise-level instance families corresponds to an Intel ® Xeon ® core hyperthread. For the differences between enterprise-level and shared instance families, see FAQs about ECS instances.

You can upgrade or downgrade instances within the same instance family or across certain instance families. For more information, see Instance families that support instance type changes.

Alibaba Cloud ECS instances are categorized into the following instance families:

g6, general purpose instance family

Features
  • Provides predictable and stable high performance and can reduce virtualization overheads with the use of the X-Dragon architecture
  • I/O optimized
  • Supports enhanced SSDs, standard SSDs, and ultra disks
    Note The maximum performance of cloud disks varies depending on instance families. A single g6 instance can deliver an IOPS of up to 200,000. For higher storage I/O performance, we recommend that you use g5se. For more information, see g5se, storage optimized instance family with enhanced performance.
  • Provides strong storage I/O performance with a large computing capacity
  • CPU-to-memory ratio of 1:4
  • Supports enabling or disabling Hyper-Threading
    Note By default, Hyper-Threading is enabled on ECS instances. For more information, see Customize CPU options.
  • Ultra high packet forwarding rate
    Note The maximum network performance varies depending on instance families. For higher packet forwarding rate, we recommend that you use g5ne. For more information, seeg5ne, network enhanced instance family.
  • Equipped with 2.5 GHz Intel ® Xeon ® Platinum 8269CY (Cascade Lake) processors, up to 3.2 GHz Turbo Boost
  • Provides strong network performance with a large computing capacity
  • Supports instance type changes to c6 or r6
  • Suitable for the following scenarios:
    • Scenarios where large volumes of packets are received and transmitted, such as on-screen video comments and telecom data forwarding
    • Enterprise-grade applications of various types and sizes
    • Websites and application servers
    • Game servers
    • Small and medium-sized database systems, caches, and search clusters
    • Data analysis and computing
    • Computing clusters and memory-intensive data processing
Instance types
Instance type vCPUs Memory (GiB) Local storage (GiB) Basic bandwidth (Gbit/s) Burstable bandwidth (Gbit/s) Packet forwarding rate (Kpps) IPv6 support NIC queues ENIs (including one primary ENI) Private IP addresses per ENI Disk IOPS (K) Disk bandwidth (Gbit/s)
ecs.g6.large 2 8.0 None 1.0 3.0 300 Yes 2 2 6 10 1
ecs.g6.xlarge 4 16.0 None 1.5 5.0 500 Yes 4 3 10 20 1.5
ecs.g6.2xlarge 8 32.0 None 2.5 8.0 800 Yes 8 4 10 25 2
ecs.g6.3xlarge 12 48.0 None 4.0 10.0 900 Yes 8 6 10 30 2.5
ecs.g6.4xlarge 16 64.0 None 5.0 10.0 1,000 Yes 8 8 20 40 3
ecs.g6.6xlarge 24 96.0 None 7.5 10.0 1,500 Yes 12 8 20 50 4
ecs.g6.8xlarge 32 128.0 None 10.0 None 2,000 Yes 16 8 20 60 5
ecs.g6.13xlarge 52 192.0 None 12.5 None 3,000 Yes 32 7 20 100 8
ecs.g6.26xlarge 104 384.0 None 25.0 None 6,000 Yes 32 15 20 200 16
Note

g5, general purpose instance family

Features
  • I/O optimized
  • Supports enhanced SSDs, standard SSDs, and ultra disks
    Note The maximum performance of cloud disks varies depending on instance families. A single g5 instance can deliver an IOPS of up to 200,000. For higher storage I/O performance, we recommend that you use g5se. For more information, see g5se, storage optimized instance family with enhanced performance.
  • CPU-to-memory ratio of 1:4
  • Ultra high packet forwarding rate
    Note The maximum network performance varies depending on instance families. For higher packet forwarding rate, we recommend that you use g5ne. For more information, seeg5ne, network enhanced instance family.
  • Equipped with 2.5 GHz Intel ® Xeon ® Platinum 8163 (Skylake) or 8269CY (Cascade Lake) processors
  • Provides strong network performance with a large computing capacity
  • Suitable for the following scenarios:
    • Scenarios where large volumes of packets are received and transmitted, such as on-screen video comments and telecom data forwarding
    • Enterprise-grade applications of various types and sizes
    • Small and medium-sized database systems, caches, and search clusters
    • Data analysis and computing
    • Computing clusters and memory-intensive data processing
Instance types
Instance type vCPUs Memory (GiB) Local storage (GiB) Bandwidth (Gbit/s) Packet forwarding rate (Kpps) IPv6 support NIC queues ENIs (including one primary ENI) Private IP addresses per ENI
ecs.g5.large 2 8.0 None 1.0 300 Yes 2 2 6
ecs.g5.xlarge 4 16.0 None 1.5 500 Yes 2 3 10
ecs.g5.2xlarge 8 32.0 None 2.5 800 Yes 2 4 10
ecs.g5.3xlarge 12 48.0 None 4.0 900 Yes 4 6 10
ecs.g5.4xlarge 16 64.0 None 5.0 1,000 Yes 4 8 20
ecs.g5.6xlarge 24 96.0 None 7.5 1,500 Yes 6 8 20
ecs.g5.8xlarge 32 128.0 None 10.0 2,000 Yes 8 8 20
ecs.g5.16xlarge 64 256.0 None 20.0 4,000 Yes 16 8 20
Note

ic5, compute intensive instance family

Features
  • I/O optimized
  • Supports enhanced SSDs, standard SSDs, and ultra disks
  • CPU-to-memory ratio of 1:1
  • Ultra high packet forwarding rate
  • Equipped with 2.5 GHz Intel ® Xeon ® Platinum 8163 (Skylake) or 8269CY (Cascade Lake) processors
  • Provides strong network performance with a large computing capacity
  • Suitable for the following scenarios:
    • Web frontend servers
    • Data analysis, batch compute, and video encoding
    • Scenarios where large volumes of packets are received and transmitted, such as on-screen video comments and telecom data forwarding
    • Frontend servers of massive multiplayer online (MMO) games
Instance types
Instance type vCPUs Memory (GiB) Local storage (GiB) Bandwidth (Gbit/s) Packet forwarding rate (Kpps) IPv6 support NIC queues ENIs (including one primary ENI) Private IP addresses per ENI
ecs.ic5.large 2 2.0 None 1.0 300 No 2 2 6
ecs.ic5.xlarge 4 4.0 None 1.5 500 No 2 3 10
ecs.ic5.2xlarge 8 8.0 None 2.5 800 No 2 4 10
ecs.ic5.3xlarge 12 12.0 None 4.0 900 No 4 6 10
ecs.ic5.4xlarge 16 16.0 None 5.0 1,000 No 4 8 20
Note

c6, compute optimized instance family

Features
  • Provides predictable and stable high performance and can reduce virtualization overheads with the use of the X-Dragon architecture
  • I/O optimized
  • Supports enhanced SSDs, standard SSDs, and ultra disks
    Note The maximum performance of cloud disks varies depending on instance families. A single c6 instance can deliver an IOPS of up to 200,000. For higher storage I/O performance, we recommend that you use g5se. For more information, see g5se, storage optimized instance family with enhanced performance.
  • Provides strong storage I/O performance with a large computing capacity
  • Supports enabling or disabling Hyper-Threading
    Note By default, Hyper-Threading is enabled on ECS instances. For more information, see Customize CPU options.
  • CPU-to-memory ratio of 1:2
  • Ultra high packet forwarding rate
  • Equipped with 2.5 GHz Intel ® Xeon ® Platinum 8269CY (Cascade Lake) processors, up to 3.2 GHz Turbo Boost
  • Provides strong network performance with a large computing capacity
  • Supports instance type changes to g6 or r6
  • Suitable for the following scenarios:
    • Scenarios where large volumes of packets are received and transmitted, such as on-screen video comments and telecom data forwarding
    • Web frontend servers
    • Frontend servers of massive multiplayer online (MMO) games
    • Data analysis, batch compute, and video encoding
    • High-performance science and engineering applications
Instance types
Instance type vCPUs Memory (GiB) Local storage (GiB) Basic bandwidth (Gbit/s) Burstable bandwidth (Gbit/s) Packet forwarding rate (Kpps) IPv6 support NIC queues ENIs (including one primary ENI) Private IP addresses per ENI Disk IOPS (K) Disk bandwidth (Gbit/s)
ecs.c6.large 2 4.0 None 1.0 3.0 300 Yes 2 2 6 10 1
ecs.c6.xlarge 4 8.0 None 1.5 5.0 500 Yes 4 3 10 20 1.5
ecs.c6.2xlarge 8 16.0 None 2.5 8.0 800 Yes 8 4 10 25 2
ecs.c6.3xlarge 12 24.0 None 4.0 10.0 900 Yes 8 6 10 30 2.5
ecs.c6.4xlarge 16 32.0 None 5.0 10.0 1,000 Yes 8 8 20 40 3
ecs.c6.6xlarge 24 48.0 None 7.5 10.0 1,500 Yes 12 8 20 50 4
ecs.c6.8xlarge 32 64.0 None 10.0 None 2,000 Yes 16 8 20 60 5
ecs.c6.13xlarge 52 96 None 12.5 None 3,000 Yes 32 7 20 100 8
ecs.c6.26xlarge 104 192.0 None 25.0 None 6,000 Yes 32 15 20 200 16
Note

c5, compute optimized instance family

Features
  • I/O optimized
  • Supports enhanced SSDs, standard SSDs, and ultra disks
    Note The maximum performance of cloud disks varies depending on instance families. A single c5 instance can deliver an IOPS of up to 200,000. For higher storage I/O performance, we recommend that you use g5se. For more information, see g5se, storage optimized instance family with enhanced performance.
  • CPU-to-memory ratio of 1:2
  • Ultra high packet forwarding rate
  • Equipped with 2.5 GHz Intel ® Xeon ® Platinum 8163 (Skylake) or 8269CY (Cascade Lake) processors
  • Provides strong network performance with a large computing capacity
  • Suitable for the following scenarios:
    • Scenarios where large volumes of packets are received and transmitted, such as on-screen video comments and telecom data forwarding
    • Web frontend servers
    • Frontend servers of massive multiplayer online (MMO) games
    • Data analysis, batch compute, and video encoding
    • High-performance science and engineering applications
Instance types
Instance type vCPUs Memory (GiB) Local storage (GiB) Bandwidth (Gbit/s) Packet forwarding rate (Kpps) IPv6 support NIC queues ENIs (including one primary ENI) Private IP addresses per ENI
ecs.c5.large 2 4.0 None 1.0 300 Yes 2 2 6
ecs.c5.xlarge 4 8.0 None 1.5 500 Yes 2 3 10
ecs.c5.2xlarge 8 16.0 None 2.5 800 Yes 2 4 10
ecs.c5.3xlarge 12 24.0 None 4.0 900 Yes 4 6 10
ecs.c5.4xlarge 16 32.0 None 5.0 1,000 Yes 4 8 20
ecs.c5.6xlarge 24 48.0 None 7.5 1,500 Yes 6 8 20
ecs.c5.8xlarge 32 64.0 None 10.0 2,000 Yes 8 8 20
ecs.c5.16xlarge 64 128.0 None 20.0 4,000 Yes 16 8 20
Note

r6, memory optimized instance family

Features
  • Provides predictable and stable high performance and can reduce virtualization overheads with the use of the X-Dragon architecture
  • I/O optimized
  • Supports enhanced SSDs (ESSDs), standard SSDs, and ultra disks
    Note The maximum performance of cloud disks varies depending on instance families. A single r6 instance can deliver an IOPS of up to 200,000. For higher storage I/O performance, we recommend that you use g5se. For more information, see g5se, storage enhanced instance family.
  • Provides strong storage I/O performance with a large computing capacity
  • Ultra high packet forwarding rate
  • Equipped with 2.5 GHz Intel ® Xeon ® Platinum 8269CY (Cascade Lake) processors, up to 3.2 GHz Turbo Boost
  • Supports enabling or disabling Hyper-Threading
    Note By default, Hyper-Threading is enabled on ECS instances. For more information, see Customize CPU options.
  • Provides strong network performance with a large computing capacity
  • Supports instance type changes to g6 or c6
  • Suitable for the following scenarios:
    • Scenarios where large volumes of packets are received and transmitted, such as on-screen video comments and telecom data forwarding
    • High-performance databases and in-memory databases
    • Data analysis and mining, and distributed memory cache
    • Hadoop clusters, Spark clusters, and other memory-intensive enterprise applications
Instance types
Instance type vCPUs Memory (GiB) Local storage (GiB) Basic bandwidth (Gbit/s) Burstable bandwidth (Gbit/s) Packet forwarding rate (Kpps) IPv6 support NIC queues ENIs (including one primary ENI) Private IP addresses per ENI Disk IOPS (K) Disk bandwidth (Gbit/s)
ecs.r6.large 2 16.0 None 1.0 3.0 300 Yes 2 2 6 10 1
ecs.r6.xlarge 4 32.0 None 1.5 5.0 500 Yes 4 3 10 20 1.5
ecs.r6.2xlarge 8 64.0 None 2.5 8.0 800 Yes 8 4 10 25 2
ecs.r6.3xlarge 12 96.0 None 4.0 10.0 900 Yes 8 6 10 30 2.5
ecs.r6.4xlarge 16 128.0 None 5.0 10.0 1,000 Yes 8 8 20 40 3
ecs.r6.6xlarge 24 192.0 None 7.5 10.0 1,500 Yes 12 8 20 50 4
ecs.r6.8xlarge 32 256.0 None 10.0 None 2,000 Yes 16 8 20 60 5
ecs.r6.13xlarge 52 384 None 12.5 None 3,000 Yes 32 7 20 100 8
ecs.r6.26xlarge 104 768.0 None 25.0 None 6,000 Yes 32 15 20 200 16
Note

r5, memory optimized instance family

Features
  • I/O optimized
  • Supports enhanced SSDs (ESSDs), standard SSDs, and ultra disks
    Note The maximum performance of cloud disks varies depending on instance families. A single r5 instance can deliver an IOPS of up to 200,000. For higher storage I/O performance, we recommend that you use g5se. For more information, see g5se, storage enhanced instance family.
  • Ultra high packet forwarding rate
  • Equipped with 2.5 GHz Intel ® Xeon ® Platinum 8163 (Skylake) or 8269CY (Cascade Lake) processors
  • Provides strong network performance with a large computing capacity
  • Suitable for the following scenarios:
    • Scenarios where large volumes of packets are received and transmitted, such as on-screen video comments and telecom data forwarding
    • High-performance databases and in-memory databases
    • Data analysis and mining, and distributed memory cache
    • Hadoop clusters, Spark clusters, and other memory-intensive enterprise applications
Instance types
Instance type vCPUs Memory (GiB) Local storage (GiB) Bandwidth (Gbit/s) Packet forwarding rate (Kpps) IPv6 support NIC queues ENIs (including one primary ENI) Private IP addresses per ENI
ecs.r5.large 2 16.0 None 1.0 300 Yes 2 2 6
ecs.r5.xlarge 4 32.0 None 1.5 500 Yes 2 3 10
ecs.r5.2xlarge 8 64.0 None 2.5 800 Yes 2 4 10
ecs.r5.3xlarge 12 96.0 None 4.0 900 Yes 4 6 10
ecs.r5.4xlarge 16 128.0 None 5.0 1,000 Yes 4 8 20
ecs.r5.6xlarge 24 192.0 None 7.5 1,500 Yes 6 8 20
ecs.r5.8xlarge 32 256.0 None 10.0 2,000 Yes 8 8 20
ecs.r5.16xlarge 64 512.0 None 20.0 4,000 Yes 16 8 20
Note

Big data instance family with enhanced network performance, d1ne

Features:
  • Optimized I/O performance
  • Supports SSDs and ultra disks
  • High-capacity local SATA hard disk drives (SATA HDDs) with high throughput and a maximum of 35 Gbit/s bandwidth among instances
  • 1:4 CPU-to-memory ratio, designed for big data scenarios
  • 2.5 GHz Intel ® Xeon ® E5-2682 v4 (Broadwell) processors
  • Higher network performance along with more computing capacity
  • Scenarios:
    • Hadoop MapReduce, HDFS, Hive, and HBase
    • Spark in-memory computing and MLlib
    • Enterprises in Internet, finance, and other industries that need to compute, store, and analyze large volumes of data
    • Elasticsearch and logging
The following table describes the instance types that belong to the d1ne instance family and specifications of these instance types.
Instance type vCPUs Memory (GiB) Local storage (GiB) Bandwidth (Gbit/s, inbound or outbound) Packet forwarding rate (Kpps, inbound or outbound) Supports IPv6 NIC queues ENIs (including one primary ENI) Private IP addresses for a single ENI
ecs.d1ne.2xlarge 8 32.0 4 × 5,500 6.0 1,000 Yes 4 4 10
ecs.d1ne.4xlarge 16 64.0 8 × 5,500 12.0 1,600 Yes 4 8 20
ecs.d1ne.6xlarge 24 96.0 12 × 5,500 16.0 2,000 Yes 6 8 20
ecs.d1ne-c8d3.8xlarge 32 128.0 12 × 5,500 20.0 2,000 Yes 6 8 20
ecs.d1ne.8xlarge 32 128.0 16 × 5,500 20.0 2,500 Yes 8 8 20
ecs.d1ne-c14d3.14xlarge 56 160.0 12 × 5,500 35.0 4,500 Yes 14 8 20
ecs.d1ne.14xlarge 56 224.0 28 × 5,500 35.0 4,500 Yes 14 8 20
Note

Instance family with local SSDs, i2

Features:
  • Optimized I/O performance
  • Supports SSDs and ultra disks
  • High-performance local NVMe SSDs with high IOPS, high I/O throughput, and low latency
  • 1:8 CPU-to-memory ratio, designed for high-performance databases
  • 2.5 GHz Intel ® Xeon ® Platinum 8163 (Skylake) processors
  • Higher network performance along with more computing capacity
  • Scenarios:
    • Online transaction processing (OLTP) and high-performance relational databases
    • NoSQL databases such as Cassandra and MongoDB
    • Search scenarios such as Elasticsearch
The following table describes the instance types that belong to the i2 instance family and specifications of these instance types.
Instance type vCPUs Memory (GiB) Local storage (GiB) Bandwidth (Gbit/s, inbound or outbound) Packet forwarding rate (Kpps, inbound or outbound) Supports IPv6 NIC queues ENIs (including one primary ENI) Private IP addresses for a single ENI
ecs.i2.xlarge 4 32.0 1 × 894 1.0 500 Yes 2 3 10
ecs.i2.2xlarge 8 64.0 1 × 1,788 2.0 1,000 Yes 2 4 10
ecs.i2.4xlarge 16 128.0 2 × 1,788 3.0 1,500 Yes 4 8 20
ecs.i2.8xlarge 32 256.0 4 × 1,788 6.0 2,000 Yes 8 8 20
ecs.i2.16xlarge 64 512.0 8 × 1,788 10.0 4,000 Yes 16 8 20
Note

Instance family with local SSDs, i2g

Features:
  • Optimized I/O performance
  • Supports SSDs and ultra disks
  • High-performance local NVMe SSDs with high IOPS, high I/O throughput, and low latency
  • 1:4 CPU-to-memory ratio, designed for high-performance databases
  • 2.5 GHz Intel ® Xeon ® Platinum 8163 (Skylake) processors
  • Higher network performance along with more computing capacity
  • Scenarios:
    • OLTP and high-performance relational databases
    • NoSQL databases such as Cassandra and MongoDB
    • Search scenarios such as Elasticsearch
The following table describes the instance types that belong to the i2g instance family and specifications of these instance types.
Instance type vCPUs Memory (GiB) Local storage (GiB) Bandwidth (Gbit/s, inbound or outbound) Packet forwarding rate (Kpps, inbound or outbound) Supports IPv6 NIC queues ENIs (including one primary ENI) Private IP addresses for a single ENI
ecs.i2g.2xlarge 8 32.0 1 × 894 2.0 1,000 No 2 4 10
ecs.i2g.4xlarge 16 64.0 1 × 1,788 3.0 1,500 No 4 8 20
ecs.i2g.8xlarge 32 128.0 2 × 1,788 6.0 2,000 No 8 8 20
ecs.i2g.16xlarge 64 256.0 4 × 1,788 10.0 4,000 No 16 8 20
Note

hfc6, compute optimized instance family with high clock speed

Features
  • Provides predictable stable and high performance and reduces virtualization overheads with the use of the ECS Bare Metal Instance architecture.
  • I/O optimized.
  • Supports enhanced SSDs, standard SSDs, and ultra disks.
  • Provides strong storage I/O performance proportional to computing capacity.
  • CPU-to-memory ratio of 1:2.
  • Ultra high packet forwarding rate.
  • Equipped with 3.1 GHz Intel ® Xeon ® Platinum 8269 (Cascade Lake), up to 3.5 GHz Turbo Boost.
    Note The CPU of the current instance family provides a 3.1 GHz clock speed. Due to the Intel ISS feature, a lower clock speed may be displayed. Alibaba Cloud is trying its best to fix the display problem. The problem does not affect the actual clock speed of your instances purchased.
  • Supports enabling or disabling Hyper-Threading
    Note By default, Hyper-Threading is enabled on ECS instances. For more information, see Customize CPU options.
  • Provides strong network performance proportional to computing capacity.
  • Suitable for the following scenarios:
    • Scenarios where large volumes of packets are received and transmitted, such as bullet screen and re-transmission of telecommunication information.
    • Web frontend servers.
    • Frontends of massive multiplayer online (MMO) gaming.
    • Data analysis, batch processing, and video encoding.
    • High-performance scientific and engineering applications.
Instance types
Instance type vCPU Memory (GiB) Local storage (GiB) Base bandwidth (inbound or outbound in Gbit/s) Burstable bandwidth ((inbound or outbound in Gbit/s) Packet forwarding rate (inbound or outbound in Kpps) IPv6 support NIC queues ENI Number of private IP addresses for a single ENI Disk IOPS (K) Disk bandwidth (Gbit/s)
ecs.hfc6.large 2 4.0 None 1.0 3.0 300 Yes 2 2 6 10 1.0
ecs.hfc6.xlarge 4 8.0 None 1.5 5.0 500 Yes 4 3 10 20 1.5
ecs.hfc6.2xlarge 8 16.0 None 2.5 8.0 800 Yes 8 4 10 25 2.0
ecs.hfc6.3xlarge 12 24.0 None 4.0 10.0 900 Yes 8 6 10 30 2.5
ecs.hfc6.4xlarge 16 32.0 None 5.0 10.0 1,000 Yes 8 8 20 40 3.0
ecs.hfc6.6xlarge 24 48.0 None 7.5 10.0 1,500 Yes 12 8 20 50 4.0
ecs.hfc6.8xlarge 32 64.0 None 10.0 None 2,000 Yes 16 8 20 60 5.0
ecs.hfc6.10xlarge 40 96.0 None 12.5 None 3,000 Yes 32 7 20 100 8.0
ecs.hfc6.16xlarge 64 128.0 None 20.0 None 4,000 Yes 32 8 20 120 10.0
ecs.hfc6.20xlarge 80 192.0 None 25.0 None 6,000 Yes 32 15 20 200 16.0
Note

hfg6, general purpose instance family with high clock speed

Features
  • Provides predictable stable and high performance and reduces virtualization overheads with the use of the ECS Bare Metal Instance architecture.
  • I/O optimized.
  • Supports enhanced SSDs, standard SSDs, and ultra disks.
  • Provides strong storage I/O performance proportional to computing capacity.
  • CPU-to-memory ratio of 1:4.
  • Ultra high packet forwarding rate.
  • Equipped with 3.1 GHz Intel ® Xeon ® Platinum 8269 (Cascade Lake), up to 3.5 GHz Turbo Boost.
    Note The CPU of the current instance family provides a 3.1 GHz clock speed. Due to the Intel ISS feature, a lower clock speed may be displayed. Alibaba Cloud is trying its best to fix the display problem. The problem does not affect the actual clock speed of your instances purchased.
  • Supports enabling or disabling Hyper-Threading
    Note By default, Hyper-Threading is enabled on ECS instances. For more information, see Customize CPU options.
  • Provides strong network performance proportional to computing capacity.
  • Suitable for the following scenarios:
    • Scenarios where large volumes of packets are received and transmitted, such as bullet screen and re-transmission of telecommunication information.
    • Enterprise-level applications of various types and sizes.
    • Websites and application servers.
    • Game servers.
    • Small and medium-sized database systems, cache, and search clusters.
    • Data analysis and computing.
    • Compute clusters and memory-intensive data processing.
Instance types
Instance type vCPU Memory (GiB) Local storage (GiB) Base bandwidth (inbound or outbound in Gbit/s) Burstable bandwidth ((inbound or outbound in Gbit/s) Packet forwarding rate (inbound or outbound in Kpps) IPv6 support NIC queues ENI Number of private IP addresses for a single ENI Disk IOPS (K) Disk bandwidth (Gbit/s)
ecs.hfg6.large 2 8.0 None 1.0 3.0 300 Yes 2 2 6 10 1.0
ecs.hfg6.xlarge 4 16.0 None 1.5 5.0 500 Yes 4 3 10 20 1.5
ecs.hfg6.2xlarge 8 32.0 None 2.5 8.0 800 Yes 8 4 10 25 2.0
ecs.hfg6.3xlarge 12 48.0 None 4.0 10.0 900 Yes 8 6 10 30 2.5
ecs.hfg6.4xlarge 16 64.0 None 5.0 10.0 1,000 Yes 8 8 20 40 3.0
ecs.hfg6.6xlarge 24 96.0 None 7.5 10.0 1,500 Yes 12 8 20 50 4.0
ecs.hfg6.8xlarge 32 128.0 None 10.0 None 2,000 Yes 16 8 20 60 5.0
ecs.hfg6.10xlarge 40 192.0 None 12.5 None 3,000 Yes 32 7 20 100 8.0
ecs.hfg6.16xlarge 64 256.0 None 20.0 None 4,000 Yes 32 8 20 120 10.0
ecs.hfg6.20xlarge 80 384.0 None 25.0 None 6,000 Yes 32 15 20 200 16.0
Note

hfr6, memory optimized instance family with high clock speed

Features
  • Provides predictable stable and high performance and reduces virtualization overheads with the use of the ECS Bare Metal Instance architecture.
  • I/O optimized.
  • Supports enhanced SSDs, standard SSDs, and ultra disks.
  • Provides strong storage I/O performance proportional to computing capacity.
  • CPU-to-memory ratio of 1:8.
  • Ultra high packet forwarding rate.
  • Equipped with 3.1 GHz Intel ® Xeon ® Platinum 8269 (Cascade Lake), up to 3.5 GHz Turbo Boost.
    Note The CPU of the current instance family provides a 3.1 GHz clock speed. Due to the Intel ISS feature, a lower clock speed may be displayed. Alibaba Cloud is trying its best to fix the display problem. The problem does not affect the actual clock speed of your instances purchased.
  • Supports enabling or disabling Hyper-Threading
    Note By default, Hyper-Threading is enabled on ECS instances. For more information, see Customize CPU options.
  • Provides strong network performance proportional to computing capacity.
  • Suitable for the following scenarios:
    • Scenarios where large volumes of packets are received and transmitted, such as bullet screen and re-transmission of telecommunication information.
    • High-performance and in-memory databases.
    • Data analysis and mining, and distributed memory cache.
    • Hadoop, Spark, and other memory-intensive enterprise applications.
Instance types
Instance type vCPU Memory (GiB) Local storage (GiB) Base bandwidth (inbound or outbound in Gbit/s) Burstable bandwidth ((inbound or outbound in Gbit/s) Packet forwarding rate (inbound or outbound in Kpps) IPv6 support NIC queues ENI Number of private IP addresses for a single ENI Disk IOPS (K) Disk bandwidth (Gbit/s)
ecs.hfr6.large 2 16.0 None 1.0 3.0 300 Yes 2 2 6 10 1.0
ecs.hfr6.xlarge 4 32.0 None 1.5 5.0 500 Yes 4 3 10 20 1.5
ecs.hfr6.2xlarge 8 64.0 None 2.5 8.0 800 Yes 8 4 10 25 2.0
ecs.hfr6.3xlarge 12 96.0 None 4.0 10.0 900 Yes 8 6 10 30 2.5
ecs.hfr6.4xlarge 16 128.0 None 5.0 10.0 1,000 Yes 8 8 20 40 3.0
ecs.hfr6.6xlarge 24 192.0 None 7.5 10.0 1,500 Yes 12 8 20 50 4.0
ecs.hfr6.8xlarge 32 256.0 None 10.0 None 2,000 Yes 16 8 20 60 5.0
ecs.hfr6.10xlarge 40 384.0 None 12.5 None 3,000 Yes 32 7 20 100 8.0
ecs.hfr6.16xlarge 64 512.0 None 20.0 None 4,000 Yes 32 8 20 120 10.0
ecs.hfr6.20xlarge 80 768.0 None 25.0 None 6,000 Yes 32 15 20 200 16.0
Note

sn2ne, general purpose instance family with enhanced network performance

Features
  • I/O optimized
  • Supports standard SSDs and ultra disks
  • CPU-to-memory ratio of 1:4
  • Ultra high packet forwarding rate
  • Equipped with 2.5 GHz Intel ® Xeon ® E5-2682 v4 (Broadwell) or Platinum 8163 (Skylake) processors
  • Provides strong network performance with a large computing capacity
  • Suitable for the following scenarios:
    • Scenarios where large volumes of packets are received and transmitted, such as on-screen video comments and telecom data forwarding
    • Enterprise-grade applications of various types and sizes
    • Small and medium-sized database systems, caches, and search clusters
    • Data analysis and computing
    • Computing clusters and memory-intensive data processing
Instance types
Instance type vCPUs Memory (GiB) Local storage (GiB) Bandwidth (Gbit/s) Packet forwarding rate (Kpps) IPv6 support NIC queues ENIs (including one primary ENI) Private IP addresses per ENI
ecs.sn2ne.large 2 8.0 None 1.0 300 Yes 2 2 6
ecs.sn2ne.xlarge 4 16.0 None 1.5 500 Yes 2 3 10
ecs.sn2ne.2xlarge 8 32.0 None 2.0 1,000 Yes 4 4 10
ecs.sn2ne.3xlarge 12 48.0 None 2.5 1,300 Yes 4 6 10
ecs.sn2ne.4xlarge 16 64.0 None 3.0 1,600 Yes 4 8 20
ecs.sn2ne.6xlarge 24 96.0 None 4.5 2,000 Yes 6 8 20
ecs.sn2ne.8xlarge 32 128.0 None 6.0 2,500 Yes 8 8 20
ecs.sn2ne.14xlarge 56 224.0 None 10.0 4,500 Yes 14 8 20
Note

sn1ne, compute optimized instance family with enhanced network performance

Features
  • I/O optimized
  • Supports standard SSDs and ultra disks
  • CPU-to-memory ratio of 1:2
  • Ultra high packet forwarding rate
  • Equipped with 2.5 GHz Intel ® Xeon ® E5-2682 v4 (Broadwell) or Platinum 8163 (Skylake) processors
  • Provides strong network performance with a large computing capacity
  • Suitable for the following scenarios:
    • Scenarios where large volumes of packets are received and transmitted, such as on-screen video comments and telecom data forwarding
    • Web frontend servers
    • Frontend servers of massive multiplayer online (MMO) games
    • Data analysis, batch compute, and video encoding
    • High-performance science and engineering applications
Instance types
Instance type vCPUs Memory (GiB) Local storage (GiB) Bandwidth (Gbit/s) Packet forwarding rate (Kpps) IPv6 support NIC queues ENIs (including one primary ENI) Private IP addresses per ENI
ecs.sn1ne.large 2 4.0 None 1.0 300 Yes 2 2 6
ecs.sn1ne.xlarge 4 8.0 None 1.5 500 Yes 2 3 10
ecs.sn1ne.2xlarge 8 16.0 None 2.0 1,000 Yes 4 4 10
ecs.sn1ne.3xlarge 12 24.0 None 2.5 1,300 Yes 4 6 10
ecs.sn1ne.4xlarge 16 32.0 None 3.0 1,600 Yes 4 8 20
ecs.sn1ne.6xlarge 24 48.0 None 4.5 2,000 Yes 6 8 20
ecs.sn1ne.8xlarge 32 64.0 None 6.0 2,500 Yes 8 8 20
Note

re4, memory optimized instance family with enhanced performance

Features
  • I/O optimized.
  • Supports standard SSDs and ultra disks.
  • Optimized for high-performance databases, in-memory databases, and other memory-intensive enterprise applications.
  • Equipped with 2.2 GHz Intel ® Xeon ® E7 8880 v4 (Broadwell) processors, up to 2.4 GHz Turbo Boost.
  • CPU-to-memory ratio of 1:12 and up to 1920.0 GiB memory.
  • The ecs.re4.20xlarge and ecs.re4.40xlarge instance types are SAP HANA-certified.
  • Suitable for the following scenarios:
    • High-performance databases and in-memory databases such as SAP HANA.
    • Memory-intensive applications.
    • Big data processing engines such as Apache Spark and Presto.
Instance types
Instance type vCPUs Memory (GiB) Local storage (GiB) Bandwidth (Gbit/s) Packet forwarding rate (Kpps) IPv6 support NIC queues ENIs (including one primary ENI) Private IP addresses per ENI
ecs.re4.20xlarge 80 960.0 None 15.0 2,000 Yes 16 8 20
ecs.re4.40xlarge 160 1920.0 None 30.0 4,500 Yes 16 8 20
Note

re4e, memory optimized instance family with enhanced performance

Features
  • I/O optimized
  • Supports standard SSDs and ultra disks
  • Optimized for high-performance databases, in-memory databases, and other memory-intensive enterprise applications
  • Equipped with 2.2 GHz Intel ® Xeon ® E7 8880 v4 (Broadwell) processors, up to 2.4 GHz Turbo Boost
  • CPU-to-memory ratio of 1:24 and up to 3840.0 GiB memory
  • Suitable for the following scenarios:
    • High-performance databases and in-memory databases such as SAP HANA
    • Memory-intensive applications
    • Big data processing engines such as Apache Spark and Presto
Instance types
Instance type vCPUs Memory (GiB) Local storage (GiB) Bandwidth (Gbit/s) Packet forwarding rate (Kpps) IPv6 support NIC queues ENIs (including one primary ENI) Private IP addresses per ENI
ecs.re4e.40xlarge 160 3840.0 None 30.0 4,500 Yes 16 15 20
Note

se1ne, memory optimized instance family with enhanced network performance

Features
  • I/O optimized
  • Supports standard SSDs and ultra disks
  • CPU-to-memory ratio of 1:8
  • Ultra high packet forwarding rate
  • Equipped with 2.5 GHz Intel ® Xeon ® E5-2682 v4 (Broadwell) or Platinum 8163 (Skylake) processors
  • Provides strong network performance with a large computing capacity
  • Suitable for the following scenarios:
    • Scenarios where large volumes of packets are received and transmitted, such as on-screen video comments and telecom data forwarding
    • High-performance databases and in-memory databases
    • Data analysis and mining, and distributed memory cache
    • Hadoop clusters, Spark clusters, and other memory-intensive enterprise applications
Instance types
Instance type vCPUs Memory (GiB) Local storage (GiB) Bandwidth (Gbit/s) Packet forwarding rate (Kpps) IPv6 support NIC queues ENIs (including one primary ENI) Private IP addresses per ENI
ecs.se1ne.large 2 16.0 None 1.0 300 Yes 2 2 6
ecs.se1ne.xlarge 4 32.0 None 1.5 500 Yes 2 3 10
ecs.se1ne.2xlarge 8 64.0 None 2.0 1,000 Yes 4 4 10
ecs.se1ne.3xlarge 12 96.0 None 2.5 1,300 Yes 4 6 10
ecs.se1ne.4xlarge 16 128.0 None 3.0 1,600 Yes 4 8 20
ecs.se1ne.6xlarge 24 192.0 None 4.5 2,000 Yes 6 8 20
ecs.se1ne.8xlarge 32 256.0 None 6.0 2,500 Yes 8 8 20
ecs.se1ne.14xlarge 56 480.0 None 10.0 4,500 Yes 14 8 20
Note

se1, memory optimized instance family

Features
  • I/O optimized
  • Supports standard SSDs and ultra disks
  • CPU-to-memory ratio of 1:8
  • Equipped with 2.5 GHz Intel ® Xeon ® E5-2682 v4 (Broadwell) processors
  • Provides strong network performance with a large computing capacity
  • Suitable for the following scenarios:
    • High-performance databases and in-memory databases
    • Data analysis and mining, and distributed memory cache
    • Hadoop clusters, Spark clusters, and other memory-intensive enterprise applications
Instance types
Instance type vCPUs Memory (GiB) Local storage (GiB) Bandwidth (Gbit/s) Packet forwarding rate (Kpps) IPv6 support NIC queues ENIs (including one primary ENI) Private IP addresses per ENI
ecs.se1.large 2 16.0 None 0.5 100 No 1 2 6
ecs.se1.xlarge 4 32.0 None 0.8 200 No 1 3 10
ecs.se1.2xlarge 8 64.0 None 1.5 400 No 1 4 10
ecs.se1.4xlarge 16 128.0 None 3.0 500 No 2 8 20
ecs.se1.8xlarge 32 256.0 None 6.0 800 No 3 8 20
ecs.se1.14xlarge 56 480.0 None 10.0 1,200 No 4 8 20
Note

Big data instance family, d1

Features:
  • Optimized I/O performance
  • Supports SSDs and ultra disks
  • High-capacity local SATA HDDs with high throughput and up to 17 Gbit/s of bandwidth among instances
  • 1:4 CPU-to-memory ratio, designed for big data scenarios
  • 2.5 GHz Intel ® Xeon ® E5-2682 v4 (Broadwell) processors
  • Higher network performance along with more computing capacity
  • Scenarios:
    • Hadoop MapReduce, HDFS, Hive, and HBase
    • Spark in-memory computing and MLlib
    • Enterprises in Internet, finance, other industries that need to compute, store, and analyze large volumes of data
    • Elasticsearch and logging
The following table describes the instance types that belong to the d1 instance family and specifications of these instance types.
Instance type vCPUs Memory (GiB) Local storage (GiB) Bandwidth (Gbit/s, inbound or outbound) Packet forwarding rate (Kpps, inbound or outbound) Supports IPv6 NIC queues ENIs (including one primary ENI) Private IP addresses for a single ENI
ecs.d1.2xlarge 8 32.0 4 × 5,500 3.0 300 No 1 4 10
ecs.d1.3xlarge 12 48.0 6 × 5,500 4.0 400 No 1 6 10
ecs.d1.4xlarge 16 64.0 8 × 5,500 6.0 600 No 2 8 20
ecs.d1.6xlarge 24 96.0 12 × 5,500 8.0 800 No 2 8 20
ecs.d1-c8d3.8xlarge 32 128.0 12 × 5,500 10.0 1,000 No 4 8 20
ecs.d1.8xlarge 32 128.0 16 × 5,500 10.0 1,000 No 4 8 20
ecs.d1-c14d3.14xlarge 56 160.0 12 × 5,500 17.0 1,800 No 6 8 20
ecs.d1.14xlarge 56 224.0 28 × 5,500 17.0 1,800 No 6 8 20
Note

Instance family with local SSDs, i1

Features:
  • Optimized I/O performance
  • Supports SSDs and ultra disks
  • High-performance local NVMe SSDs with high IOPS, high I/O throughput, and low latency
  • 1:4 CPU-to-memory ratio, designed for high-performance databases
  • 2.5 GHz Intel ® Xeon ® E5-2682 v4 (Broadwell) processors
  • Higher network performance along with more computing capacity
  • Scenarios:
    • Online transaction processing (OLTP) and high-performance relational databases
    • NoSQL databases such as Cassandra and MongoDB
    • Search scenarios such as Elasticsearch
The following table describes the instance types that belong to the i1 instance family and specifications of these instance types.
Instance type vCPUs Memory (GiB) Local storage (GiB) Bandwidth (Gbit/s, inbound or outbound) Packet forwarding rate (Kpps, inbound or outbound) Supports IPv6 NIC queues ENIs (including one primary ENI) Private IP addresses for a single ENI
ecs.i1.xlarge 4 16.0 2 × 104 0.8 200 No 1 3 10
ecs.i1.2xlarge 8 32.0 2 × 208 1.5 400 No 1 4 10
ecs.i1.3xlarge 12 48.0 2 × 312 2.0 400 No 1 6 10
ecs.i1.4xlarge 16 64.0 2 × 416 3.0 500 No 2 8 20
ecs.i1-c5d1.4xlarge 16 64.0 2 × 1,456 3.0 400 No 2 8 20
ecs.i1.6xlarge 24 96.0 2 × 624 4.5 600 No 2 8 20
ecs.i1.8xlarge 32 128.0 2 × 832 6.0 800 No 3 8 20
ecs.i1-c10d1.8xlarge 32 128.0 2 × 1,456 6.0 800 No 3 8 20
ecs.i1.14xlarge 56 224.0 2 × 1,456 10.0 1,200 No 4 8 20
Note

hfc5, compute optimized instance family with high clock speed

Features
  • I/O optimized.
  • Supports standard SSDs and ultra disks.
  • Consistent computing performance.
  • Equipped with 3.1 GHz Intel ® Xeon ® Gold 6149 (Skylake) processors.
  • CPU-to-memory ratio of 1:2.
  • Provides strong network performance proportional to computing capacity.
  • Suitable for the following scenarios:
    • High-performance Web frontend servers.
    • High-performance scientific and engineering applications.
    • MMO gaming and video coding.
Instance types
Instance type vCPU Memory (GiB) Local storage (GiB) Base bandwidth (inbound or outbound in Gbit/s) Packet forwarding rate (inbound or outbound in Kpps) IPv6 support NIC queues ENI Private IP addresses of a single ENI
ecs.hfc5.large 2 4.0 None 1.0 300 No 2 2 6
ecs.hfc5.xlarge 4 8.0 None 1.5 500 No 2 3 10
ecs.hfc5.2xlarge 8 16.0 None 2.0 1,000 No 2 4 10
ecs.hfc5.3xlarge 12 24.0 None 2.5 1,300 No 4 6 10
ecs.hfc5.4xlarge 16 32.0 None 3.0 1,600 No 4 8 20
ecs.hfc5.6xlarge 24 48.0 None 4.5 2,000 No 6 8 20
ecs.hfc5.8xlarge 32 64.0 None 6.0 2,500 No 8 8 20
Note

hfg5, general purpose instance family with high clock speed

Features
  • I/O optimized.
  • Supports standard SSDs and ultra disks.
  • Consistent computing performance.
  • Equipped with 3.1 GHz Intel ® Xeon ® Gold 6149 (Skylake) processors.
  • CPU-to-memory ratio of 1:4 (excluding the instance type with 56 vCPUs).
  • Provides strong network performance proportional to computing capacity.
  • Suitable for the following scenarios:
    • High-performance web frontend servers.
    • High-performance scientific and engineering applications.
    • MMO gaming and video coding.
Instance types
Instance type vCPU Memory (GiB) Local storage (GiB) Base bandwidth (inbound or outbound in Gbit/s) Packet forwarding rate (inbound or outbound in Kpps) IPv6 support NIC queues ENI Private IP addresses of a single ENI
ecs.hfg5.large 2 8.0 None 1.0 300 No 2 2 6
ecs.hfg5.xlarge 4 16.0 None 1.5 500 No 2 3 10
ecs.hfg5.2xlarge 8 32.0 None 2.0 1,000 No 2 4 10
ecs.hfg5.3xlarge 12 48.0 None 2.5 1,300 No 4 6 10
ecs.hfg5.4xlarge 16 64.0 None 3.0 1,600 No 4 8 20
ecs.hfg5.6xlarge 24 96.0 None 4.5 2,000 No 6 8 20
ecs.hfg5.8xlarge 32 128.0 None 6.0 2,500 No 8 8 20
ecs.hfg5.14xlarge 56 160.0 None 10.0 4,000 No 14 8 20
Note

gn6i, compute optimized instance family with GPU capabilities

Features
  • I/O optimized
  • CPU-to-memory ratio of 1:4
  • Equipped with 2.5 GHz Intel ® Xeon ® Platinum 8163 (Skylake) processors
  • Supports standard SSDs, ultra disks, and enhanced SSDs that deliver millions of IOPS
  • Achieves better performance with the new-generation X-Dragon compute architecture
  • Uses NVIDIA T4 GPU computing accelerators
    • Powered by the new NVIDIA Turing architecture
    • Up to 320 Turing Tensor Cores
    • Up to 2,560 CUDA Cores
    • Mixed-precision Tensor Cores support 65 FP16 TFLOPS, 130 INT8 TOPS, and 260 INT4 TOPS
    • 16 GB memory capacity (320 GB/s bandwidth)
  • Provides strong network performance with large computing capacity
  • Suitable for the following scenarios:
    • AI (deep learning and machine learning) inference for computer vision, voice recognition, speech synthesis, natural language processing, machine translation, and reference systems
    • Real-time rendering for cloud gaming
    • Real-time rendering for AR and VR applications
    • Reload graphics computing or Graphics Workstation
    • GPU-accelerated databases
    • High-performance computing
Instance types
Instance type vCPUs Memory (GiB) Local storage (GiB) GPU GPU memory (GB) Bandwidth (Gbit/s) Packet forwarding rate (Kpps) IPv6 support NIC queues ENIs
ecs.gn6i-c4g1.xlarge 4 15 N/A T4*1 16 4 500 Yes 2 2
ecs.gn6i-c8g1.2xlarge 8 31 N/A T4*1 16 5 800 Yes 2 2
ecs.gn6i-c16g1.4xlarge 16 62 N/A T4*1 16 6 1,000 Yes 4 3
ecs.gn6i-c24g1.6xlarge 24 93 N/A T4*1 16 7.5 1,200 Yes 6 4
ecs.gn6i-c24g1.12xlarge 48 186 N/A T4*2 32 15 2,400 Yes 12 6
ecs.gn6i-c24g1.24xlarge 96 372 N/A T4*4 64 30 4,800 Yes 24 8
Note

gn6e, compute optimized instance family with GPU capabilities

Features
  • I/O optimized
  • Supports enhanced SSDs, standard SSDs, and ultra disks
  • Uses NVIDIA V100 (32 GB NVLink) GPU processors
  • CPU-to-memory ratio of 1:4
  • Equipped with 2.5 GHz Intel ® Xeon ® Platinum 8163 (Skylake) processors
  • Uses NVIDIA V100 GPU computing accelerators (SXM2-based)
    • Powered by the new NVIDIA Volta architecture
    • 16 GB of HBM2 memory (900 GB/s bandwidth)
    • 5,120 CUDA Cores
    • 640 Tensor Cores
    • Supports up to six NVLink connections for a total bandwidth of 300 GB/s with 25 GB/s each
  • Provides strong network performance based on large computing capacity
  • Suitable for the following scenarios:
    • Deep learning applications such as training and inference applications of AI algorithms used in image classification, autonomous vehicles, and speech recognition
    • Scientific computing applications, such as fluid dynamics, finance, molecular dynamics, and environmental analysis
Instance types
Instance type vCPUs Memory (GiB) Local storage (GiB) GPUs GPU memory (GB) Bandwidth (Gbit/s) Packet forwarding rate (Kpps) IPv6 support NIC queues ENIs (including one primary ENI) Private IP addresses per ENI
ecs.gn6e-c12g1.3xlarge 12 92.0 None 1 × V100 32 5.0 800 Yes 8 6 10
ecs.gn6e-c12g1.12xlarge 48 368.0 None 4 × V100 128 16.0 2,400 Yes 8 8 20
ecs.gn6e-c12g1.24xlarge 96 736.0 None 8 × V100 256 32.0 4,800 Yes 16 8 20
Note

gn6v, compute optimized instance family with GPU capabilities

Features
  • I/O optimized
  • Supports enhanced SSDs, standard SSDs, and ultra disks.
  • Uses NVIDIA V100 GPU processors
  • CPU-to-memory ratio of 1:4
  • Equipped with 2.5 GHz Intel ® Xeon ® Platinum 8163 (Skylake) processors
  • Uses NVIDIA V100 GPU computing accelerators (SXM2-based)
    • Powered by the innovative NVIDIA Volta architecture
    • 16 GB HBM2 GPU memory
    • 5,120 CUDA Cores
    • 640 Tensor Cores
    • Memory bandwidth of 900 GB/s
    • Supports up to six NVLink connections for a total bandwidth of 300 GB/s with 25 GB/s each
  • Provides strong network performance with large computing capacity
  • Suitable for the following scenarios:
    • Deep learning, image classification, autonomous vehicles, voice recognition, and other AI applications
    • Scientific computing applications, such as fluid dynamics, finance, molecular dynamics, and environmental analysis
Instance types
Instance type vCPUs Memory (GiB) Local storage (GiB) GPU GPU memory (GB) Bandwidth (Gbit/s) Packet forwarding rate (Kpps) IPv6 support NIC queues ENIs Private IP address of a single ENI
ecs.gn6v-c8g1.2xlarge 8 32.0 N/A 1 * NVIDIA V100 1 * 16 2.5 800 Yes 4 4 10
ecs.gn6v-c8g1.8xlarge 32 128.0 N/A 4 * NVIDIA V100 4 * 16 10.0 2,000 Yes 8 8 20
ecs.gn6v-c8g1.16xlarge 64 256.0 N/A 8 * NVIDIA V100 8 * 16 20.0 2,500 Yes 16 8 20
Note

FPGA-based compute-optimized instance family, f3

Features:
  • Optimized I/O performance
  • Supports SSDs and ultra disks
  • Xilinx 16nm Virtex UltraScale+ VU9P
  • 1:4 CPU-to-memory ratio
  • 2.5 GHz Intel ® Xeon ® Platinum 8163 (Skylake) processors
  • Higher network performance along with more computing capacity
  • Scenarios:
    • Deep learning and reasoning
    • Genomics research
    • Database acceleration
    • Image transcoding such as converting JPEG images to WebP images
    • Real-time video processing such as H.265 video compression
The following table describes the instance types that belong to the f3 instance family and specifications of these instance types.
Instance type vCPUs Memory (GiB) Local storage (GiB) FPGAs Bandwidth (Gbit/s, inbound or outbound) Packet forwarding rate (Kpps, inbound or outbound) Supports IPv6 NIC queues ENIs (including one primary ENI) Private IP addresses for a single ENI
ecs.f3-c4f1.xlarge 4 16.0 None 1 × Xilinx VU9P 1.5 300 No 2 3 10
ecs.f3-c8f1.2xlarge 8 32.0 None 1 × Xilinx VU9P 2.5 500 No 4 4 10
ecs.f3-c16f1.4xlarge 16 64.0 None 1 × Xilinx VU9P 5.0 1,000 No 4 8 20
ecs.f3-c16f1.8xlarge 32 128.0 None 2 × Xilinx VU9P 10.0 2,000 No 8 8 20
ecs.f3-c16f1.16xlarge 64 256.0 None 4 × Xilinx VU9P 20.0 2,500 No 16 8 20
Note

vgn5i, lightweight compute optimized instance family with GPU capabilities

Features
  • I/O optimized
  • Supports standard SSDs and ultra disks
  • Uses NVIDIA P4 GPU computing accelerators
  • Contains virtual GPUs (vGPUs), which are the result of GPU virtualization with mediated pass-through
    • Supports the 1/8, 1/4, 1/2, and 1/1 computing capacity of NVIDIA® Tesla® P4 GPUs
    • Supports 1, 2, 4, and 8 GB of GPU memory
  • CPU-to-memory ratio of 1:3
  • Equipped with 2.5 GHz Intel ® Xeon ® E5-2682 v4 (Broadwell) processors
  • Provides strong network performance with large computing capacity
  • Suitable for the following scenarios:
    • Real-time rendering for cloud gaming
    • Real-time rendering for AR and VR applications
    • AI (deep learning and machine learning) inference for the elastic deployment of Internet services
    • Educational environment of deep learning
    • Modeling experiment environment of deep learning
Instance types
Instance type vCPUs Memory (GiB) Local storage (GiB) GPU GPU memory (GB) Bandwidth (Gbit/s) Packet forwarding rate (Kpps) IPv6 support NIC queues ENIs
ecs.vgn5i-m1.large 2 6 N/A P4*1/8 1 1 300 Yes 2 2
ecs.vgn5i-m2.xlarge 4 12 N/A P4*1/4 2 2 500 Yes 2 3
ecs.vgn5i-m4.2xlarge 8 24 N/A P4 * 1/2 4 3 800 Yes 2 4
ecs.vgn5i-m8.4xlarge 16 48 N/A P4*1 8 5 1,000 Yes 4 5
Note

gn5, compute optimized instance family with GPU capabilities

Features
  • I/O optimized
  • Supports standard SSDs and ultra disks
  • Uses NVIDIA P100 GPU processors
  • Multiple CPU-to-memory ratios
  • High-performance local NVMe SSDs
  • Equipped with 2.5 GHz Intel ® Xeon ® E5-2682 v4 (Broadwell) processors
  • Provides strong network performance with large computing capacity
  • Suitable for the following scenarios:
    • Deep learning
    • Scientific computing applications, such as fluid dynamics, finance, genomics, and environmental analysis
    • High-performance computing, rendering, multimedia encoding and decoding, and other server-side GPU compute workloads
Instance types
Instance type vCPUs Memory (GiB) Local storage (GiB) GPU GPU memory (GB) Bandwidth (Gbit/s) Packet forwarding rate (Kpps) IPv6 support NIC queues ENIs Private IP address of a single ENI
ecs.gn5-c4g1.xlarge 4 30.0 440 1 * NVIDIA P100 1 * 16 3.0 300 No 1 3 10
ecs.gn5-c8g1.2xlarge 8 60.0 440 1 * NVIDIA P100 1 * 16 3.0 400 No 1 4 10
ecs.gn5-c4g1.2xlarge 8 60.0 880 2 * NVIDIA P100 2 * 16 5.0 1,000 No 2 4 10
ecs.gn5-c8g1.4xlarge 16 120.0 880 2 * NVIDIA P100 2 * 16 5.0 1,000 No 4 8 20
ecs.gn5-c28g1.7xlarge 28 112.0 440 1 * NVIDIA P100 1 * 16 5.0 1,000 No 8 8 20
ecs.gn5-c8g1.8xlarge 32 240.0 1,760 4 * NVIDIA P100 4 * 16 10.0 2,000 No 8 8 20
ecs.gn5-c28g1.14xlarge 56 224.0 880 2 * NVIDIA P100 2 * 16 10.0 2,000 No 14 8 20
ecs.gn5-c8g1.14xlarge 54 480.0 3,520 8 * NVIDIA P100 8 * 16 25.0 4,000 No 14 8 20
Note

gn5i, compute optimized instance family with GPU capabilities

Features
  • I/O optimized
  • Supports standard SSDs and ultra disks
  • Uses NVIDIA P4 GPU processors
  • CPU-to-memory ratio of 1:4
  • Equipped with 2.5 GHz Intel ® Xeon ® E5-2682 v4 (Broadwell) processors
  • Provides strong network performance with large computing capacity
  • Suitable for the following scenarios:
    • Deep learning inference
    • Multimedia encoding and decoding, and other server-side GPU compute workloads
Instance types
Instance type vCPUs Memory (GiB) Local storage (GiB) GPU GPU memory (GB) Bandwidth (Gbit/s) Packet forwarding rate (Kpps) IPv6 support NIC queues ENIs Private IP address of a single ENI
ecs.gn5i-c2g1.large 2 8.0 N/A 1 * NVIDIA P4 1 * 8 1.0 100 Yes 2 2 6
ecs.gn5i-c4g1.xlarge 4 16.0 N/A 1 * NVIDIA P4 1 * 8 1.5 200 Yes 2 3 10
ecs.gn5i-c8g1.2xlarge 8 32.0 N/A 1 * NVIDIA P4 1 * 8 2.0 400 Yes 4 4 10
ecs.gn5i-c16g1.4xlarge 16 64.0 N/A 1 * NVIDIA P4 1 * 8 3.0 800 Yes 4 8 20
ecs.gn5i-c16g1.8xlarge 32 128.0 N/A 2 * NVIDIA P4 2 * 8 6.0 1,200 Yes 8 8 20
ecs.gn5i-c28g1.14xlarge 56 224.0 N/A 2 * NVIDIA P4 2 * 8 10.0 2,000 Yes 14 8 20
Note

gn4, compute optimized family with GPU capabilities

Features
  • I/O optimized
  • Supports standard SSDs and ultra disks
  • Uses NVIDIA M40 GPU processors
  • Multiple CPU-to-memory ratios
  • Equipped with 2.5 GHz Intel ® Xeon ® E5-2682 v4 (Broadwell) processors
  • Provides strong network performance with large computing capacity
  • Suitable for the following scenarios:
    • Deep learning
    • Scientific computing applications, such as fluid dynamics, finance, genomics, and environmental analysis
    • High-performance computing, rendering, multimedia encoding and decoding, and other server-side GPU compute workloads
Instance types
Instance type vCPUs Memory (GiB) Local storage (GiB) GPU GPU memory (GB) Bandwidth (Gbit/s) Packet forwarding rate (Kpps) IPv6 support NIC queues ENIs Private IP address of a single ENI
ecs.gn4-c4g1.xlarge 4 30.0 N/A 1 * NVIDIA M40 1 * 12 3.0 300 No 1 3 10
ecs.gn4-c8g1.2xlarge 8 30.0 N/A 1 * NVIDIA M40 1 * 12 3.0 400 No 1 4 10
ecs.gn4.8xlarge 32 48.0 N/A 1 * NVIDIA M40 1 * 12 6.0 800 No 3 8 20
ecs.gn4-c4g1.2xlarge 8 60.0 N/A 2 * NVIDIA M40 2 * 12 5.0 500 No 1 4 10
ecs.gn4-c8g1.4xlarge 16 60.0 N/A 2 * NVIDIA M40 2 * 12 5.0 500 No 1 8 20
ecs.gn4.14xlarge 56 96.0 N/A 2 * NVIDIA M40 2 * 12 10.0 1,200 No 4 8 20
Note

GPU-based compute-optimized instance family, ga1

Features:
  • Optimized I/O performance
  • Supports SSDs and ultra disks
  • AMD S7150 GPUs
  • 1:2.5 CPU-to-memory ratio
  • 2.5 GHz Intel ® Xeon ® E5-2682 v4 (Broadwell) processors
  • High-performance local NVMe SSDs
  • Higher network performance along with more computing capacity
  • Scenarios:
    • Rendering and multimedia encoding and decoding
    • Machine learning, high-performance computing, and high-performance databases
    • Other server-end workloads that require powerful concurrent floating-point compute capabilities
The following table describes the instance types that belong to the ga1 instance family and specifications of these instance types.
Instance type vCPUs Memory (GiB) Local storage (GiB) GPU GPU memory (GB) Bandwidth (Gbit/s, inbound or outbound) Packet forwarding rate (Kpps, inbound or outbound) Supports IPv6 NIC queues ENIs (including one primary ENI) Private IP addresses for a single ENI
ecs.ga1.xlarge 4 10.0 1 × 87 0.25 × AMD S7150 2 1.0 200 No 1 3 10
ecs.ga1.2xlarge 8 20.0 1 × 175 0.5 × AMD S7150 4 1.5 300 No 1 4 10
ecs.ga1.4xlarge 16 40.0 1 × 350 1 × AMD S7150 8 3.0 500 No 2 8 20
ecs.ga1.8xlarge 32 80.0 1 × 700 2 × AMD S7150 2 × 8 6.0 800 No 3 8 20
ecs.ga1.14xlarge 56 160.0 1 × 1,400 4 × AMD S7150 4 × 8 10.0 1,200 No 4 8 20
Note

FPGA-based compute-optimized instance family, f1

Features:
  • Optimized I/O performance
  • Supports SSDs and ultra disks
  • Intel ® Arria ® 10 GX 1150 FPGAs
  • 1:7.5 CPU-to-memory ratio
  • 2.5 GHz Intel ® Xeon ® E5-2682 v4 (Broadwell) processors
  • Higher network performance along with more computing capacity
  • Scenarios:
    • Deep learning and reasoning
    • Genomics research
    • Financial analysis
    • Image transcoding
    • Computational workloads such as real-time video processing and security management
The following table describes the instance types that belong to the f1 instance family and specifications of these instance types.
Instance type vCPUs Memory (GiB) Local storage (GiB) FPGAs Bandwidth (Gbit/s, inbound or outbound) Packet forwarding rate (Kpps, inbound or outbound) Supports IPv6 NIC queues ENIs (including one primary ENI) Private IP addresses for a single ENI
ecs.f1-c8f1.2xlarge 8 60.0 None Intel Arria 10 GX 1150 3.0 400 Yes 4 4 10
ecs.f1-c8f1.4xlarge 16 120.0 None 2 × Intel Arria 10 GX 1150 5.0 1,000 Yes 4 8 20
ecs.f1-c28f1.7xlarge 28 112.0 None Intel Arria 10 GX 1150 5.0 2,000 Yes 8 8 20
ecs.f1-c28f1.14xlarge 56 224.0 None 2 × Intel Arria 10 GX 1150 10.0 2,000 Yes 14 8 20
Note

ebmgn6e, compute optimized ECS Bare Metal Instance family with GPU capabilities

Features
  • Provides flexible and powerful software-defined compute based on the X-Dragon architecture
  • I/O optimized
  • Supports enhanced SSDs, standard SSDs, and ultra disks
  • Uses NVIDIA V100 (32 GB NVLink) GPU processors
  • CPU-to-memory ratio of 1:4
  • Equipped with 2.5 GHz Intel ® Xeon ® Platinum 8163 (Skylake) processors
  • Uses NVIDIA V100 GPU computing accelerators (SXM2-based)
    • Powered by the innovative NVIDIA Volta architecture
    • 32 GB of HBM2 memory (900 GB/s bandwidth)
    • 5,120 CUDA Cores
    • 640 Tensor Cores
    • Supports up to six NVLink connections for a total bandwidth of 300 GB/s (25 GB/s per connection)
  • Provides strong network performance based on large computing capacity
  • Suitable for the following scenarios:
    • Deep learning applications such as training and inference applications of AI algorithms used in image classification, autonomous vehicles, and speech recognition
    • Scientific computing applications such as computational fluid dynamics, computational finance, molecular dynamics, and environmental analysis
Instance types
Instance type vCPUs Memory (GiB) Local storage (GiB) GPUs GPU memory (GB) Bandwidth (Gbit/s) Packet forwarding rate (Kpps) IPv6 support NIC queues ENIs (including one primary ENI) Private IP addresses per ENI
ecs.ebmgn6e.24xlarge 96 768.0 None 8 × V100 256 32.0 4,800 Yes 16 8 20
Note

ebmgn6v, compute optimized ECS Bare Metal Instance family with GPU capabilities

Features
  • Provides flexible and powerful software-defined compute based on the X-Dragon architecture
  • I/O optimized
  • Supports enhanced SSDs, standard SSDs, and ultra disks
  • Uses NVIDIA V100 GPU processors
  • CPU-to-memory ratio of 1:4
  • Equipped with 2.5 GHz Intel ® Xeon ® Platinum 8163 (Skylake) processors
  • Uses NVIDIA V100 GPU computing accelerators (SXM2-based)
    • Powered by the innovative NVIDIA Volta architecture
    • 16 GB of HBM2 memory (900 GB/s bandwidth)
    • 5,120 CUDA Cores
    • 640 Tensor Cores
    • Supports up to six NVLink connections for a total bandwidth of 300 GB/s (25 GB/s per connection)
  • Provides strong network performance based on large computing capacity
  • Suitable for the following scenarios:
    • Deep learning applications such as training and inference applications of AI algorithms used in image classification, autonomous vehicles, and speech recognition
    • Scientific computing applications such as computational fluid dynamics, computational finance, molecular dynamics, and environmental analysis
Instance types
Instance type vCPUs Memory (GiB) Local storage (GiB) GPUs GPU memory (GB) Bandwidth (Gbit/s) Packet forwarding rate (Kpps) IPv6 support NIC queues ENIs (including one primary ENI) Private IP addresses per ENI
ecs.ebmgn6v.24xlarge 96 384.0 None 8 × V100 128 30.0 4,500 Yes 8 32 10
Note

ebmgn6i, compute optimized ECS Bare Metal Instance family with GPU capabilities

Features
  • Powered by flexible and powerful software-defined compute based on the X-Dragon architecture
  • I/O optimized
  • CPU-to-memory ratio of 1:4
  • Equipped with 2.5 GHz Intel ® Xeon ® Platinum 8163 (Skylake) processors
  • Supports enhanced SSDs (million-level IOPS), standard SSDs, and ultra disks
  • Uses NVIDIA T4 GPU computing accelerators:
    • Powered by the new NVIDIA Turing architecture
    • 16 GB memory capacity (320 GB/s bandwidth)
    • 2,560 CUDA Cores
    • Up to 320 Turing Tensor Cores
    • Mixed-precision Tensor Cores support 65 FP16 TFLOPS, 130 INT8 TOPS, and 260 INT4 TOPS
  • Provides strong network performance based on large computing capacity
  • Suitable for the following scenarios:
    • AI (deep learning and machine learning) inference for computer vision, speech recognition, speech synthesis, natural language processing (NLP), machine translation, and recommendation systems
    • Real-time rendering for cloud gaming
    • Real-time rendering for AR and VR applications
    • Graphics workstations or overloaded graphics computing
    • GPU-accelerated databases
    • High-performance computing
Instance types
Instance type vCPU Memory (GiB) Local storage (GiB) GPU GPU memory (GB) Bandwidth (Gbit/s) Packet forwarding rate (Kpps) IPv6 support NIC queue ENI (including one primary ENI) Private IP address for a single ENI
ecs.ebmgn6i.24xlarge 96 384.0 N/A 4 × T4 64 30.0 4,500 Yes 8 32 10
Note

ebmc6, compute optimized ECS Bare Metal Instance family

Features
  • I/O optimized
  • Supports enhanced SSDs, standard SSDs, and ultra disks
  • CPU-to-memory ratio of 1:1.8
  • Equipped with 2.5 GHz Intel ® Xeon ® Platinum 8269 (Cascade Lake) processors, 104 vCPUs, up to 3.2 GHz Turbo Boost
  • High network performance: 6,000 Kpps packet forwarding rate
  • Supports VPCs only
  • Provides dedicated hardware resources and physical isolation
  • Suitable for the following scenarios:
    • Scenarios where large volumes of packets are received and transmitted, such as on-screen video comments and telecom data forwarding
    • Workloads that require direct access to physical resources or scenarios that require a license to be bound to the hardware
    • Third-party virtualization (including but not limited to Xen and KVM), and AnyStack (including but not limited to OpenStack and ZStack)
    • Containers (including but not limited to Docker, Clear Containers, and Pouch)
    • Video encoding, decoding, and rendering
    • Massively Multiplayer Online (MMO) game frontends
    • High-performance scientific and engineering applications
Instance types
Instance type vCPU Memory (GiB) Local storage (GiB) Bandwidth (Gbit/s) Packet forwarding rate (Kpps) IPv6 support NIC queue ENI (including one primary ENI) Private IP address for a single ENI
ecs.ebmc6.26xlarge 104 192.0 N/A 30.0 6,000 Yes 8 32 10
Note

ebmg6, general purpose ECS Bare Metal Instance family

Features
  • I/O optimized
  • Supports enhanced SSDs, standard SSDs, and ultra disks
  • CPU-to-memory ratio of 1:3.7
  • Equipped with 2.5 GHz Intel ® Xeon ® Platinum 8269 (Cascade Lake) processors, 104 vCPUs, up to 3.2 GHz Turbo Boost
  • High network performance: 6,000 Kpps packet forwarding rate
  • Supports VPCs only
  • Provides dedicated hardware resources and physical isolation
  • Suitable for the following scenarios:
    • Workloads that require direct access to physical resources or scenarios that require a license to be bound to the hardware
    • Third-party virtualization (including but not limited to Xen and KVM), and AnyStack (including but not limited to OpenStack and ZStack)
    • Containers (including but not limited to Docker, Clear Containers, and Pouch)
    • Video encoding, decoding, and rendering
    • Enterprise-level applications, such as large and medium-sized databases
    • Computing clusters and memory-intensive data processing
    • Data analysis and computing
Instance types
Instance type vCPU Memory (GiB) Local storage (GiB) Bandwidth (Gbit/s) Packet forwarding rate (Kpps) IPv6 support NIC queue ENI (including one primary ENI) Private IP address for a single ENI
ecs.ebmg6.26xlarge 104 384.0 N/A 30.0 6,000 Yes 8 32 10
Note

ebmr6, memory optimized ECS Bare Metal Instance family

Features
  • I/O optimized
  • Supports enhanced SSDs, standard SSDs, and ultra disks
  • CPU-to-memory ratio of 1:7.4
  • Equipped with 2.5 GHz Intel ® Xeon ® Platinum 8269 (Cascade Lake) processors, 104 vCPUs, up to 3.2 GHz Turbo Boost
  • High network performance: 6,000 Kpps packet forwarding rate
  • Supports VPCs only
  • Provides dedicated hardware resources and physical isolation
  • Suitable for the following scenarios:
    • Workloads that require direct access to physical resources or scenarios that require a license to be bound to the hardware
    • Third-party virtualization (including but not limited to Xen and KVM), and AnyStack (including but not limited to OpenStack and ZStack)
    • Containers (including but not limited to Docker, Clear Containers, and Pouch)
    • High-performance databases and in-memory databases
    • Data analysis and mining, and distributed memory cache
    • Hadoop, Spark, and other memory-intensive enterprise applications
Instance types
Instance type vCPU Memory (GiB) Local storage (GiB) Bandwidth (Gbit/s) Packet forwarding rate (Kpps) IPv6 support NIC queue ENI (including one primary ENI) Private IP address for a single ENI
ecs.ebmr6.26xlarge 104 768.0 N/A 30.0 6,000 Yes 8 32 10
Note

ebmhfc6, compute optimized ECS Bare Metal Instance family with high clock speed

Features
  • I/O optimized
  • Supports enhanced SSDs, standard SSDs, and ultra disks
  • CPU-to-memory ratio of 1:2.4
  • Equipped with 3.1 GHz Intel ® Xeon ® Platinum 8269 (Cascade Lake) processors, 80 vCPUs, up to 3.5 GHz Turbo Boost
  • High network performance: 6,000 Kpps packet forwarding rate
  • Supports VPCs only
  • Provides dedicated hardware resources and physical isolation
  • Suitable for the following scenarios:
    • Scenarios where large volumes of packets are received and transmitted, such as on-screen video comments and telecom data forwarding
    • Workloads that require direct access to physical resources or scenarios that require a license to be bound to the hardware
    • Third-party virtualization (including but not limited to Xen and KVM), and AnyStack (including but not limited to OpenStack and ZStack)
    • Containers (including but not limited to Docker, Clear Containers, and Pouch)
    • Video encoding, decoding, and rendering
Instance types
Instance type vCPU Memory (GiB) Local storage (GiB) Bandwidth (Gbit/s) Packet forwarding rate (Kpps) IPv6 support NIC queue ENI (including one primary ENI) Private IP address for a single ENI
ecs.ebmhfc6.20xlarge 80 192.0 N/A 30.0 6,000 Yes 8 32 10
Note

ebmhfg6, general purpose ECS Bare Metal Instance family with high clock speed

Features
  • I/O optimized
  • Supports enhanced SSDs, standard SSDs, and ultra disks
  • CPU-to-memory ratio of 1:4.8
  • Equipped with 3.1 GHz Intel ® Xeon ® Platinum 8269 (Cascade Lake) processors, 80 vCPUs, up to 3.5 GHz Turbo Boost
  • High network performance: 6,000 Kpps packet forwarding rate
  • Supports VPCs only
  • Provides dedicated hardware resources and physical isolation
  • Suitable for the following scenarios:
    • Workloads that require direct access to physical resources or scenarios that require a license to be bound to the hardware
    • Third-party virtualization (including but not limited to Xen and KVM), and AnyStack (including but not limited to OpenStack and ZStack)
    • Containers (including but not limited to Docker, Clear Containers, and Pouch)
    • Enterprise-level applications, such as large and medium-sized databases
    • Video encoding, decoding, and rendering
Instance types
Instance type vCPU Memory (GiB) Local storage (GiB) Bandwidth (Gbit/s) Packet forwarding rate (Kpps) IPv6 support NIC queue ENI (including one primary ENI) Private IP address for a single ENI
ecs.ebmhfg6.20xlarge 80 384.0 N/A 30.0 6,000 Yes 8 32 10
Note

ebmhfr6, memory optimized ECS Bare Metal Instance family with high clock speed

Features
  • I/O optimized
  • Supports enhanced SSDs, standard SSDs, and ultra disks
  • CPU-to-memory ratio of 1:9.6
  • Equipped with 3.1 GHz Intel ® Xeon ® Platinum 8269 (Cascade Lake) processors, 80 vCPUs, up to 3.5 GHz Turbo Boost
  • High network performance: 6,000 Kpps packet forwarding rate
  • Supports VPCs only
  • Provides dedicated hardware resources and physical isolation
  • Suitable for the following scenarios:
    • Workloads that require direct access to physical resources or scenarios that require a license to be bound to the hardware
    • Third-party virtualization (including but not limited to Xen and KVM), and AnyStack (including but not limited to OpenStack and ZStack)
    • Containers (including but not limited to Docker, Clear Containers, and Pouch)
    • High-performance databases and in-memory databases
    • Data analysis and mining, and distributed memory cache
    • Hadoop, Spark, and other memory-intensive enterprise applications
Instance types
Instance type vCPU Memory (GiB) Local storage (GiB) Bandwidth (Gbit/s) Packet forwarding rate (Kpps) IPv6 support NIC queue ENI (including one primary ENI) Private IP address for a single ENI
ecs.ebmhfr6.20xlarge 80 768.0 N/A 30.0 6,000 Yes 8 32 10
Note

ebmhfg5, ECS Bare Metal Instance family with high clock speed

Features
  • I/O optimized
  • Supports standard SSDs and ultra disks
  • CPU-to-memory ratio of 1:4
  • Equipped with 3.7 GHz Intel ® Xeon ® E3-1240 v6 (Skylake) processors, 8 vCPUs, up to 4.1 GHz Turbo Boost
  • High network performance: 2,000 Kpps packet forwarding rate
  • Supports VPCs only
  • Provides dedicated hardware resources and physical isolation
  • Does not support automatic recovery
  • Supports Intel ® SGX
  • Suitable for the following scenarios:
    • Workloads that require direct access to physical resources or scenarios that require a license to be bound to the hardware
    • Gaming or finance applications requiring high performance
    • High-performance Web servers
    • Enterprise-level applications such as high-performance databases
Instance types
Instance type vCPU Memory (GiB) Local storage (GiB) Bandwidth (Gbit/s) Packet forwarding rate (Kpps) IPv6 support NIC queue ENI (including one primary ENI) Private IP address for a single ENI
ecs.ebmhfg5.2xlarge 8 32.0 N/A 6.0 2,000 No 8 6 8
Note

ebmc4, compute optimized ECS Bare Metal Instance family

Features
  • I/O optimized
  • Supports standard SSDs and ultra disks
  • CPU-to-memory ratio of 1:2
  • Equipped with 2.5 GHz Intel ® Xeon ® E5-2682 v4 (Broadwell) processors, up to 2.9 GHz Turbo Boost
  • High network performance: 4,000 Kpps packet forwarding rate
  • Supports VPCs only
  • Provides dedicated hardware resources and physical isolation
  • Suitable for the following scenarios:
    • Workloads that require direct access to physical resources or scenarios that require a license to be bound to the hardware
    • Third-party virtualization (including but not limited to Xen and KVM), and AnyStack (including but not limited to OpenStack and ZStack)
    • Containers (including but not limited to Docker, Clear Containers, and Pouch)
    • Enterprise-level applications, such as large and medium-sized databases
    • Video encoding
Instance types
Instance type vCPU Memory (GiB) Local storage (GiB) Bandwidth (Gbit/s) Packet forwarding rate (Kpps) IPv6 support NIC queue ENI (including one primary ENI) Private IP address for a single ENI
ecs.ebmc4.8xlarge 32 64.0 N/A 10.0 4,000 No 8 12 10
Note

scch5, SCC instance family with high clock speed

Features
  • I/O optimized
  • Supports standard SSDs and ultra disks
  • Supports both RoCE and VPCs, of which RoCE is dedicated to RDMA communication
  • Has all features of ECS Bare Metal Instances
  • Equipped with 3.1 GHz Intel ® Xeon ® Gold 6149 (Skylake)
  • CPU-to-memory ratio of 1:3
  • Suitable for the following scenarios:
    • Large-scale machine learning training
    • Large-scale high performance scientific computing and simulations
    • Large-scale data analysis, batch computing, and video encoding
Instance types
Instance type vCPUs Physical cores Memory (GiB) GPUs Bandwidth (Gbit/s) Packet forwarding rate (Kpps) RoCE (Gbit/s) IPv6 support NIC queues ENIs (including one primary ENI) Private IP addresses per ENI
ecs.scch5.16xlarge 64 32 192.0 N/A 10.0 4,500 2 × 25 No 8 32 10
Note

sccg5, general purpose SCC instance family

Features
  • I/O optimized
  • Supports standard SSDs and ultra disks
  • Supports both RoCE and VPCs, of which RoCE is dedicated to RDMA communication
  • Has all features of ECS Bare Metal Instances
  • Equipped with 2.5 GHz Intel ® Xeon ® Platinum 8163 (Skylake) processors
  • CPU-to-memory ratio of 1:4
  • Suitable for the following scenarios:
    • Large-scale machine learning training
    • Large-scale high performance scientific computing and simulations
    • Large-scale data analysis, batch computing, and video encoding
Instance types
Instance type vCPUs Physical cores Memory (GiB) GPUs Bandwidth (Gbit/s) Packet forwarding rate (Kpps) RoCE (Gbit/s) IPv6 support NIC queues ENIs (including one primary ENI) Private IP addresses per ENI
ecs.sccg5.24xlarge 96 48 384.0 N/A 10.0 4,500 2 × 25 No 8 32 10
Note

sccgn6, compute optimized SCC GPU instance family

Features
  • I/O optimized
  • CPU-to-memory ratio of 1:4
  • Equipped with 2.5 GHz Intel ® Xeon ® Platinum 8163 (Skylake) processors
  • Has all features of ECS Bare Metal Instances
  • Storage:
    • Supports enhanced SSDs, standard SSDs, and ultra disks
    • Supports high performance Cloud Parallel File System (CPFS)
  • Networking:
    • Supports VPCs
    • Supports RoCE v2 networks, which is dedicated to RDMA communication
  • Uses NVIDIA V100 GPU processors (with the SXM2 module):
    • Based on the new NVIDIA Volta architecture
    • 16 GB HBM2 GPU memory
    • 5,120 CUDA Cores
    • 640 Tensor Cores
    • Memory bandwidth of 900 GB/s
    • Supports up to six NVLink connections and total bandwidth of 300 GB/s (25 GB/s per connection)
  • Suitable for the following scenarios:
    • Ultra-large-scale machine learning training on a distributed GPU cluster
    • Large-scale high performance scientific computing and simulations
    • Large-scale data analysis, batch computing, and video encoding
Instance types
Instance type vCPUs Memory (GiB) Local storage (GiB) GPUs Bandwidth (Gbit/s) Packet forwarding rate (Kpps) RoCE (Gbit/s) IPv6 support NIC queues ENIs (including one primary ENI) Private IP addresses per ENI
ecs.sccgn6.24xlarge 96 384 N/A 8 × V100 30 4,500 2 × 25 Yes 8 32 10
Note

ebmc5s, compute optimized ECS Bare Metal Instance family with enhanced network performance

Features
  • I/O optimized
  • Supports enhanced SSDs, standard SSDs, and ultra disks
  • CPU-to-memory ratio of 1:2
  • Equipped with 2.5 GHz Intel ® Xeon ® Platinum 8163 (Skylake) processors, 96 vCPUs, up to 2.7 GHz Turbo Boost
  • High network performance: 4,500 Kpps packet forwarding rate
  • Supports VPCs only
  • Provides dedicated hardware resources and physical isolation
  • Suitable for the following scenarios:
    • Scenarios where large volumes of packets are received and transmitted, such as on-screen video comments and telecom data forwarding
    • Workloads that require direct access to physical resources or scenarios that require a license to be bound to the hardware
    • Third-party virtualization (including but not limited to Xen and KVM), and AnyStack (including but not limited to OpenStack and ZStack)
    • Containers (including but not limited to Docker, Clear Containers, and Pouch)
    • Video encoding, decoding, and rendering
Instance types
Instance type vCPU Memory (GiB) Local storage (GiB) Bandwidth (Gbit/s) Packet forwarding rate (Kpps) IPv6 support NIC queue ENI (including one primary ENI) Private IP address for a single ENI
ecs.ebmc5s.24xlarge 96 192.0 N/A 30.0 4,500 No 8 32 10
Note

ebmg5s, general purpose ECS Bare Metal Instance family with enhanced network performance

Features
  • I/O optimized
  • Supports enhanced SSDs, standard SSDs, and ultra disks
  • CPU-to-memory ratio of 1:4
  • Equipped with 2.5 GHz Intel ® Xeon ® Platinum 8163 (Skylake) processors, 96 vCPUs, up to 2.7 GHz Turbo Boost
  • High network performance: 4,500 Kpps packet forwarding rate
  • Supports VPCs only
  • Provides dedicated hardware resources and physical isolation
  • Suitable for the following scenarios:
    • Workloads that require direct access to physical resources or scenarios that require a license to be bound to the hardware
    • Third-party virtualization (including but not limited to Xen and KVM), and AnyStack (including but not limited to OpenStack and ZStack)
    • Containers (including but not limited to Docker, Clear Containers, and Pouch)
    • Enterprise-level applications, such as large and medium-sized databases
    • Video encoding
Instance types
Instance type vCPU Memory (GiB) Local storage (GiB) Bandwidth (Gbit/s) Packet forwarding rate (Kpps) IPv6 support NIC queue ENI (including one primary ENI) Private IP address for a single ENI
ecs.ebmg5s.24xlarge 96 384.0 N/A 30.0 4,500 No 8 32 10
Note

ebmr5s, memory optimized ECS Bare Metal Instance family with enhanced network performance

Features
  • I/O optimized
  • Supports enhanced SSDs, standard SSDs, and ultra disks
  • CPU-to-memory ratio of 1:8
  • Equipped with 2.5 GHz Intel ® Xeon ® Platinum 8163 (Skylake) processors, 96 vCPUs, up to 2.7 GHz Turbo Boost
  • High network performance: 4,500 Kpps packet forwarding rate
  • Supports VPCs only
  • Provides dedicated hardware resources and physical isolation
  • Suitable for the following scenarios:
    • Workloads that require direct access to physical resources or scenarios that require a license to be bound to the hardware
    • Third-party virtualization (including but not limited to Xen and KVM), and AnyStack (including but not limited to OpenStack and ZStack)
    • Containers (including but not limited to Docker, Clear Containers, and Pouch)
    • High-performance databases and in-memory databases
    • Data analysis and mining, and distributed memory cache
    • Hadoop, Spark, and other memory-intensive enterprise applications
Instance types
Instance type vCPU Memory (GiB) Local storage (GiB) Bandwidth (Gbit/s) Packet forwarding rate (Kpps) IPv6 support NIC queue ENI (including one primary ENI) Private IP address for a single ENI
ecs.ebmr5s.24xlarge 96 768.0 N/A 30.0 4,500 No 8 32 10
Note