General purpose

General purpose instances are suitable for a wide range of workloads, offering a balanced distribution of CPU and RAM resources. They are ideal for common computing scenarios such as web servers, small to medium-sized databases, application servers, and development/test environments.

• Applicable types: m3az, m2a, m1ie, t1i

m3az

The m3az instance uses the latest 4th-generation AMD EPYC 9004 series processors. It supports high turbo clock speeds of up to 3.95GHz. Designed for single-threaded CPU environments, this instance type is optimized for workloads like certain HPC applications that operate on a single-threaded CPU.

Hardware specifications

• Up to 4.1GHz 4th-generation AMD EPYC processor (AMD EPYC 9474F Genoa)
• Up to 12.5Gbps network bandwidth
• Instance sizes supporting high-performance vCPUs mapped 1:1 to up to 32 physical CPUs and 128GiB of memory

Details

Instance size	vCPU	Memory (GiB)	Network bandwidth (Gbps)
m3az.large	2	8	up to 10
m3az.xlarge	4	16	up to 10
m3az.2xlarge	8	32	up to 10
m3az.4xlarge	16	64	up to 10
m3az.6xlarge	24	96	12.5
m3az.8xlarge	32	128	12.5

NUMA topology

In Non-Uniform Memory Access (NUMA) architecture, each CPU can access its own local memory. By allowing multiple processors to share memory, NUMA architecture ensures high scalability.

The m3az instance has the following NUMA topology based on instance size:

Instance size	NUMA domain count	Default cores per NUMA
m3az.large	1	2
m3az.xlarge	1	4
m3az.2xlarge	1	8
m3az.4xlarge	1	16
m3az.6xlarge	1	24
m3az.8xlarge	1	32

NUMA topology architecture

m3az.large

m3az.xlarge

m3az.2xlarge

m3az.4xlarge

m3az.6xlarge

m3az.8xlarge

m2a

The m2a instance uses the 3rd-generation AMD EPYC 7000 series processors. It supports an all-core turbo clock speed of up to 3.6GHz and is designed for workloads that do not require full computing capacity at all times. This enables users to optimize configuration based on cost and performance.

Hardware specifications

• Up to 3.6GHz 3rd-generation AMD EPYC processor (AMD EPYC 7643)
• Up to 25Gbps network bandwidth
• Instance sizes supporting up to 96 vCPUs and 384GiB of memory
• Supports AMD instruction sets (AVX, AVX2)

Details

Instance size	vCPU	Memory (GiB)	Network bandwidth (Gbps)
m2a.large	2	8	Max 10
m2a.xlarge	4	16	Max 10
m2a.2xlarge	8	32	Max 10
m2a.4xlarge	16	64	Max 10
m2a.8xlarge	32	128	Max 10
m2a.12xlarge	48	192	Max 12.5
m2a.16xlarge	64	256	Max 12.5
m2a.24xlarge	96	384	Max 25

NUMA topology

The m2a instance has the following NUMA topology based on instance size:

Instance size	NUMA domain count	Default cores per NUMA
m2a.large	1	1
m2a.xlarge	1	2
m2a.2xlarge	1	4
m2a.4xlarge	1	8
m2a.8xlarge	2	8
m2a.12xlarge	2	12
m2a.16xlarge	2	16
m2a.24xlarge	2	24

NUMA topology architecture

m2a.large

m2a.xlarge

m2a.2xlarge

m2a.4xlarge

m2a.8xlarge

m2a.12xlarge

m2a.16xlarge

m2a.24xlarge

m1ie

The m1ie instance uses 2nd-generation Intel Xeon Scalable processors and supports a base clock speed of 2.2GHz. This instance type offers a balanced mix of compute, memory, and network resources, and supports larger-than-standard storage capacities, making it ideal for storage-intensive applications.

Hardware specifications

• Up to 4.0GHz 2nd-generation Intel Xeon Scalable processor (Gold 5220R Cascade Lake)
• Up to 50Gbps network bandwidth
• Instance sizes supporting up to 96 vCPUs and 512GiB of memory
• Supports up to 8TB of local disk capacity

Details

Instance size	vCPU	Memory (GiB)	Network bandwidth (Gbps)	Storage or local disk size (TB)
m1ie.baremetal	96	512	50	8

t1i

The t1i instance is a next-generation burstable general purpose instance that balances compute, memory, and network resources. It is suitable for applications with moderate CPU usage and occasional usage spikes.
For more details on burstable instances, refer to Burstable instances.

caution

Ensure the instance size meets the minimum memory requirements of your OS and application. For Windows OS, most instances must be at least micro-sized.

Hardware specifications

• Up to 3.9GHz 2nd-generation Intel Xeon Scalable processor (Cascade Lake 5220 or Skylake 5120)
• Up to 5Gbps network bandwidth
• Instance sizes supporting up to 8 vCPUs and 32GiB of memory
• Supports Intel instruction sets (AVX, AVX2, AVX-512)
• Provides baseline CPU performance for standard workloads
• Burst CPU performance managed via credits

Details

Instance size	vCPU	Memory (GiB)	Network bandwidth (Gbps)	CPU credits/hour	Credit cap (24h)	Baseline usage/vCPU
t1i.nano	2	0.5	up to 5	6	144	5%
t1i.micro	2	1	up to 5	12	288	10%
t1i.small	2	2	up to 5	24	576	20%
t1i.medium	2	4	up to 5	24	576	20%
t1i.large	2	8	up to 5	36	864	30%
t1i.xlarge	4	16	up to 5	96	2304	40%
t1i.2xlarge	8	32	up to 5	192	4608	40%