Memory optimized instances - Amazon Elastic Compute Cloud
Hardware specificationsMemory performanceInstance performanceNetwork performanceAmazon EBS I/O performanceSSD-based instance store volume I/O performanceHigh availability and reliability (X1)Support for vCPUsRelease notes
Services or capabilities described in Amazon Web Services documentation might vary by Region. To see the differences applicable to the China Regions, see Getting Started with Amazon Web Services in China (PDF).

Memory optimized instances

Memory optimized instances are designed to deliver fast performance for workloads that process large data sets in memory.

R5, R5a, R5b, and R5n instances

These instances are well suited for the following:

  • High-performance, including relational MySQL and NoSQL, for example MongoDB and Cassandra databases.

  • Distributed web scale cache stores that provide in-memory caching of key-value type data, for example Memcached and Redis

  • In-memory databases using optimized data storage formats and analytics for business intelligence; for example, SAP HANA

  • Applications performing real-time processing of big unstructured data, using Hadoop and Spark clusters

  • High-performance computing (HPC) and Electronic Design Automation (EDA) applications.

Bare metal instances, such as r5.metal, provide your applications with direct access to physical resources of the host server, such as processors and memory.

R6a instances

These instances are ideal for running memory-intensive workloads, such as the following:

  • High-performance databases, both relational and NoSQL

  • Distributed web scale in-memory caches, such as Memcached and Redis

  • Real-time big data analytics, such as Hadoop and Spark clusters

Hpc6id instances

These instances are ideal for running high performance computing (HPC) workloads, such as the following:

  • Seismic and Reservoir

  • Crash Simulation

  • Finite Element Analysis

R6i and R6id instances

These instances are ideal for running memory-intensive workloads, such as the following:

  • High-performance databases, relational and NoSQL

  • In-memory databases, for example SAP HANA

  • Distributed web scale in-memory caches, for example Memcached and Redis

  • Real-time big data analytics, including Hadoop and Spark clusters

R6in and R6idn instances

These instances are well suited for network-intensive workloads such as the following:

  • High-performance relational, MySQL and NoSQL. For example, MongoDB and Cassandra databases

  • Distributed web scale cache stores that provide in-memory caching of key-value type data, including Memcached and Redis

  • In-memory databases using optimized data storage formats and analytics for business intelligence, for example SAP HANA

  • Real-time big data analytics for financial services, for example Hadoop and Spark clusters

For more information, see Amazon EC2 R6i Instances.

High memory (u-*) instances

These instances offer 3 TiB, 6 TiB, 9 TiB, 12 TiB, 18 TiB, and 24 TiB of memory per instance. They are designed to run large in-memory databases, including production deployments of the SAP HANA in-memory database.

X1 instances

These instances are well suited for the following:

  • In-memory databases such as SAP HANA, including SAP-certified support for Business Suite S/4HANA, Business Suite on HANA (SoH), Business Warehouse on HANA (BW), and Data Mart Solutions on HANA. For more information, see SAP HANA on the Amazon Cloud.

  • Big-data processing engines such as Apache Spark or Presto.

  • High-performance computing (HPC) applications.

X1e instances

These instances are well suited for the following:

  • High-performance databases.

  • In-memory databases such as SAP HANA. For more information, see SAP HANA on the Amazon Cloud.

  • Memory-intensive enterprise applications.

X2idn, X2iedn, and X2iezn instances

These instances are well suited for the following:

  • In-memory databases, such as Redis and Memcached.

  • Relational databases, such as MySQL and PostGreSQL.

  • Electronic design automation (EDA) workloads, such as physical verification and layout tools.

  • Memory-intensive workloads, such as real-time analytics and real-time caching servers.

X2idn and X2iedn instances support io2 Block Express volumes. All io2 volumes attached to X2idn and X2iedn instances, during or after launch, automatically run on EBS Block Express. For more information, see io2 Block Express volumes.

z1d instances

These instances deliver both high compute and high memory and are well-suited for the following:

  • Electronic Design Automation (EDA)

  • Relational database workloads

z1d.metal instances provide your applications with direct access to physical resources of the host server, such as processors and memory.

Hardware specifications

The following is a summary of the hardware specifications for memory optimized instances. A virtual central processing unit (vCPU) represents a portion of the physical CPU assigned to a virtual machine (VM). For x86 instances, there are two vCPUs per core. For Graviton instances, there is one vCPU per core.

Instance type Default vCPUs Memory (GiB)
cr1.8xlarge 32 243.76
hpc6id.32xlarge 64 1024.00
r3.large 2 15.00
r3.xlarge 4 30.50
r3.2xlarge 8 61.00
r3.4xlarge 16 122.00
r3.8xlarge 32 244.00
r4.large 2 15.25
r4.xlarge 4 30.50
r4.2xlarge 8 61.00
r4.4xlarge 16 122.00
r4.8xlarge 32 244.00
r4.16xlarge 64 488.00
r5.large 2 16.00
r5.xlarge 4 32.00
r5.2xlarge 8 64.00
r5.4xlarge 16 128.00
r5.8xlarge 32 256.00
r5.12xlarge 48 384.00
r5.16xlarge 64 512.00
r5.24xlarge 96 768.00
r5.metal 96 768.00
r5a.large 2 16.00
r5a.xlarge 4 32.00
r5a.2xlarge 8 64.00
r5a.4xlarge 16 128.00
r5a.8xlarge 32 256.00
r5a.12xlarge 48 384.00
r5a.16xlarge 64 512.00
r5a.24xlarge 96 768.00
r5ad.large 2 16.00
r5ad.xlarge 4 32.00
r5ad.2xlarge 8 64.00
r5ad.4xlarge 16 128.00
r5ad.8xlarge 32 256.00
r5ad.12xlarge 48 384.00
r5ad.16xlarge 64 512.00
r5ad.24xlarge 96 768.00
r5b.large 2 16.00
r5b.xlarge 4 32.00
r5b.2xlarge 8 64.00
r5b.4xlarge 16 128.00
r5b.8xlarge 32 256.00
r5b.12xlarge 48 384.00
r5b.16xlarge 64 512.00
r5b.24xlarge 96 768.00
r5b.metal 96 768.00
r5d.large 2 16.00
r5d.xlarge 4 32.00
r5d.2xlarge 8 64.00
r5d.4xlarge 16 128.00
r5d.8xlarge 32 256.00
r5d.12xlarge 48 384.00
r5d.16xlarge 64 512.00
r5d.24xlarge 96 768.00
r5d.metal 96 768.00
r5dn.large 2 16.00
r5dn.xlarge 4 32.00
r5dn.2xlarge 8 64.00
r5dn.4xlarge 16 128.00
r5dn.8xlarge 32 256.00
r5dn.12xlarge 48 384.00
r5dn.16xlarge 64 512.00
r5dn.24xlarge 96 768.00
r5dn.metal 96 768.00
r5n.large 2 16.00
r5n.xlarge 4 32.00
r5n.2xlarge 8 64.00
r5n.4xlarge 16 128.00
r5n.8xlarge 32 256.00
r5n.12xlarge 48 384.00
r5n.16xlarge 64 512.00
r5n.24xlarge 96 768.00
r5n.metal 96 768.00
r6a.large 2 16.00
r6a.xlarge 4 32.00
r6a.2xlarge 8 64.00
r6a.4xlarge 16 128.00
r6a.8xlarge 32 256.00
r6a.12xlarge 48 384.00
r6a.16xlarge 64 512.00
r6a.24xlarge 96 768.00
r6a.32xlarge 128 1024.00
r6a.48xlarge 192 1536.00
r6a.metal 192 1536.00
r6i.large 2 16.00
r6i.xlarge 4 32.00
r6i.2xlarge 8 64.00
r6i.4xlarge 16 128.00
r6i.8xlarge 32 256.00
r6i.12xlarge 48 384.00
r6i.16xlarge 64 512.00
r6i.24xlarge 96 768.00
r6i.32xlarge 128 1024.00
r6i.metal 128 1024.00
r6idn.large 2 16.00
r6idn.xlarge 4 32.00
r6idn.2xlarge 8 64.00
r6idn.4xlarge 16 128.00
r6idn.8xlarge 32 256.00
r6idn.12xlarge 48 384.00
r6idn.16xlarge 64 512.00
r6idn.24xlarge 96 768.00
r6idn.32xlarge 128 1024.00
r6idn.metal 128 1024.00
r6in.large 2 16.00
r6in.xlarge 4 32.00
r6in.2xlarge 8 64.00
r6in.4xlarge 16 128.00
r6in.8xlarge 32 256.00
r6in.12xlarge 48 384.00
r6in.16xlarge 64 512.00
r6in.24xlarge 96 768.00
r6in.32xlarge 128 1024.00
r6in.metal 128 1024.00
r6id.large 2 16.00
r6id.xlarge 4 32.00
r6id.2xlarge 8 64.00
r6id.4xlarge 16 128.00
r6id.8xlarge 32 256.00
r6id.12xlarge 48 384.00
r6id.16xlarge 64 512.00
r6id.24xlarge 96 768.00
r6id.32xlarge 128 1024.00
r6id.metal 128 1024.00
u-3tb1.56xlarge 224 3072.00
u-6tb1.56xlarge 224 6144.00
u-6tb1.112xlarge 448 6144.00
u-6tb1.metal 448 6144.00
u-9tb1.112xlarge 448 9216.00
u-9tb1.metal 448 9216.00
u-12tb1.112xlarge 448 12288.00
u-12tb1.metal 448 12288.00
u-18tb1.metal 448 18432.00
u-24tb1.metal 448 24576.00
x1.16xlarge 64 976.00
x1.32xlarge 128 1952.00
x2idn.16xlarge 64 1024.00
x2idn.24xlarge 96 1536.00
x2idn.32xlarge 128 2048.00
x2idn.metal 128 2048.00
x2iedn.xlarge 4 128.00
x2iedn.2xlarge 8 256.00
x2iedn.4xlarge 16 512.00
x2iedn.8xlarge 32 1024.00
x2iedn.16xlarge 64 2048.00
x2iedn.24xlarge 96 3072.00
x2iedn.32xlarge 128 4096.00
x2iedn.metal 128 4096.00
x2iezn.2xlarge 8 256.00
x2iezn.4xlarge 16 512.00
x2iezn.6xlarge 24 768.00
x2iezn.8xlarge 32 1024.00
x2iezn.12xlarge 48 1536.00
x2iezn.metal 48 1536.00
x1e.xlarge 4 122.00
x1e.2xlarge 8 244.00
x1e.4xlarge 16 488.00
x1e.8xlarge 32 976.00
x1e.16xlarge 64 1952.00
x1e.32xlarge 128 3904.00
z1d.large 2 16.00
z1d.xlarge 4 32.00
z1d.2xlarge 8 64.00
z1d.3xlarge 12 96.00
z1d.6xlarge 24 192.00
z1d.12xlarge 48 384.00
z1d.metal 48 384.00

* Each logical processor is a hyperthread on 224 cores.

The memory optimized instances use the following processors.

AMD processors

  • AMD EPYC 7000 series processors (AMD EPYC 7571): R5a, R5ad

  • 3rd generation AMD EPYC processors (AMD EPYC 7R13): R6a

Intel processors

  • Intel Xeon Scalable processors (Haswell E7-8880 v3): X1, X1e

  • Intel Xeon Scalable processors (Broadwell E5-2686 v4): R4

  • Intel Xeon Scalable processors (Skylake 8151): z1d

  • Intel Xeon Scalable processors (Skylake 8175M or Cascade Lake 8259CL): R5, R5d

  • 2nd generation Intel Xeon Scalable processors (Cascade Lake 8259CL): R5b, R5n

  • 2nd generation Intel Xeon Scalable processors (Cascade Lake 8252C): X2iezn

  • 3rd generation Intel Xeon Scalable processors (Ice Lake 8375C): R6i, R6id, X2idn, X2iedn

For more information, see Amazon EC2 Instance Types.

Memory performance

X1 instances include Intel Scalable Memory Buffers, providing 300 GiB/s of sustainable memory-read bandwidth and 140 GiB/s of sustainable memory-write bandwidth.

For more information about how much RAM can be enabled for memory optimized instances, see Hardware specifications.

Memory optimized instances have high memory and require 64-bit HVM AMIs to take advantage of that capacity. HVM AMIs provide superior performance in comparison to paravirtual (PV) AMIs on memory optimized instances. .

Instance performance

Memory optimized instances enable increased cryptographic performance through the latest Intel AES-NI feature and support Advanced Vector Extensions 2 (Intel AVX2) processor instructions to expand most integer commands to 256 bits.

Network performance

You can enable enhanced networking on supported instance types to provide lower latencies, lower network jitter, and higher packet-per-second (PPS) performance. Most applications do not consistently need a high level of network performance, but can benefit from access to increased bandwidth when they send or receive data. For more information, see Enhanced networking on Windows.

The following is a summary of network performance for memory optimized instances that support enhanced networking.

Note

Instance types indicated with a † have a baseline bandwidth and can use a network I/O credit mechanism to burst beyond their baseline bandwidth on a best effort basis. For more information, see instance network bandwidth.

Instance type Network performance Enhanced networking features
hpc6id.32xlarge 200 Gigabit ENA | EFA
r3.large † Moderate Not supported
r3.xlarge † Moderate Not supported
r3.2xlarge † High Not supported
r3.4xlarge † High Not supported
r3.8xlarge † 10 Gigabit Not supported
r4.large † Up to 10 Gigabit ENA
r4.xlarge † Up to 10 Gigabit ENA
r4.2xlarge † Up to 10 Gigabit ENA
r4.4xlarge † Up to 10 Gigabit ENA
r4.8xlarge 10 Gigabit ENA
r4.16xlarge 25 Gigabit ENA
r5.large † Up to 10 Gigabit ENA
r5.xlarge † Up to 10 Gigabit ENA
r5.2xlarge † Up to 10 Gigabit ENA
r5.4xlarge † Up to 10 Gigabit ENA
r5.8xlarge 10 Gigabit ENA
r5.12xlarge 12 Gigabit ENA
r5.16xlarge 20 Gigabit ENA
r5.24xlarge 25 Gigabit ENA
r5.metal 25 Gigabit ENA
r5a.large † Up to 10 Gigabit ENA
r5a.xlarge † Up to 10 Gigabit ENA
r5a.2xlarge † Up to 10 Gigabit ENA
r5a.4xlarge † Up to 10 Gigabit ENA
r5a.8xlarge † Up to 10 Gigabit ENA
r5a.12xlarge 10 Gigabit ENA
r5a.16xlarge 12 Gigabit ENA
r5a.24xlarge 20 Gigabit ENA
r5ad.large † Up to 10 Gigabit ENA
r5ad.xlarge † Up to 10 Gigabit ENA
r5ad.2xlarge † Up to 10 Gigabit ENA
r5ad.4xlarge † Up to 10 Gigabit ENA
r5ad.8xlarge † Up to 10 Gigabit ENA
r5ad.12xlarge 10 Gigabit ENA
r5ad.16xlarge 12 Gigabit ENA
r5ad.24xlarge 20 Gigabit ENA
r5b.large † Up to 10 Gigabit ENA
r5b.xlarge † Up to 10 Gigabit ENA
r5b.2xlarge † Up to 10 Gigabit ENA
r5b.4xlarge † Up to 10 Gigabit ENA
r5b.8xlarge 10 Gigabit ENA
r5b.12xlarge 12 Gigabit ENA
r5b.16xlarge 20 Gigabit ENA
r5b.24xlarge 25 Gigabit ENA
r5b.metal 25 Gigabit ENA
r5d.large † Up to 10 Gigabit ENA
r5d.xlarge † Up to 10 Gigabit ENA
r5d.2xlarge † Up to 10 Gigabit ENA
r5d.4xlarge † Up to 10 Gigabit ENA
r5d.8xlarge 10 Gigabit ENA
r5d.12xlarge 12 Gigabit ENA
r5d.16xlarge 20 Gigabit ENA
r5d.24xlarge 25 Gigabit ENA
r5d.metal 25 Gigabit ENA
r5dn.large † Up to 25 Gigabit ENA
r5dn.xlarge † Up to 25 Gigabit ENA
r5dn.2xlarge † Up to 25 Gigabit ENA
r5dn.4xlarge † Up to 25 Gigabit ENA
r5dn.8xlarge 25 Gigabit ENA
r5dn.12xlarge 50 Gigabit ENA
r5dn.16xlarge 75 Gigabit ENA
r5dn.24xlarge 100 Gigabit ENA | EFA
r5dn.metal 100 Gigabit ENA | EFA
r5n.large † Up to 25 Gigabit ENA
r5n.xlarge † Up to 25 Gigabit ENA
r5n.2xlarge † Up to 25 Gigabit ENA
r5n.4xlarge † Up to 25 Gigabit ENA
r5n.8xlarge 25 Gigabit ENA
r5n.12xlarge 50 Gigabit ENA
r5n.16xlarge 75 Gigabit ENA
r5n.24xlarge 100 Gigabit ENA | EFA
r5n.metal 100 Gigabit ENA | EFA
r6a.large † Up to 12.5 Gigabit ENA
r6a.xlarge † Up to 12.5 Gigabit ENA
r6a.2xlarge † Up to 12.5 Gigabit ENA
r6a.4xlarge † Up to 12.5 Gigabit ENA
r6a.8xlarge 12.5 Gigabit ENA
r6a.12xlarge 18.75 Gigabit ENA
r6a.16xlarge 25 Gigabit ENA
r6a.24xlarge 37.5 Gigabit ENA
r6a.32xlarge 50 Gigabit ENA
r6a.48xlarge 50 Gigabit ENA | EFA
r6a.metal 50 Gigabit ENA | EFA
r6i.large † Up to 12.5 Gigabit ENA
r6i.xlarge † Up to 12.5 Gigabit ENA
r6i.2xlarge † Up to 12.5 Gigabit ENA
r6i.4xlarge † Up to 12.5 Gigabit ENA
r6i.8xlarge 12.5 Gigabit ENA
r6i.12xlarge 18.75 Gigabit ENA
r6i.16xlarge 25 Gigabit ENA
r6i.24xlarge 37.5 Gigabit ENA
r6i.32xlarge 50 Gigabit ENA | EFA
r6i.metal 50 Gigabit ENA | EFA
r6idn.large † Up to 25 Gigabit ENA
r6idn.xlarge † Up to 30 Gigabit ENA
r6idn.2xlarge † Up to 40 Gigabit ENA
r6idn.4xlarge † Up to 50 Gigabit ENA
r6idn.8xlarge 50 Gigabit ENA
r6idn.12xlarge 75 Gigabit ENA
r6idn.16xlarge 100 Gigabit ENA
r6idn.24xlarge 150 Gigabit ENA
r6idn.32xlarge 200 Gigabit ENA | EFA
r6idn.metal 200 Gigabit ENA | EFA
r6in.large † Up to 25 Gigabit ENA
r6in.xlarge † Up to 30 Gigabit ENA
r6in.2xlarge † Up to 40 Gigabit ENA
r6in.4xlarge † Up to 50 Gigabit ENA
r6in.8xlarge 50 Gigabit ENA
r6in.12xlarge 75 Gigabit ENA
r6in.16xlarge 100 Gigabit ENA
r6in.24xlarge 150 Gigabit ENA
r6in.32xlarge 200 Gigabit ENA | EFA
r6in.metal 200 Gigabit ENA | EFA
r6id.large † Up to 12.5 Gigabit ENA
r6id.xlarge † Up to 12.5 Gigabit ENA
r6id.2xlarge † Up to 12.5 Gigabit ENA
r6id.4xlarge † Up to 12.5 Gigabit ENA
r6id.8xlarge 12.5 Gigabit ENA
r6id.12xlarge 18.75 Gigabit ENA
r6id.16xlarge 25 Gigabit ENA
r6id.24xlarge 37.5 Gigabit ENA
r6id.32xlarge 50 Gigabit ENA | EFA
r6id.metal 50 Gigabit ENA | EFA
u-3tb1.56xlarge 50 Gigabit ENA
u-6tb1.56xlarge 100 Gigabit ENA
u-6tb1.112xlarge 100 Gigabit ENA
u-6tb1.metal 100 ENA
u-9tb1.112xlarge 100 Gigabit ENA
u-9tb1.metal 100 ENA
u-12tb1.112xlarge 100 Gigabit ENA
u-12tb1.metal 100 ENA
u-18tb1.metal 100 Gigabit ENA
u-24tb1.metal 100 Gigabit ENA
x1.16xlarge 10 Gigabit ENA
x1.32xlarge 25 Gigabit ENA
x2idn.16xlarge 50 Gigabit ENA
x2idn.24xlarge 75 Gigabit ENA
x2idn.32xlarge 100 Gigabit ENA | EFA
x2idn.metal 100 Gigabit ENA | EFA
x2iedn.xlarge † Up to 25 Gigabit ENA
x2iedn.2xlarge † Up to 25 Gigabit ENA
x2iedn.4xlarge † Up to 25 Gigabit ENA
x2iedn.8xlarge 25 Gigabit ENA
x2iedn.16xlarge 50 Gigabit ENA
x2iedn.24xlarge 75 Gigabit ENA
x2iedn.32xlarge 100 Gigabit ENA | EFA
x2iedn.metal 100 Gigabit ENA | EFA
x2iezn.2xlarge † Up to 25 Gigabit ENA
x2iezn.4xlarge † Up to 25 Gigabit ENA
x2iezn.6xlarge 50 Gigabit ENA
x2iezn.8xlarge 75 Gigabit ENA
x2iezn.12xlarge 100 Gigabit ENA | EFA
x2iezn.metal 100 Gigabit ENA | EFA
x1e.xlarge † Up to 10 Gigabit ENA
x1e.2xlarge † Up to 10 Gigabit ENA
x1e.4xlarge † Up to 10 Gigabit ENA
x1e.8xlarge † Up to 10 Gigabit ENA
x1e.16xlarge 10 Gigabit ENA
x1e.32xlarge 25 Gigabit ENA
z1d.large † Up to 10 Gigabit ENA
z1d.xlarge † Up to 10 Gigabit ENA
z1d.2xlarge † Up to 10 Gigabit ENA
z1d.3xlarge † Up to 10 Gigabit ENA
z1d.6xlarge 12 Gigabit ENA
z1d.12xlarge 25 Gigabit ENA
z1d.metal 25 Gigabit ENA

For 32xlarge and metal instance types that support 200 Gbps, at least 2 ENIs, each attached to a different network card, are required on the instance to achieve 200 Gbps throughput. Each ENI attached to a network card can achieve a max of 170 Gbps.

u-6tb1.metal, u-9tb1.metal, and u-12tb1.metal instances launched after March 12, 2020 provide network performance of 100 Gbps. u-6tb1.metal, u-9tb1.metal, and u-12tb1.metal instances launched before March 12, 2020 mightonly provide network performance of 25 Gbps. To ensure that instances launched before March 12, 2020 have a network performance of 100 Gbps, contact your account team to upgrade your instance at no additional cost.

The following table shows the baseline and burst bandwidth for instance types that use the network I/O credit mechanism to burst beyond their baseline bandwidth.

Instance type Baseline bandwidth (Gbps) Burst bandwidth (Gbps)
r3.large 0.5 1.2
r3.xlarge 0.7 2.8
r3.2xlarge 1.0 10.0
r3.4xlarge 2.0 10.0
r3.8xlarge 5.0 10.0
r4.large 0.75 10.0
r4.xlarge 1.25 10.0
r4.2xlarge 2.5 10.0
r4.4xlarge 5.0 10.0
r5.large 0.75 10.0
r5.xlarge 1.25 10.0
r5.2xlarge 2.5 10.0
r5.4xlarge 5.0 10.0
r5a.large 0.75 10.0
r5a.xlarge 1.25 10.0
r5a.2xlarge 2.5 10.0
r5a.4xlarge 5.0 10.0
r5a.8xlarge 7.5 10.0
r5ad.large 0.75 10.0
r5ad.xlarge 1.25 10.0
r5ad.2xlarge 2.5 10.0
r5ad.4xlarge 5.0 10.0
r5ad.8xlarge 7.5 10.0
r5b.large 0.75 10.0
r5b.xlarge 1.25 10.0
r5b.2xlarge 2.5 10.0
r5b.4xlarge 5.0 10.0
r5d.large 0.75 10.0
r5d.xlarge 1.25 10.0
r5d.2xlarge 2.5 10.0
r5d.4xlarge 5.0 10.0
r5dn.large 2.1 25.0
r5dn.xlarge 4.1 25.0
r5dn.2xlarge 8.125 25.0
r5dn.4xlarge 16.25 25.0
r5n.large 2.1 25.0
r5n.xlarge 4.1 25.0
r5n.2xlarge 8.125 25.0
r5n.4xlarge 16.25 25.0
r6a.large 0.781 12.5
r6a.xlarge 1.562 12.5
r6a.2xlarge 3.125 12.5
r6a.4xlarge 6.25 12.5
r6i.large 0.781 12.5
r6i.xlarge 1.562 12.5
r6i.2xlarge 3.125 12.5
r6i.4xlarge 6.25 12.5
r6idn.large 3.125 25.0
r6idn.xlarge 6.25 30.0
r6idn.2xlarge 12.5 40.0
r6idn.4xlarge 25.0 50.0
r6in.large 3.125 25.0
r6in.xlarge 6.25 30.0
r6in.2xlarge 12.5 40.0
r6in.4xlarge 25.0 50.0
r6id.large 0.781 12.5
r6id.xlarge 1.562 12.5
r6id.2xlarge 3.125 12.5
r6id.4xlarge 6.25 12.5
x2iedn.xlarge 1.875 25.0
x2iedn.2xlarge 5.0 25.0
x2iedn.4xlarge 12.5 25.0
x2iezn.2xlarge 12.5 25.0
x2iezn.4xlarge 15.0 25.0
x1e.xlarge 0.625 10.0
x1e.2xlarge 1.25 10.0
x1e.4xlarge 2.5 10.0
x1e.8xlarge 5.0 10.0
z1d.large 0.75 10.0
z1d.xlarge 1.25 10.0
z1d.2xlarge 2.5 10.0
z1d.3xlarge 5.0 10.0

Amazon EBS I/O performance

Amazon EBS optimized instances use an optimized configuration stack and provide additional, dedicated capacity for Amazon EBS I/O. This optimization provides the best performance for your Amazon EBS volumes by minimizing contention between Amazon EBS I/O and other traffic from your instance.

For more information, see Amazon EBS–optimized instances.

SSD-based instance store volume I/O performance

If you use all the SSD-based instance store volumes available to your instance, you can get up to the IOPS (4,096 byte block size) performance listed in the following table (at queue depth saturation). Otherwise, you get lower IOPS performance.

Instance Size 100% Random Read IOPS Write IOPS
hpc6id.32xlarge 2,146,664 1,073,336
r5ad.large 30,000 15,000
r5ad.xlarge 59,000 29,000
r5ad.2xlarge 117,000 57,000
r5ad.4xlarge 234,000 114,000
r5ad.8xlarge 466,666 233,333
r5ad.12xlarge 700,000 340,000
r5ad.16xlarge 933,333 466,666
r5ad.24xlarge 1,400,000 680,000
r5d.large 30,000 15,000
r5d.xlarge 59,000 29,000
r5d.2xlarge 117,000 57,000
r5d.4xlarge 234,000 114,000
r5d.8xlarge 466,666 233,333
r5d.12xlarge 700,000 340,000
r5d.16xlarge 933,333 466,666
r5d.24xlarge 1,400,000 680,000
r5d.metal 1,400,000 680,000
r5dn.large 30,000 15,000
r5dn.xlarge 59,000 29,000
r5dn.2xlarge 117,000 57,000
r5dn.4xlarge 234,000 114,000
r5dn.8xlarge 466,666 233,333
r5dn.12xlarge 700,000 340,000
r5dn.16xlarge 933,333 466,666
r5dn.24xlarge 1,400,000 680,000
r5dn.metal 1,400,000 680,000
r6id.metal 3,219,995 1,610,005
r6id.large 33,542 16,771
r6id.xlarge 67,083 33,542
r6id.2xlarge 134,167 67,084
r6id.4xlarge 268,333 134,167
r6id.8xlarge 536,666 268,334
r6id.12xlarge 804,999 402,501
r6id.16xlarge 1,073,332 536,668
r6id.24xlarge 1,609,998 805,002
r6id.32xlarge 2,146,664 1,073,336
r6id.metal 2,146,664 1,073,336
r6idn.large 33,542 16,771
r6idn.xlarge 67,083 33,542
r6idn.2xlarge 134,167 67,084
r6idn.4xlarge 268,333 134,167
r6idn.8xlarge 536,666 268,334
r6idn.12xlarge 804,999 402,501
r6idn.16xlarge 1,073,332 536,668
r6idn.24xlarge 1,609,998 805,002
r6idn.32xlarge 2,146,664 1,073,336
r6idn.metal 2,146,664 1,073,336
x2idn.16xlarge 430,000 180,000
x2idn.24xlarge 645,000 270,000
x2idn.32xlarge 860,000 360,000
x2idn.metal 860,000 360,000
x2iedn.xlarge 26,875 11,250
x2iedn.2xlarge 53,750 22,500
x2iedn.4xlarge 107,500 45,000
x2iedn.8xlarge 215,000 90,000
x2iedn.16xlarge 430,000 180,000
x2iedn.24xlarge 645,000 270,000
x2iedn.32xlarge 860,000 360,000
x2iedn.metal 860,000 360,000
z1d.large 30,000 15,000
z1d.xlarge 59,000 29,000
z1d.2xlarge 117,000 57,000
z1d.3xlarge 175,000 75,000
z1d.6xlarge 350,000 170,000
z1d.12xlarge 700,000 340,000
z1d.metal 700,000 340,000

As you fill the SSD-based instance store volumes for your instance, the number of write IOPS that you can achieve decreases. This is due to the extra work the SSD controller must do to find available space, rewrite existing data, and erase unused space so that it can be rewritten. This process of garbage collection results in internal write amplification to the SSD, expressed as the ratio of SSD write operations to user write operations. This decrease in performance is even larger if the write operations are not in multiples of 4,096 bytes or not aligned to a 4,096-byte boundary. If you write a smaller amount of bytes or bytes that are not aligned, the SSD controller must read the surrounding data and store the result in a new location. This pattern results in significantly increased write amplification, increased latency, and dramatically reduced I/O performance.

SSD controllers can use several strategies to reduce the impact of write amplification. One such strategy is to reserve space in the SSD instance storage so that the controller can more efficiently manage the space available for write operations. This is called over-provisioning. The SSD-based instance store volumes provided to an instance don't have any space reserved for over-provisioning. To reduce write amplification, we recommend that you leave 10% of the volume unpartitioned so that the SSD controller can use it for over-provisioning. This decreases the storage that you can use, but increases performance even if the disk is close to full capacity.

For instance store volumes that support TRIM, you can use the TRIM command to notify the SSD controller whenever you no longer need data that you've written. This provides the controller with more free space, which can reduce write amplification and increase performance. For more information, see Instance store volume TRIM support.

High availability and reliability (X1)

X1 instances support Single Device Data Correction (SDDC +1), which detects and corrects multi-bit errors. SDDC +1 uses error checking and correction code to identify and disable a failed single DRAM device.

In addition, you can implement high availability (HA) and disaster recovery (DR) solutions to meet recovery point objective (RPO), recovery time objective (RTO), and cost requirements by leveraging Amazon CloudFormation and Recover your instance.

If you run an SAP HANA production environment, you also have the option of using HANA System Replication (HSR) on X1 instances. For more information about architecting HA and DR solutions on X1 instances, see SAP HANA on the Amazon Web Services Cloud: Quick Start Reference Deployment.

Support for vCPUs

Memory optimized instances provide a high number of vCPUs, which can cause launch issues with operating systems that have a lower vCPU limit. We strongly recommend that you use the latest AMIs when you launch memory optimized instances.

The following AMIs support launching memory optimized instances:

  • Amazon Linux 2 (HVM)

  • Amazon Linux AMI 2016.03 (HVM) or later

  • Ubuntu Server 14.04 LTS (HVM)

  • Red Hat Enterprise Linux 7.1 (HVM)

  • SUSE Linux Enterprise Server 12 SP1 (HVM)

  • Windows Server 2019

  • Windows Server 2016

  • Windows Server 2012 R2

  • Windows Server 2012

  • Windows Server 2008 R2 64-bit

  • Windows Server 2008 SP2 64-bit

Release notes

  • Instances built on the Nitro System have the following requirements:

    The current Amazon Windows AMIs meet these requirements.

  • To get the best performance from your R6i instances, ensure that they have ENA driver version 2.2.3 or later. Using an ENA driver earlier than version 2.0.0 with these instances causes network interface attachment failures. The following AMIs have a compatible ENA driver.

    • Amazon Windows AMI from May 2021 or later

  • Instances built on the Nitro System instances support a maximum of 28 attachments, including network interfaces, EBS volumes, and NVMe instance store volumes. For more information, see Nitro System volume limits.

  • All io2 volumes attached to C6a, C6in, C7g, Inf2, M6a, M6in, M6idn, M7g, R5b, R6in, R6idn, R7g, Trn1, Trn1n, X2idn, and X2iedn instances, during or after launch, automatically run on EBS Block Express. For more information, see io2 Block Express volumes.

  • Launching a bare metal instance boots the underlying server, which includes verifying all hardware and firmware components. This means that it can take 20 minutes from the time the instance enters the running state until it becomes available over the network.

  • To attach or detach EBS volumes or secondary network interfaces from a bare metal instance requires PCIe native hotplug support.

  • Bare metal instances use a PCI-based serial device rather than an I/O port-based serial device. The upstream Linux kernel and the latest Amazon Linux AMIs support this device. Bare metal instances also provide an ACPI SPCR table to enable the system to automatically use the PCI-based serial device. The latest Windows AMIs automatically use the PCI-based serial device.

  • You can't launch X1 instances using a Windows Server 2008 SP2 64-bit AMI, except for x1.16xlarge instances.

  • You can't launch X1e instances using a Windows Server 2008 SP2 64-bit AMI.

  • With earlier versions of the Windows Server 2008 R2 64-bit AMI, you can't launch r4.large and r4.4xlarge instances. If you experience this issue, update to the latest version of this AMI.

  • There is a limit on the total number of instances that you can launch in a Region, and there are additional limits on some instance types. For more information, see How many instances can I run in Amazon EC2? in the Amazon EC2 FAQ.