General purpose instances - Amazon Elastic Compute Cloud
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.

General purpose instances

General purpose instances provide a balance of compute, memory, and networking resources, and can be used for a wide range of workloads.

M5 and M5a instances

These instances provide an ideal cloud infrastructure, offering a balance of compute, memory, and networking resources for a broad range of applications that are deployed in the cloud. They are well-suited for the following:

  • Small and midsize databases

  • Data processing tasks that require additional memory

  • Caching fleets

  • Backend servers for SAP, Microsoft SharePoint, cluster computing, and other enterprise applications

Bare metal instances, such as m5.metal, m5n.metal, and m5zn.metal provide your applications with direct access to physical resources of the host server, such as processors and memory.

M5zn

These instances are ideal for applications that benefit from extremely high single-thread performance, high throughput, and low latency networking. They are well-suited for the following:

  • Gaming

  • High performance computing

  • Simulation modeling

M6i and M6id instances

These instances are well suited for general-purpose workloads such as the following:

  • Application servers and web servers

  • Microservices

  • High performance computing

  • App development

  • Small and midsize databases

  • Caching fleets

Bare metal instances such as m6i.metal provide your applications with direct access to physical resources of the host server, such as processors and memory.

M6in and M6idn instances

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

  • High-performance file systems

  • Distributed web scale in-memory caches

  • Caching fleets

  • Real-time big data analytics

  • Telco applications such as 5G User Plane Function (UPF)

For more information, see Amazon EC2 M6i Instances.

T2, T3, and T3a instances

These instances provide a baseline level of CPU performance with the ability to burst to a higher level when required by your workload. An Unlimited instance can sustain high CPU performance for any period of time whenever required. For more information, see Burstable performance instances. They are well-suited for the following:

  • Websites and web applications

  • Code repositories

  • Development, build, test, and staging environments

  • Microservices

Hardware specifications

The following is a summary of the hardware specifications for general purpose 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)
m1.small 1 1.70
m1.medium 1 3.70
m1.large 2 7.50
m1.xlarge 4 15.00
m2.xlarge 2 17.10
m2.2xlarge 4 34.20
m2.4xlarge 8 68.40
m3.medium 1 3.75
m3.large 2 7.50
m3.xlarge 4 15.00
m3.2xlarge 8 30.00
m4.large 2 8.00
m4.xlarge 4 16.00
m4.2xlarge 8 32.00
m4.4xlarge 16 64.00
m4.10xlarge 40 160.00
m4.16xlarge 64 256.00
m5.large 2 8.00
m5.xlarge 4 16.00
m5.2xlarge 8 32.00
m5.4xlarge 16 64.00
m5.8xlarge 32 128.00
m5.12xlarge 48 192.00
m5.16xlarge 64 256.00
m5.24xlarge 96 384.00
m5.metal 96 384.00
m5a.large 2 8.00
m5a.xlarge 4 16.00
m5a.2xlarge 8 32.00
m5a.4xlarge 16 64.00
m5a.8xlarge 32 128.00
m5a.12xlarge 48 192.00
m5a.16xlarge 64 256.00
m5a.24xlarge 96 384.00
m5ad.large 2 8.00
m5ad.xlarge 4 16.00
m5ad.2xlarge 8 32.00
m5ad.4xlarge 16 64.00
m5ad.8xlarge 32 128.00
m5ad.12xlarge 48 192.00
m5ad.16xlarge 64 256.00
m5ad.24xlarge 96 384.00
m5d.large 2 8.00
m5d.xlarge 4 16.00
m5d.2xlarge 8 32.00
m5d.4xlarge 16 64.00
m5d.8xlarge 32 128.00
m5d.12xlarge 48 192.00
m5d.16xlarge 64 256.00
m5d.24xlarge 96 384.00
m5d.metal 96 384.00
m5dn.large 2 8.00
m5dn.xlarge 4 16.00
m5dn.2xlarge 8 32.00
m5dn.4xlarge 16 64.00
m5dn.8xlarge 32 128.00
m5dn.12xlarge 48 192.00
m5dn.16xlarge 64 256.00
m5dn.24xlarge 96 384.00
m5dn.metal 96 384.00
m5n.large 2 8.00
m5n.xlarge 4 16.00
m5n.2xlarge 8 32.00
m5n.4xlarge 16 64.00
m5n.8xlarge 32 128.00
m5n.12xlarge 48 192.00
m5n.16xlarge 64 256.00
m5n.24xlarge 96 384.00
m5n.metal 96 384.00
m5zn.large 2 8.00
m5zn.xlarge 4 16.00
m5zn.2xlarge 8 32.00
m5zn.3xlarge 12 48.00
m5zn.6xlarge 24 96.00
m5zn.12xlarge 48 192.00
m5zn.metal 48 192.00
m6a.large 2 8.00
m6a.xlarge 4 16.00
m6a.2xlarge 8 32.00
m6a.4xlarge 16 64.00
m6a.8xlarge 32 128.00
m6a.12xlarge 48 192.00
m6a.16xlarge 64 256.00
m6a.24xlarge 96 384.00
m6a.32xlarge 128 512.00
m6a.48xlarge 192 768.00
m6a.metal 192 768.00
m6i.large 2 8.00
m6i.xlarge 4 16.00
m6i.2xlarge 8 32.00
m6i.4xlarge 16 64.00
m6i.8xlarge 32 128.00
m6i.12xlarge 48 192.00
m6i.16xlarge 64 256.00
m6i.24xlarge 96 384.00
m6i.32xlarge 128 512.00
m6i.metal 128 512.00
m6id.large 2 8.00
m6id.xlarge 4 16.00
m6id.2xlarge 8 32.00
m6id.4xlarge 16 64.00
m6id.8xlarge 32 128.00
m6id.12xlarge 48 192.00
m6id.16xlarge 64 256.00
m6id.24xlarge 96 384.00
m6id.32xlarge 128 512.00
m6id.metal 128 512.00
m6idn.large 2 8.00
m6idn.xlarge 4 16.00
m6idn.2xlarge 8 32.00
m6idn.4xlarge 16 64.00
m6idn.8xlarge 32 128.00
m6idn.12xlarge 48 192.00
m6idn.16xlarge 64 256.00
m6idn.24xlarge 96 384.00
m6idn.32xlarge 128 512.00
m6in.large 2 8.00
m6in.xlarge 4 16.00
m6in.2xlarge 8 32.00
m6in.4xlarge 16 64.00
m6in.8xlarge 32 128.00
m6in.12xlarge 48 192.00
m6in.16xlarge 64 256.00
m6in.24xlarge 96 384.00
m6in.32xlarge 128 512.00
t1.micro 1 0.61
t2.nano 1 0.50
t2.micro 1 1.00
t2.small 1 2.00
t2.medium 2 4.00
t2.large 2 8.00
t2.xlarge 4 16.00
t2.2xlarge 8 32.00
t3.nano 2 0.50
t3.micro 2 1.00
t3.small 2 2.00
t3.medium 2 4.00
t3.large 2 8.00
t3.xlarge 4 16.00
t3.2xlarge 8 32.00
t3a.nano 2 0.50
t3a.micro 2 1.00
t3a.small 2 2.00
t3a.medium 2 4.00
t3a.large 2 8.00
t3a.xlarge 4 16.00
t3a.2xlarge 8 32.00

The general purpose instances use the following processors.

AMD processors
  • AMD EPYC 7000 series processors (AMD EPYC 7571): M5a, M5ad, T3a

  • 3rd generation AMD EPYC processors (AMD EPYC 7R13): M6a

Intel processors
  • Intel Xeon Scalable processors (Haswell E5-2676 v3 or Broadwell E5-2686 v4): M4, T2

  • Intel Xeon Scalable processors (Skylake 8175M or Cascade Lake 8259CL): M5, M5d, T3

  • 2nd generation Intel Xeon Scalable processors (Cascade Lake 8259CL): M5n

  • 2nd generation Intel Xeon Scalable processors (Cascade Lake 8252C): M5zn

  • 3rd generation Intel Xeon Scalable processors (Ice Lake 8375C): M6i, M6id

For more information, see Amazon EC2 Instance Types.

Instance performance

EBS-optimized instances enable you to get consistently high performance for your EBS volumes by eliminating contention between Amazon EBS I/O and other network traffic from your instance. Some general purpose instances are EBS-optimized by default at no additional cost. For more information, see Amazon EBS–optimized instances.

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 general purpose instances that support enhanced networking.

Instance type Baseline bandwidth (Gbps) Burst bandwidth (Gbps) Enhanced networking features
m1.small 0.3 1.2 Not supported
m1.medium 0.3 1.2 Not supported
m1.large 0.7 2.8 Not supported
m1.xlarge 1.0 10.0 Not supported
m2.xlarge 0.3 1.2 Not supported
m2.2xlarge 0.7 2.8 Not supported
m2.4xlarge 1.0 10.0 Not supported
m3.medium 0.3 1.2 Not supported
m3.large 0.7 2.8 Not supported
m3.xlarge 1.0 10.0 Not supported
m3.2xlarge 1.0 10.0 Not supported
m4.large 0.45 1.2 Not supported
m4.xlarge 0.75 2.8 Not supported
m4.2xlarge 1.0 10.0 Not supported
m4.4xlarge 2.0 10.0 Not supported
m4.10xlarge 5.0 10.0 Not supported
m4.16xlarge 5.0 10.0 ENA
m5.large 0.75 10.0 ENA
m5.xlarge 1.25 10.0 ENA
m5.2xlarge 2.5 10.0 ENA
m5.4xlarge 5.0 10.0 ENA
m5.8xlarge 10.0 - ENA
m5.12xlarge 12.0 - ENA
m5.16xlarge 20.0 - ENA
m5.24xlarge 25.0 - ENA
m5.metal 25.0 - ENA
m5a.large 0.75 10.0 ENA
m5a.xlarge 1.25 10.0 ENA
m5a.2xlarge 2.5 10.0 ENA
m5a.4xlarge 5.0 10.0 ENA
m5a.8xlarge 7.5 10.0 ENA
m5a.12xlarge 10.0 - ENA
m5a.16xlarge 12.0 - ENA
m5a.24xlarge 20.0 - ENA
m5ad.large 0.75 10.0 ENA
m5ad.xlarge 1.25 10.0 ENA
m5ad.2xlarge 2.5 10.0 ENA
m5ad.4xlarge 5.0 10.0 ENA
m5ad.8xlarge 7.5 10.0 ENA
m5ad.12xlarge 10.0 - ENA
m5ad.16xlarge 12.0 - ENA
m5ad.24xlarge 20.0 - ENA
m5d.large 0.75 10.0 ENA
m5d.xlarge 1.25 10.0 ENA
m5d.2xlarge 2.5 10.0 ENA
m5d.4xlarge 5.0 10.0 ENA
m5d.8xlarge 10.0 - ENA
m5d.12xlarge 12.0 - ENA
m5d.16xlarge 20.0 - ENA
m5d.24xlarge 25.0 - ENA
m5d.metal 25.0 - ENA
m5dn.large 2.1 25.0 ENA
m5dn.xlarge 4.1 25.0 ENA
m5dn.2xlarge 8.125 25.0 ENA
m5dn.4xlarge 16.25 25.0 ENA
m5dn.8xlarge 25.0 - ENA
m5dn.12xlarge 50.0 - ENA
m5dn.16xlarge 75.0 - ENA
m5dn.24xlarge 100.0 - ENA | EFA
m5dn.metal 100.0 - ENA | EFA
m5n.large 2.1 25.0 ENA
m5n.xlarge 4.1 25.0 ENA
m5n.2xlarge 8.125 25.0 ENA
m5n.4xlarge 16.25 25.0 ENA
m5n.8xlarge 25.0 - ENA
m5n.12xlarge 50.0 - ENA
m5n.16xlarge 75.0 - ENA
m5n.24xlarge 100.0 - ENA | EFA
m5n.metal 100.0 - ENA | EFA
m5zn.large 3.0 25.0 ENA
m5zn.xlarge 5.0 25.0 ENA
m5zn.2xlarge 10.0 25.0 ENA
m5zn.3xlarge 15.0 25.0 ENA
m5zn.6xlarge 50.0 - ENA
m5zn.12xlarge 100.0 - ENA | EFA
m5zn.metal 100.0 - ENA | EFA
m6a.large 0.781 12.5 ENA
m6a.xlarge 1.562 12.5 ENA
m6a.2xlarge 3.125 12.5 ENA
m6a.4xlarge 6.25 12.5 ENA
m6a.8xlarge 12.5 - ENA
m6a.12xlarge 18.75 - ENA
m6a.16xlarge 25.0 - ENA
m6a.24xlarge 37.5 - ENA
m6a.32xlarge 50.0 - ENA
m6a.48xlarge 50.0 - ENA | EFA
m6a.metal 50.0 - ENA | EFA
m6i.large 0.781 12.5 ENA
m6i.xlarge 1.562 12.5 ENA
m6i.2xlarge 3.125 12.5 ENA
m6i.4xlarge 6.25 12.5 ENA
m6i.8xlarge 12.5 - ENA
m6i.12xlarge 18.75 - ENA
m6i.16xlarge 25.0 - ENA
m6i.24xlarge 37.5 - ENA
m6i.32xlarge 50.0 - ENA | EFA
m6i.metal 50.0 - ENA | EFA
m6id.large 0.781 12.5 ENA
m6id.xlarge 1.562 12.5 ENA
m6id.2xlarge 3.125 12.5 ENA
m6id.4xlarge 6.25 12.5 ENA
m6id.8xlarge 12.5 - ENA
m6id.12xlarge 18.75 - ENA
m6id.16xlarge 25.0 - ENA
m6id.24xlarge 37.5 - ENA
m6id.32xlarge 50.0 - ENA | EFA
m6id.metal 50.0 - ENA | EFA
m6idn.large 3.125 25.0 ENA
m6idn.xlarge 6.25 30.0 ENA
m6idn.2xlarge 12.5 40.0 ENA
m6idn.4xlarge 25.0 50.0 ENA
m6idn.8xlarge 50.0 - ENA
m6idn.12xlarge 75.0 - ENA
m6idn.16xlarge 100.0 - ENA
m6idn.24xlarge 150.0 - ENA
m6idn.32xlarge 400.0 - ENA | EFA
m6in.large 3.125 25.0 ENA
m6in.xlarge 6.25 30.0 ENA
m6in.2xlarge 12.5 40.0 ENA
m6in.4xlarge 25.0 50.0 ENA
m6in.8xlarge 50.0 - ENA
m6in.12xlarge 75.0 - ENA
m6in.16xlarge 100.0 - ENA
m6in.24xlarge 150.0 - ENA
m6in.32xlarge 400.0 - ENA | EFA
t1.micro 0.07 0.28 Not supported
t2.nano 0.032 0.512 Not supported
t2.micro 0.064 1.024 Not supported
t2.small 0.128 1.024 Not supported
t2.medium 0.256 1.024 Not supported
t2.large 0.512 1.024 Not supported
t2.xlarge 0.75 1.024 Not supported
t2.2xlarge 1.0 1.024 Not supported
t3.nano 0.032 5.0 ENA
t3.micro 0.064 5.0 ENA
t3.small 0.128 5.0 ENA
t3.medium 0.256 5.0 ENA
t3.large 0.512 5.0 ENA
t3.xlarge 1.024 5.0 ENA
t3.2xlarge 2.048 5.0 ENA
t3a.nano 0.032 5.0 ENA
t3a.micro 0.064 5.0 ENA
t3a.small 0.128 5.0 ENA
t3a.medium 0.256 5.0 ENA
t3a.large 0.512 5.0 ENA
t3a.xlarge 1.024 5.0 ENA
t3a.2xlarge 2.048 5.0 ENA

† These instances 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.

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.

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
m1.small 14400 4800
m1.medium 28800 8000
m1.large 57600 16000
m1.xlarge 115200 32000
m2.xlarge 36800 4800
m2.2xlarge 75200 11200
m2.4xlarge 166400 21440
m3.medium 9600 6400
m3.large 19200 12800
m3.xlarge 38400 22400
m3.2xlarge 73600 51200
m5ad.large 30000 15000
m5ad.xlarge 59000 29000
m5ad.2xlarge 117000 57000
m5ad.4xlarge 234000 114000
m5ad.8xlarge 466666 233334
m5ad.12xlarge 700000 340000
m5ad.16xlarge 933332 466668
m5ad.24xlarge 1400000 680000
m5d.large 30000 15000
m5d.xlarge 59000 29000
m5d.2xlarge 117000 57000
m5d.4xlarge 234000 114000
m5d.8xlarge 466666 233334
m5d.12xlarge 700000 340000
m5d.16xlarge 933332 466668
m5d.24xlarge 1400000 680000
m5d.metal 1400000 680000
m5dn.large 29000 14500
m5dn.xlarge 58000 29000
m5dn.2xlarge 116000 58000
m5dn.4xlarge 232000 116000
m5dn.8xlarge 464000 232000
m5dn.12xlarge 700000 350000
m5dn.16xlarge 930000 465000
m5dn.24xlarge 1400000 700000
m5dn.metal 1400000 700000
m6id.large 33542 16771
m6id.xlarge 134167 67084
m6id.2xlarge 134167 45000
m6id.4xlarge 268333 134167
m6id.8xlarge 536666 268334
m6id.12xlarge 804998 402500
m6id.16xlarge 1073332 536668
m6id.24xlarge 1609996 805000
m6id.32xlarge 2146664 1073336
m6id.metal 2146664 1073336
m6idn.large 33542 16771
m6idn.xlarge 67083 33542
m6idn.2xlarge 134167 67084
m6idn.4xlarge 268333 134167
m6idn.8xlarge 536666 268334
m6idn.12xlarge 804998 402500
m6idn.16xlarge 1073332 536668
m6idn.24xlarge 1609996 805000
m6idn.32xlarge 2146664 1073336

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.

Release notes

  • Instances built on the Nitro System, M4, t2.large and larger, t3.large and larger, and t3a.large and larger instance types require 64-bit HVM AMIs. They have high-memory, and require a 64-bit operating system to take advantage of that capacity. HVM AMIs provide superior performance in comparison to paravirtual (PV) AMIs on high-memory instance types. In addition, you must use an HVM AMI to take advantage of enhanced networking.

  • 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 M6i 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 support a maximum of 28 attachments, including network interfaces, EBS volumes, and NVMe instance store volumes. For more information, see Nitro System volume limits.

  • 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.

  • 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.