Instance types - Amazon Elastic Compute Cloud

Instance types


For detailed instance type specifications, see the Amazon EC2 Instance Types Guide. For pricing information, see Amazon EC2 Instance Types.

When you launch an instance, the instance type that you specify determines the hardware of the host computer used for your instance. Each instance type offers different compute, memory, and storage capabilities, and is grouped in an instance family based on these capabilities. Select an instance type based on the requirements of the application or software that you plan to run on your instance.

Amazon EC2 dedicates some resources of the host computer, such as CPU, memory, and instance storage, to a particular instance. Amazon EC2 shares other resources of the host computer, such as the network and the disk subsystem, among instances. If each instance on a host computer tries to use as much of one of these shared resources as possible, each receives an equal share of that resource. However, when a resource is underused, an instance can consume a higher share of that resource while it's available.

Each instance type provides higher or lower minimum performance from a shared resource. For example, instance types with high I/O performance have a larger allocation of shared resources. Allocating a larger share of shared resources also reduces the variance of I/O performance. For most applications, moderate I/O performance is more than enough. However, for applications that require greater or more consistent I/O performance, consider an instance type with higher I/O performance.

Instance type naming convention

Amazon EC2 provides a variety of instance types so you can choose the type that best meets your requirements. Instance types are named based on their family, generation, processor family, additional capabilities, and size. The first position of the instance type name indicates the instance family, for example c. The second position indicates the instance generation, for example 7. The third position indicates the processor family, for example g. The remaining letters before the period indicate additional capabilities, such as instance store volumes. After the period (.) is the instance size, such as small or 4xlarge, or metal for bare metal instances.

            The image shows the instance type c7gn.xlarge, with a label for each part of the instance name.
Instance families
  • C – Compute optimized

  • D – Dense storage

  • F – FPGA

  • G – Graphics intensive

  • Hpc – High performance computing

  • I – Storage optimized

  • Im – Storage optimized with a one to four ratio of vCPU to memory

  • Is – Storage optimized with a one to six ratio of vCPU to memory

  • Inf – AWS Inferentia

  • M – General purpose

  • Mac – macOS

  • P – GPU accelerated

  • R – Memory optimized

  • T – Burstable performance

  • Trn – AWS Trainium

  • U – High memory

  • VT – Video transcoding

  • X – Memory intensive

Processor families
  • a – AMD processors

  • g – AWS Graviton processors

  • i – Intel processors

Additional capabilities
  • b – EBS optimized

  • d – Instance store volumes

  • n – Network and EBS optimized

  • e – Extra storage or memory

  • z – High performance

  • q – Qualcomm inference accelerators

  • flex – Flex instance

Available instance types

Amazon EC2 provides a wide selection of instance types optimized to fit different use cases. Instance types comprise varying combinations of CPU, memory, storage, and networking capacity and give you the flexibility to choose the appropriate mix of resources for your applications. Each instance type includes one or more instance sizes, allowing you to scale your resources to the requirements of your target workload.


Previous generation instances are still fully supported and retain the same features and functionality. We encourage you to use the latest generation of instances to get the best performance.

To determine which instance types meet your requirements, such as supported Regions, compute resources, or storage resources, see Find an Amazon EC2 instance type.

Current generation instances

For the best performance, we recommend that you use the following instance types when you launch new instances. For more information, see Amazon EC2 Instance Types.

Sixth generation and later Amazon EC2 instance types
  • General purpose: M6a, M6g, M6gd, M6i, M6id, M6idn, M6in, M7a, M7g, M7gd, M7i, M7i-flex, T4g

  • Compute optimized: C6a, C6g, C6gd, C6gn, C6i, C6id, C6in, C7a, C7g, C7gd, C7gn, C7i

  • Memory optimized: R6a, R6g, R6gd, R6i, R6id, R6idn, R6in, R7a, R7g, R7gd, R7i, R7iz, X2gd, X2idn, X2iedn

  • Storage optimized: I4g, I4i, Im4gn, Is4gen

  • Accelerated computing: DL2q, G5g, Inf2, P5, Trn1, Trn1n

  • High-performance computing: Hpc6a, Hpc6id, Hpc7a, Hpc7g

General purpose

Type Sizes
M5 m5.large | m5.xlarge | m5.2xlarge | m5.4xlarge | m5.8xlarge | m5.12xlarge | m5.16xlarge | m5.24xlarge | m5.metal
M5a m5a.large | m5a.xlarge | m5a.2xlarge | m5a.4xlarge | m5a.8xlarge | m5a.12xlarge | m5a.16xlarge | m5a.24xlarge
M5ad m5ad.large | m5ad.xlarge | m5ad.2xlarge | m5ad.4xlarge | m5ad.8xlarge | m5ad.12xlarge | m5ad.16xlarge | m5ad.24xlarge
M5d m5d.large | m5d.xlarge | m5d.2xlarge | m5d.4xlarge | m5d.8xlarge | m5d.12xlarge | m5d.16xlarge | m5d.24xlarge | m5d.metal
M5dn m5dn.large | m5dn.xlarge | m5dn.2xlarge | m5dn.4xlarge | m5dn.8xlarge | m5dn.12xlarge | m5dn.16xlarge | m5dn.24xlarge | m5dn.metal
M5n m5n.large | m5n.xlarge | m5n.2xlarge | m5n.4xlarge | m5n.8xlarge | m5n.12xlarge | m5n.16xlarge | m5n.24xlarge | m5n.metal
M5zn m5zn.large | m5zn.xlarge | m5zn.2xlarge | m5zn.3xlarge | m5zn.6xlarge | m5zn.12xlarge | m5zn.metal
M6a m6a.large | m6a.xlarge | m6a.2xlarge | m6a.4xlarge | m6a.8xlarge | m6a.12xlarge | m6a.16xlarge | m6a.24xlarge | m6a.32xlarge | m6a.48xlarge | m6a.metal
M6g m6g.medium | m6g.large | m6g.xlarge | m6g.2xlarge | m6g.4xlarge | m6g.8xlarge | m6g.12xlarge | m6g.16xlarge | m6g.metal
M6gd m6gd.medium | m6gd.large | m6gd.xlarge | m6gd.2xlarge | m6gd.4xlarge | m6gd.8xlarge | m6gd.12xlarge | m6gd.16xlarge | m6gd.metal
M6i m6i.large | m6i.xlarge | m6i.2xlarge | m6i.4xlarge | m6i.8xlarge | m6i.12xlarge | m6i.16xlarge | m6i.24xlarge | m6i.32xlarge | m6i.metal
M6id m6id.large | m6id.xlarge | m6id.2xlarge | m6id.4xlarge | m6id.8xlarge | m6id.12xlarge | m6id.16xlarge | m6id.24xlarge | m6id.32xlarge | m6id.metal
M6idn m6idn.large | m6idn.xlarge | m6idn.2xlarge | m6idn.4xlarge | m6idn.8xlarge | m6idn.12xlarge | m6idn.16xlarge | m6idn.24xlarge | m6idn.32xlarge | m6idn.metal
M6in m6in.large | m6in.xlarge | m6in.2xlarge | m6in.4xlarge | m6in.8xlarge | m6in.12xlarge | m6in.16xlarge | m6in.24xlarge | m6in.32xlarge | m6in.metal
M7a m7a.medium | m7a.large | m7a.xlarge | m7a.2xlarge | m7a.4xlarge | m7a.8xlarge | m7a.12xlarge | m7a.16xlarge | m7a.24xlarge | m7a.32xlarge | m7a.48xlarge | m7a.metal-48xl
M7g m7g.medium | m7g.large | m7g.xlarge | m7g.2xlarge | m7g.4xlarge | m7g.8xlarge | m7g.12xlarge | m7g.16xlarge | m7g.metal
M7gd m7gd.medium | m7gd.large | m7gd.xlarge | m7gd.2xlarge | m7gd.4xlarge | m7gd.8xlarge | m7gd.12xlarge | m7gd.16xlarge
M7i m7i.large | m7i.xlarge | m7i.2xlarge | m7i.4xlarge | m7i.8xlarge | m7i.12xlarge | m7i.16xlarge | m7i.24xlarge | m7i.48xlarge | m7i.metal-24xl | m7i.metal-48xl
M7i-flex m7i-flex.large | m7i-flex.xlarge | m7i-flex.2xlarge | m7i-flex.4xlarge | m7i-flex.8xlarge
Mac1 mac1.metal
Mac2 mac2.metal
Mac2-m2 mac2-m2.metal
Mac2-m2pro mac2-m2pro.metal
T2 t2.nano | t2.micro | t2.small | t2.medium | t2.large | t2.xlarge | t2.2xlarge
T3 t3.nano | t3.micro | t3.small | t3.medium | t3.large | t3.xlarge | t3.2xlarge
T3a t3a.nano | t3a.micro | t3a.small | t3a.medium | t3a.large | t3a.xlarge | t3a.2xlarge
T4g t4g.nano | t4g.micro | t4g.small | t4g.medium | t4g.large | t4g.xlarge | t4g.2xlarge

Compute optimized

Type Sizes
C5 c5.large | c5.xlarge | c5.2xlarge | c5.4xlarge | c5.9xlarge | c5.12xlarge | c5.18xlarge | c5.24xlarge | c5.metal
C5a c5a.large | c5a.xlarge | c5a.2xlarge | c5a.4xlarge | c5a.8xlarge | c5a.12xlarge | c5a.16xlarge | c5a.24xlarge
C5ad c5ad.large | c5ad.xlarge | c5ad.2xlarge | c5ad.4xlarge | c5ad.8xlarge | c5ad.12xlarge | c5ad.16xlarge | c5ad.24xlarge
C5d c5d.large | c5d.xlarge | c5d.2xlarge | c5d.4xlarge | c5d.9xlarge | c5d.12xlarge | c5d.18xlarge | c5d.24xlarge | c5d.metal
C5n c5n.large | c5n.xlarge | c5n.2xlarge | c5n.4xlarge | c5n.9xlarge | c5n.18xlarge | c5n.metal
C6a c6a.large | c6a.xlarge | c6a.2xlarge | c6a.4xlarge | c6a.8xlarge | c6a.12xlarge | c6a.16xlarge | c6a.24xlarge | c6a.32xlarge | c6a.48xlarge | c6a.metal
C6g c6g.medium | c6g.large | c6g.xlarge | c6g.2xlarge | c6g.4xlarge | c6g.8xlarge | c6g.12xlarge | c6g.16xlarge | c6g.metal
C6gd c6gd.medium | c6gd.large | c6gd.xlarge | c6gd.2xlarge | c6gd.4xlarge | c6gd.8xlarge | c6gd.12xlarge | c6gd.16xlarge | c6gd.metal
C6gn c6gn.medium | c6gn.large | c6gn.xlarge | c6gn.2xlarge | c6gn.4xlarge | c6gn.8xlarge | c6gn.12xlarge | c6gn.16xlarge
C6i c6i.large | c6i.xlarge | c6i.2xlarge | c6i.4xlarge | c6i.8xlarge | c6i.12xlarge | c6i.16xlarge | c6i.24xlarge | c6i.32xlarge | c6i.metal
C6id c6id.large | c6id.xlarge | c6id.2xlarge | c6id.4xlarge | c6id.8xlarge | c6id.12xlarge | c6id.16xlarge | c6id.24xlarge | c6id.32xlarge | c6id.metal
C6in c6in.large | c6in.xlarge | c6in.2xlarge | c6in.4xlarge | c6in.8xlarge | c6in.12xlarge | c6in.16xlarge | c6in.24xlarge | c6in.32xlarge | c6in.metal
C7a c7a.medium | c7a.large | c7a.xlarge | c7a.2xlarge | c7a.4xlarge | c7a.8xlarge | c7a.12xlarge | c7a.16xlarge | c7a.24xlarge | c7a.32xlarge | c7a.48xlarge | c7a.metal-48xl
C7g c7g.medium | c7g.large | c7g.xlarge | c7g.2xlarge | c7g.4xlarge | c7g.8xlarge | c7g.12xlarge | c7g.16xlarge | c7g.metal
C7gd c7gd.medium | c7gd.large | c7gd.xlarge | c7gd.2xlarge | c7gd.4xlarge | c7gd.8xlarge | c7gd.12xlarge | c7gd.16xlarge
C7gn c7gn.medium | c7gn.large | c7gn.xlarge | c7gn.2xlarge | c7gn.4xlarge | c7gn.8xlarge | c7gn.12xlarge | c7gn.16xlarge
C7i c7i.large | c7i.xlarge | c7i.2xlarge | c7i.4xlarge | c7i.8xlarge | c7i.12xlarge | c7i.16xlarge | c7i.24xlarge | c7i.48xlarge | c7i.metal-24xl | c7i.metal-48xl

Memory optimized

Type Sizes
R5 r5.large | r5.xlarge | r5.2xlarge | r5.4xlarge | r5.8xlarge | r5.12xlarge | r5.16xlarge | r5.24xlarge | r5.metal
R5a r5a.large | r5a.xlarge | r5a.2xlarge | r5a.4xlarge | r5a.8xlarge | r5a.12xlarge | r5a.16xlarge | r5a.24xlarge
R5ad r5ad.large | r5ad.xlarge | r5ad.2xlarge | r5ad.4xlarge | r5ad.8xlarge | r5ad.12xlarge | r5ad.16xlarge | r5ad.24xlarge
R5b r5b.large | r5b.xlarge | r5b.2xlarge | r5b.4xlarge | r5b.8xlarge | r5b.12xlarge | r5b.16xlarge | r5b.24xlarge | r5b.metal
R5d r5d.large | r5d.xlarge | r5d.2xlarge | r5d.4xlarge | r5d.8xlarge | r5d.12xlarge | r5d.16xlarge | r5d.24xlarge | r5d.metal
R5dn r5dn.large | r5dn.xlarge | r5dn.2xlarge | r5dn.4xlarge | r5dn.8xlarge | r5dn.12xlarge | r5dn.16xlarge | r5dn.24xlarge | r5dn.metal
R5n r5n.large | r5n.xlarge | r5n.2xlarge | r5n.4xlarge | r5n.8xlarge | r5n.12xlarge | r5n.16xlarge | r5n.24xlarge | r5n.metal
R6a r6a.large | r6a.xlarge | r6a.2xlarge | r6a.4xlarge | r6a.8xlarge | r6a.12xlarge | r6a.16xlarge | r6a.24xlarge | r6a.32xlarge | r6a.48xlarge | r6a.metal
R6g r6g.medium | r6g.large | r6g.xlarge | r6g.2xlarge | r6g.4xlarge | r6g.8xlarge | r6g.12xlarge | r6g.16xlarge | r6g.metal
R6gd r6gd.medium | r6gd.large | r6gd.xlarge | r6gd.2xlarge | r6gd.4xlarge | r6gd.8xlarge | r6gd.12xlarge | r6gd.16xlarge | r6gd.metal
R6i r6i.large | r6i.xlarge | r6i.2xlarge | r6i.4xlarge | r6i.8xlarge | r6i.12xlarge | r6i.16xlarge | r6i.24xlarge | r6i.32xlarge | r6i.metal
R6idn r6idn.large | r6idn.xlarge | r6idn.2xlarge | r6idn.4xlarge | r6idn.8xlarge | r6idn.12xlarge | r6idn.16xlarge | r6idn.24xlarge | r6idn.32xlarge | r6idn.metal
R6in r6in.large | r6in.xlarge | r6in.2xlarge | r6in.4xlarge | r6in.8xlarge | r6in.12xlarge | r6in.16xlarge | r6in.24xlarge | r6in.32xlarge | r6in.metal
R6id r6id.large | r6id.xlarge | r6id.2xlarge | r6id.4xlarge | r6id.8xlarge | r6id.12xlarge | r6id.16xlarge | r6id.24xlarge | r6id.32xlarge | r6id.metal
R7a r7a.medium | r7a.large | r7a.xlarge | r7a.2xlarge | r7a.4xlarge | r7a.8xlarge | r7a.12xlarge | r7a.16xlarge | r7a.24xlarge | r7a.32xlarge | r7a.48xlarge | r7a.metal-48xl
R7g r7g.medium | r7g.large | r7g.xlarge | r7g.2xlarge | r7g.4xlarge | r7g.8xlarge | r7g.12xlarge | r7g.16xlarge | r7g.metal
R7gd r7gd.medium | r7gd.large | r7gd.xlarge | r7gd.2xlarge | r7gd.4xlarge | r7gd.8xlarge | r7gd.12xlarge | r7gd.16xlarge
R7i r7i.large | r7i.xlarge | r7i.2xlarge | r7i.4xlarge | r7i.8xlarge | r7i.12xlarge | r7i.16xlarge | r7i.24xlarge | r7i.48xlarge | r7i.metal-24xl | r7i.metal-48xl
R7iz r7iz.large | r7iz.xlarge | r7iz.2xlarge | r7iz.4xlarge | r7iz.8xlarge | r7iz.12xlarge | r7iz.16xlarge | r7iz.32xlarge | r7iz.metal-16xl | r7iz.metal-32xl
U-3tb1 u-3tb1.56xlarge
U-6tb1 u-6tb1.56xlarge | u-6tb1.112xlarge | u-6tb1.metal
U-9tb1 u-9tb1.112xlarge | u-9tb1.metal
U-12tb1 u-12tb1.112xlarge | u-12tb1.metal
U-18tb1 u-18tb1.112xlarge | u-18tb1.metal
U-24tb1 u-24tb1.112xlarge | u-24tb1.metal
X1 x1.16xlarge | x1.32xlarge
X2gd x2gd.medium | x2gd.large | x2gd.xlarge | x2gd.2xlarge | x2gd.4xlarge | x2gd.8xlarge | x2gd.12xlarge | x2gd.16xlarge | x2gd.metal
X2idn x2idn.16xlarge | x2idn.24xlarge | x2idn.32xlarge | x2idn.metal
X2iedn x2iedn.xlarge | x2iedn.2xlarge | x2iedn.4xlarge | x2iedn.8xlarge | x2iedn.16xlarge | x2iedn.24xlarge | x2iedn.32xlarge | x2iedn.metal
X2iezn x2iezn.2xlarge | x2iezn.4xlarge | x2iezn.6xlarge | x2iezn.8xlarge | x2iezn.12xlarge | x2iezn.metal
X1e x1e.xlarge | x1e.2xlarge | x1e.4xlarge | x1e.8xlarge | x1e.16xlarge | x1e.32xlarge
z1d z1d.large | z1d.xlarge | z1d.2xlarge | z1d.3xlarge | z1d.6xlarge | z1d.12xlarge | z1d.metal

Storage optimized

Type Sizes
D2 d2.xlarge | d2.2xlarge | d2.4xlarge | d2.8xlarge
D3 d3.xlarge | d3.2xlarge | d3.4xlarge | d3.8xlarge
D3en d3en.xlarge | d3en.2xlarge | d3en.4xlarge | d3en.6xlarge | d3en.8xlarge | d3en.12xlarge
H1 h1.2xlarge | h1.4xlarge | h1.8xlarge | h1.16xlarge
I3 i3.large | i3.xlarge | i3.2xlarge | i3.4xlarge | i3.8xlarge | i3.16xlarge | i3.metal
I3en i3en.large | i3en.xlarge | i3en.2xlarge | i3en.3xlarge | i3en.6xlarge | i3en.12xlarge | i3en.24xlarge | i3en.metal
I4g i4g.large | i4g.xlarge | i4g.2xlarge | i4g.4xlarge | i4g.8xlarge | i4g.16xlarge
I4i i4i.large | i4i.xlarge | i4i.2xlarge | i4i.4xlarge | i4i.8xlarge | i4i.12xlarge | i4i.16xlarge | i4i.24xlarge | i4i.32xlarge | i4i.metal
Im4gn im4gn.large | im4gn.xlarge | im4gn.2xlarge | im4gn.4xlarge | im4gn.8xlarge | im4gn.16xlarge
Is4gen is4gen.medium | is4gen.large | is4gen.xlarge | is4gen.2xlarge | is4gen.4xlarge | is4gen.8xlarge

Accelerated computing

Type Sizes
DL1 dl1.24xlarge
DL2q dl2q.24xlarge
F1 f1.2xlarge | f1.4xlarge | f1.16xlarge
G3 g3.4xlarge | g3.8xlarge | g3.16xlarge
G4ad g4ad.xlarge | g4ad.2xlarge | g4ad.4xlarge | g4ad.8xlarge | g4ad.16xlarge
G4dn g4dn.xlarge | g4dn.2xlarge | g4dn.4xlarge | g4dn.8xlarge | g4dn.12xlarge | g4dn.16xlarge | g4dn.metal
G5 g5.xlarge | g5.2xlarge | g5.4xlarge | g5.8xlarge | g5.12xlarge | g5.16xlarge | g5.24xlarge | g5.48xlarge
G5g g5g.xlarge | g5g.2xlarge | g5g.4xlarge | g5g.8xlarge | g5g.16xlarge | g5g.metal
Inf1 inf1.xlarge | inf1.2xlarge | inf1.6xlarge | inf1.24xlarge
Inf2 inf2.xlarge | inf2.8xlarge | inf2.24xlarge | inf2.48xlarge
P2 p2.xlarge | p2.8xlarge | p2.16xlarge
P3 p3.2xlarge | p3.8xlarge | p3.16xlarge
P3dn p3dn.24xlarge
P4d p4d.24xlarge
P4de p4de.24xlarge
P5 p5.48xlarge
Trn1 trn1.2xlarge | trn1.32xlarge
Trn1n trn1n.32xlarge
VT1 vt1.3xlarge | vt1.6xlarge | vt1.24xlarge

High-performance computing

Type Sizes
Hpc6a hpc6a.48xlarge
Hpc6id hpc6id.32xlarge
Hpc7a hpc7a.12xlarge | hpc7a.24xlarge | hpc7a.48xlarge | hpc7a.96xlarge
Hpc7g hpc7g.4xlarge | hpc7g.8xlarge | hpc7g.16xlarge

Previous generation instances

Amazon Web Services offers previous generation instance types for users who have optimized their applications around them and have yet to upgrade. We encourage you to use current generation instance types to get the best performance, but we continue to support the following previous generation instance types. For more information about which current generation instance type would be a suitable upgrade, see Previous Generation Instances.

Type Sizes
A1 a1.medium | a1.large | a1.xlarge | a1.2xlarge | a1.4xlarge | a1.metal
C1 c1.medium | c1.xlarge
C3 c3.large | c3.xlarge | c3.2xlarge | c3.4xlarge | c3.8xlarge
C4 c4.large | c4.xlarge | c4.2xlarge | c4.4xlarge | c4.8xlarge
G2 g2.2xlarge | g2.8xlarge
I2 i2.xlarge | i2.2xlarge | i2.4xlarge | i2.8xlarge
M1 m1.small | m1.medium | m1.large | m1.xlarge
M2 m2.xlarge | m2.2xlarge | m2.4xlarge
M3 m3.medium | m3.large | m3.xlarge | m3.2xlarge
M4 m4.large | m4.xlarge | m4.2xlarge | m4.4xlarge | m4.10xlarge | m4.16xlarge
R3 r3.large | r3.xlarge | r3.2xlarge | r3.4xlarge | r3.8xlarge
R4 r4.large | r4.xlarge | r4.2xlarge | r4.4xlarge | r4.8xlarge | r4.16xlarge
T1 t1.micro

Hardware specifications

For detailed instance type specifications, see the Amazon EC2 Instance Types Guide. For pricing information, see Amazon EC2 Instance Types.

To determine which instance type best meets your needs, we recommend that you launch an instance and use your own benchmark application. Because you pay by the instance second, it's convenient and inexpensive to test multiple instance types before making a decision. If your needs change, even after you make a decision, you can change the instance type later. For more information, see Change the instance type.

Processor features

Intel processor features

Amazon EC2 instances that run on Intel processors may include the following features. Not all of the following processor features are supported by all instance types. For detailed information about which features are available for each instance type, see Amazon EC2 Instance Types.

  • Intel AES New Instructions (AES-NI) — Intel AES-NI encryption instruction set improves upon the original Advanced Encryption Standard (AES) algorithm to provide faster data protection and greater security. All current generation EC2 instances support this processor feature.

  • Intel Advanced Vector Extensions (Intel AVX, Intel AVX2, and Intel AVX-512) — Intel AVX and Intel AVX2 are 256-bit, and Intel AVX-512 is a 512-bit instruction set extension designed for applications that are Floating Point (FP) intensive. Intel AVX instructions improve performance for applications like image and audio/video processing, scientific simulations, financial analytics, and 3D modeling and analysis. These features are only available on instances launched with HVM AMIs.

  • Intel Turbo Boost Technology — Intel Turbo Boost Technology processors automatically run cores faster than the base operating frequency.

  • Intel Deep Learning Boost (Intel DL Boost) — Accelerates AI deep learning use cases. The 2nd Gen Intel Xeon Scalable processors extend Intel AVX-512 with a new Vector Neural Network Instruction (VNNI/INT8) that significantly increases deep learning inference performance over previous generation Intel Xeon Scalable processors (with FP32) for image recognition/segmentation, object detection, speech recognition, language translation, recommendation systems, reinforcement learning, and more. VNNI may not be compatible with all Linux distributions.

    The following instances support VNNI: M5n, R5n, M5dn, M5zn, R5b, R5dn, D3, D3en, and C6i. C5 and C5d instances support VNNI for only 12xlarge, 24xlarge, and metal instances.

Confusion may result from industry naming conventions for 64-bit CPUs. Chip manufacturer Advanced Micro Devices (AMD) introduced the first commercially successful 64-bit architecture based on the Intel x86 instruction set. Consequently, the architecture is widely referred to as AMD64 regardless of the chip manufacturer. Windows and several Linux distributions follow this practice. This explains why the internal system information on an instance running Ubuntu or Windows displays the CPU architecture as AMD64 even though the instances are running on Intel hardware.

AMI virtualization types

The virtualization type of your instance is determined by the AMI that you use to launch it. Current generation instance types support hardware virtual machine (HVM) only. Some previous generation instance types support paravirtual (PV) and some AWS Regions support PV instances. For more information, see Linux AMI virtualization types.

For best performance, we recommend that you use an HVM AMI. In addition, HVM AMIs are required to take advantage of enhanced networking. HVM virtualization uses hardware-assist technology provided by the AWS platform. With HVM virtualization, the guest VM runs as if it were on a native hardware platform, except that it still uses PV network and storage drivers for improved performance.

Instances built on the Nitro System

The Nitro System is a collection of hardware and software components built by AWS that enable high performance, high availability, and high security. For more information, see AWS Nitro System.

The Nitro System provides bare metal capabilities that eliminate virtualization overhead and support workloads that require full access to host hardware. Bare metal instances are well suited for the following:

  • Workloads that require access to low-level hardware features (for example, Intel VT) that are not available or fully supported in virtualized environments

  • Applications that require a non-virtualized environment for licensing or support

Nitro components

The following components are part of the Nitro System:

  • Nitro card

    • Local NVMe storage volumes

    • Networking hardware support

    • Management

    • Monitoring

    • Security

  • Nitro security chip, integrated into the motherboard

  • Nitro hypervisor - A lightweight hypervisor that manages memory and CPU allocation and delivers performance that is indistinguishable from bare metal for most workloads.

Virtualized instances

The following virtualized instances are built on the Nitro System:

  • General purpose: A1 | M5 | M5a | M5ad | M5d | M5dn | M5n | M5zn | M6a | M6g | M6gd | M6i | M6id | M6idn | M6in | M7a | M7g | M7gd | M7i | M7i-flex | T3 | T3a | T4g

  • Compute optimized: C5 | C5a | C5ad | C5d | C5n | C6a | C6g | C6gd | C6gn | C6i | C6id | C6in | C7a | C7g | C7gd | C7gn | C7i

  • Memory optimized: R5 | R5a | R5ad | R5b | R5d | R5dn | R5n | R6a | R6g | R6gd | R6i | R6idn | R6in | R6id | R7a | R7g | R7gd | R7i | R7iz | U-3tb1 | U-6tb1 | U-9tb1 | U-12tb1 | U-18tb1 | U-24tb1 | X2gd | X2idn | X2iedn | X2iezn | z1d

  • Storage optimized: D3 | D3en | I3en | I4g | I4i | Im4gn | Is4gen

  • Accelerated computing: DL1 | DL2q | G4ad | G4dn | G5 | G5g | Inf1 | Inf2 | P3dn | P4d | P4de | P5 | Trn1 | Trn1n | VT1

  • High-performance computing: Hpc6a | Hpc6id | Hpc7a | Hpc7g

Bare metal instances

The following bare metal instances are built on the Nitro System:

  • General purpose: a1.metal | m5.metal | m5d.metal | m5dn.metal | m5n.metal | m5zn.metal | m6a.metal | m6g.metal | m6gd.metal | m6i.metal | m6id.metal | m6idn.metal | m6in.metal | m7a.metal-48xl | m7g.metal | m7i.metal-24xl | m7i.metal-48xl | mac1.metal | mac2.metal | mac2-m2.metal | mac2-m2pro.metal

  • Compute optimized: c5.metal | c5d.metal | c5n.metal | c6a.metal | c6g.metal | c6gd.metal | c6i.metal | c6id.metal | c6in.metal | c7a.metal-48xl | c7g.metal | c7i.metal-24xl | c7i.metal-48xl

  • Memory optimized: r5.metal | r5b.metal | r5d.metal | r5dn.metal | r5n.metal | r6a.metal | r6g.metal | r6gd.metal | r6i.metal | r6idn.metal | r6in.metal | r6id.metal | r7a.metal-48xl | r7g.metal | r7i.metal-24xl | r7i.metal-48xl | r7iz.metal-16xl | r7iz.metal-32xl | u-6tb1.metal | u-9tb1.metal | u-12tb1.metal | u-18tb1.metal | u-24tb1.metal | x2gd.metal | x2idn.metal | x2iedn.metal | x2iezn.metal | z1d.metal

  • Storage optimized: i3.metal | i3en.metal | i4i.metal

  • Accelerated computing: g4dn.metal | g5g.metal

Networking and storage features

When you select an instance type, this determines the networking and storage features that are available. To describe an instance type, use the describe-instance-types command.

Networking features
  • IPv6 is supported on all instance types, except C1, M1, M2, M3, and T1.

  • To maximize the networking and bandwidth performance of your instance type, you can do the following:

    • Launch supported instance types into a cluster placement group to optimize your instances for high performance computing (HPC) applications. Instances in a common cluster placement group can benefit from high-bandwidth, low-latency networking. For more information, see Placement groups.

    • Enable enhanced networking for supported current generation instance types to get significantly higher packet per second (PPS) performance, lower network jitter, and lower latencies. For more information, see Enhanced networking on Linux.

  • Current generation instance types that are enabled for enhanced networking have the following networking performance attributes:

    • Traffic within the same Region over private IPv4 or IPv6 can support 5 Gbps for single-flow traffic and up to 25 Gbps for multi-flow traffic (depending on the instance type).

    • Traffic to and from Amazon S3 buckets within the same Region over the public IP address space or through a VPC endpoint can use all available instance aggregate bandwidth.

  • The maximum transmission unit (MTU) supported varies across instance types. All Amazon EC2 instance types support standard Ethernet V2 1500 MTU frames. All current generation instances support 9001 MTU, or jumbo frames, and some previous generation instances support them as well. For more information, see Network maximum transmission unit (MTU) for your EC2 instance.

Storage features
  • Some instance types support EBS volumes and instance store volumes, while other instance types support only EBS volumes. Some instance types that support instance store volumes use solid state drives (SSD) to deliver very high random I/O performance. Some instance types support NVMe instance store volumes. Some instance types support NVMe EBS volumes. For more information, see Amazon EBS and NVMe on Linux instances and NVMe SSD volumes.

  • To obtain additional, dedicated capacity for Amazon EBS I/O, you can launch some instance types as EBS–optimized instances. Some instance types are EBS–optimized by default. For more information, see Amazon EBS–optimized instances.

Summary of networking and storage features

The following table summarizes the networking and storage features supported by current generation instance types.

General purpose

Instance type EBS only NVME EBS Instance store Placement group Enhanced networking
M5 Yes Yes No Yes ENA
M5a Yes Yes No Yes ENA
M5ad No Yes NVMe Yes ENA
M5d No Yes NVMe Yes ENA
M5dn No Yes NVMe Yes ENA | EFA
M5n Yes Yes No Yes ENA | EFA
M5zn Yes Yes No Yes ENA | EFA
M6a Yes Yes No Yes ENA | EFA
M6g Yes Yes No Yes ENA
M6gd No Yes NVMe Yes ENA
M6i Yes Yes No Yes ENA | EFA
M6id No Yes NVMe Yes ENA | EFA
M6idn No Yes NVMe Yes ENA | EFA
M6in Yes Yes No Yes ENA | EFA
M7a Yes Yes No Yes ENA | EFA
M7g Yes Yes No Yes ENA | EFA
M7gd No Yes NVMe Yes ENA | EFA
M7i Yes Yes No Yes ENA | EFA
M7i-flex Yes Yes No Yes ENA
Mac1 Yes Yes No Yes ENA
Mac2 Yes Yes No Yes ENA
Mac2-m2 Yes Yes No Yes ENA
Mac2-m2pro Yes Yes No Yes ENA
T2 Yes No No Yes Not supported
T3 Yes Yes No Yes ENA
T3a Yes Yes No Yes ENA
T4g Yes Yes No Yes ENA

Compute optimized

Instance type EBS only NVME EBS Instance store Placement group Enhanced networking
C5 Yes Yes No Yes ENA
C5a Yes Yes No Yes ENA
C5ad No Yes NVMe Yes ENA
C5d No Yes NVMe Yes ENA
C5n Yes Yes No Yes ENA | EFA
C6a Yes Yes No Yes ENA | EFA
C6g Yes Yes No Yes ENA
C6gd No Yes NVMe Yes ENA
C6gn Yes Yes No Yes ENA | EFA
C6i Yes Yes No Yes ENA | EFA
C6id No Yes NVMe Yes ENA | EFA
C6in Yes Yes No Yes ENA | EFA
C7a Yes Yes No Yes ENA | EFA
C7g Yes Yes No Yes ENA | EFA
C7gd No Yes NVMe Yes ENA | EFA
C7gn Yes Yes No Yes ENA | EFA
C7i Yes Yes No Yes ENA | EFA

Memory optimized

Instance type EBS only NVME EBS Instance store Placement group Enhanced networking
R5 Yes Yes No Yes ENA
R5a Yes Yes No Yes ENA
R5ad No Yes NVMe Yes ENA
R5b Yes Yes No Yes ENA
R5d No Yes NVMe Yes ENA
R5dn No Yes NVMe Yes ENA | EFA
R5n Yes Yes No Yes ENA | EFA
R6a Yes Yes No Yes ENA | EFA
R6g Yes Yes No Yes ENA
R6gd No Yes NVMe Yes ENA
R6i Yes Yes No Yes ENA | EFA
R6idn No Yes NVMe Yes ENA | EFA
R6in Yes Yes No Yes ENA | EFA
R6id No Yes NVMe Yes ENA | EFA
R7a Yes Yes No Yes ENA | EFA
R7g Yes Yes No Yes ENA | EFA
R7gd No Yes NVMe Yes ENA | EFA
R7i Yes Yes No Yes ENA | EFA
R7iz Yes Yes No Yes ENA | EFA
U-3tb1 Yes Yes No Yes ENA
U-6tb1 Yes Yes No Yes ENA
U-9tb1 Yes Yes No Yes ENA
U-12tb1 Yes Yes No Yes ENA
U-18tb1 Yes Yes No Yes ENA
U-24tb1 Yes Yes No Yes ENA
X1 No No SSD Yes ENA
X2gd No Yes NVMe Yes ENA
X2idn No Yes NVMe Yes ENA | EFA
X2iedn No Yes NVMe Yes ENA | EFA
X2iezn Yes Yes No Yes ENA | EFA
X1e No No SSD Yes ENA
z1d No Yes NVMe Yes ENA

Storage optimized

Instance type EBS only NVME EBS Instance store Placement group Enhanced networking
D2 No No HDD Yes Intel 82599 VF
D3 No Yes NVMe Yes ENA
D3en No Yes NVMe Yes ENA
H1 No No HDD Yes ENA
I3 No Yes NVMe Yes ENA
I3en No Yes NVMe Yes ENA | EFA
I4g No Yes NVMe Yes ENA | EFA
I4i No Yes NVMe Yes ENA | EFA
Im4gn No Yes NVMe Yes ENA | EFA
Is4gen No Yes NVMe Yes ENA

Accelerated computing

Instance type EBS only NVME EBS Instance store Placement group Enhanced networking
DL1 No Yes NVMe Yes ENA | EFA
DL2q Yes Yes No Yes ENA | EFA
F1 No No NVMe Yes ENA
G3 Yes No No Yes ENA
G4ad No Yes NVMe Yes ENA
G4dn No Yes NVMe Yes ENA | EFA
G5 No Yes NVMe Yes ENA | EFA
G5g Yes Yes No Yes ENA
Inf1 Yes Yes No Yes ENA | EFA
Inf2 Yes Yes No Yes ENA
P2 Yes No No Yes ENA
P3 Yes No No Yes ENA
P3dn No Yes NVMe Yes ENA | EFA
P4d No Yes NVMe Yes ENA | EFA
P4de No Yes NVMe Yes ENA | EFA
P5 No Yes NVMe Yes ENA | EFA
Trn1 No Yes NVMe Yes ENA | EFA
Trn1n No Yes NVMe Yes ENA | EFA
VT1 Yes Yes No Yes ENA | EFA

High-performance computing

Instance type EBS only NVME EBS Instance store Placement group Enhanced networking
Hpc6a Yes Yes No Yes ENA | EFA
Hpc6id No Yes NVMe Yes ENA | EFA
Hpc7a Yes Yes No Yes ENA | EFA
Hpc7g Yes Yes No Yes ENA | EFA

Previous generation instance types

The following table summarizes the networking and storage features supported by previous generation instance types.

Instance type EBS only NVME EBS Instance store Placement group Enhanced networking
A1 Yes Yes No Yes ENA
C1 No No HDD Yes Not supported
C3 No No SSD Yes Intel 82599 VF
C4 Yes No No Yes Intel 82599 VF
G2 No No SSD Yes Not supported
I2 No No SSD Yes Intel 82599 VF
M1 No No HDD Yes Not supported
M2 No No HDD Yes Not supported
M3 No No SSD Yes Not supported
M4 Yes No No Yes ENA
R3 No No SSD Yes Intel 82599 VF
R4 Yes No No Yes ENA
T1 Yes No No Yes Not supported

m4.16xlarge instances support enhanced networking using an Elastic Network Adapter (ENA). All other M4 instance types support enhanced networking using an Intel 82599 VF interface.

Instance limits

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 about the default limits, see How many instances can I run in Amazon EC2?

For more information about viewing your current limits or requesting an increase in your current limits, see Amazon EC2 service quotas.