t3a.xlarge
Amazon EC2 · T3a family
The t3a.xlarge is a general purpose Amazon EC2 instance type with 4 vCPUs and 16 GiB of memory, running on AMD EPYC 7571 with up to 5 Gigabit networking, starting at $0.0986 per hour on demand.
- vCPUs
- 4
- memory
- 16 GiB
- processor
- AMD
- on-demand
- $0.1504
- spot
- $0.0689
- 1-yr reserved
- $0.0943
shared / burstable
4 GiB per vCPU
2.5 GHz
$110/mo · us-east-1
−54% vs on-demand
−37% vs on-demand
t3a.xlarge price by region
19 regions · hourly list prices in USD · On-demand · Spot · 1-yr Reserved · 3-yr Reserved
| Asia Pacific (Mumbai)ap-south-1cheapestAsia | $0.0986 | $0.0366 | $0.0621 | $0.0416 |
| US East (N. Virginia)us-east-1referenceNorth America | $0.1504 | $0.0689 | $0.0943 | $0.065 |
| US East (Ohio)us-east-2North America | $0.1504 | $0.0572 | $0.0943 | $0.065 |
| US West (Oregon)us-west-2North America | $0.1504 | $0.0658 | $0.0943 | $0.065 |
| EU (Ireland)eu-west-1Europe | $0.1632 | $0.073 | $0.1028 | $0.0705 |
| Canada (Central)ca-central-1North America | $0.167 | $0.0788 | $0.1052 | $0.0722 |
| EU (London)eu-west-2Europe | $0.1699 | $0.0716 | $0.107 | $0.0734 |
| EU (Paris)eu-west-3Europe | $0.1699 | $0.0634 | $0.107 | $0.0734 |
| Israel (Tel Aviv)il-central-1Middle East | $0.1715 | $0.0637 | $0.1081 | $0.0741 |
| EU (Milan)eu-south-1Europe | $0.1725 | $0.0475 | $0.1087 | $0.0745 |
| EU (Frankfurt)eu-central-1Europe | $0.1728 | $0.0763 | $0.1089 | $0.0746 |
| AWS GovCloud (US-East)us-gov-east-1North America | $0.1757 | – | $0.1107 | $0.0759 |
| AWS GovCloud (US)us-gov-west-1North America | $0.1757 | – | $0.1107 | $0.0759 |
| US West (N. California)us-west-1North America | $0.1786 | $0.079 | $0.1125 | $0.0771 |
| Asia Pacific (Seoul)ap-northeast-2Asia | $0.1872 | $0.0742 | $0.1131 | $0.0809 |
| Asia Pacific (Singapore)ap-southeast-1Asia | $0.1888 | $0.0817 | $0.1187 | $0.0799 |
| Asia Pacific (Sydney)ap-southeast-2Oceania | $0.1901 | $0.075 | $0.1198 | $0.0821 |
| Asia Pacific (Tokyo)ap-northeast-1Asia | $0.1958 | $0.0683 | $0.1234 | $0.0846 |
| South America (Sao Paulo)sa-east-1South America | $0.2419 | $0.1123 | $0.1389 | $0.1045 |
Specifications
compute
- vCPUs
- 4
- Memory
- 16 GiB
- Memory per vCPU
- 4 GiB
- Processor
- AMD EPYC 7571
- CPU vendor
- AMD
- Architecture
- x86-64
- Clock speed
- 2.5 GHz
- Burstable CPU
- Yes (shared cores)
storage & network
- Local storage
- Network storage only
- Storage detail
- EBS only
- Network
- Up to 5 Gigabit
- Peak bandwidth
- 5 Gbps
aws details
- API name
t3a.xlarge- Family
- T3a
- Category
- General purpose
- Generation
- Current
- Regions
- 19 (Asia, Europe, Middle East, North America, Oceania, South America)
- Cheapest region
- Asia Pacific (Mumbai) · $0.0986/hr
- Dedicated EBS throughput
- Up to 2780 Mbps
- Enhanced networking
- Yes
- Normalization size factor
- 8
- Processor features
- AVX; AVX2; AMD Turbo
- Spot interruption rate (us-east-1)
- >20%
- Spot savings vs on-demand (us-east-1)
- 61%
t3a.xlarge equivalents on other clouds
closest match by vCPU, memory, CPU architecture, accelerator and class · Linux on-demand in each provider's reference region
| Provider | Instance | vCPU | Memory | Processor | On-demand | vs t3a.xlarge |
|---|---|---|---|---|---|---|
| AWS | t3a.xlarge | 4 | 16 GiB | AMD EPYC 7571 | $0.1504 | – |
| Azure | B4as v2 | 4 | 16 GiB | AMD EPYC 7763v (Milan) | $0.15 | 0% |
| B4s v2 | 4 | 16 GiB | Intel Xeon Platinum 8370C (Ice Lake) | $0.166 | +10% | |
| Google Cloud | e2-standard-4dedicated CPU | 4 | 16 GiB | AMD Rome, Intel Broadwell, Intel Haswell, Intel Skylake | $0.134 | -11% |
| n2d-standard-4dedicated CPU | 4 | 16 GiB | AMD Milan, AMD Rome | $0.169 | +12% | |
| Akamai | G8 Dedicated 16x4dedicated CPU | 4 | 16 GiB | AMD EPYC | $0.21 | +40% |
| Linode 16GB6 vCPU (1.5×) | 6 | 16 GiB | AMD EPYC | $0.144 | -4% | |
| Vultr | vhf-4c-16gb | 4 | 16 GiB | Intel Xeon | $0.132 | -12% |
| vhp-4c-12gb-amd12 GiB (1.3× less) | 4 | 12 GiB | AMD EPYC | $0.099 | -34% |
T3a family sizes
| Size | vCPU | Memory | Local storage | On-demand | Monthly | Spot | 1-yr |
|---|---|---|---|---|---|---|---|
| t3a.nano | 2 | 0.5 GiB | – | $0.0047 | $3.431 | $0.0021 | $0.0029 |
| t3a.micro | 2 | 1 GiB | – | $0.0094 | $6.862 | $0.0044 | $0.0059 |
| t3a.small | 2 | 2 GiB | – | $0.0188 | $13.72 | $0.0084 | $0.0118 |
| t3a.medium | 2 | 4 GiB | – | $0.0376 | $27.45 | $0.016 | $0.0236 |
| t3a.large | 2 | 8 GiB | – | $0.0752 | $54.90 | $0.0318 | $0.0471 |
| t3a.xlarge | 4 | 16 GiB | – | $0.1504 | $110 | $0.0689 | $0.0943 |
| t3a.2xlarge | 8 | 32 GiB | – | $0.3008 | $220 | $0.1474 | $0.1886 |
Similar instance types on AWS
other AWS families with a similar shape, cheapest first
t3a.xlarge FAQ
How much does t3a.xlarge cost?
t3a.xlarge costs $0.1504 per hour on demand with Linux in US East (N. Virginia) (us-east-1), about $110 per month. Across 19 regions the Linux on-demand price ranges from $0.0986 in Asia Pacific (Mumbai) to $0.2419 in South America (Sao Paulo). Spot capacity costs $0.0689 per hour (54% less), but can be reclaimed. A 1-yr Reserved commitment brings it to $0.0943 per hour. With Windows it costs $0.224 per hour.
How many vCPUs and how much memory does t3a.xlarge have?
t3a.xlarge has 4 vCPUs and 16 GiB of memory (4 GiB per vCPU), running on AMD EPYC 7571. It is a general purpose instance type in the T3a family with a burstable (shared) CPU.
What is the cheapest region for t3a.xlarge?
Asia Pacific (Mumbai) (ap-south-1) at $0.0986 per hour for Linux on demand, 34% less than US East (N. Virginia). t3a.xlarge is sold in 19 AWS regions.
What are the equivalents of t3a.xlarge on other clouds?
The closest matches by vCPU, memory, CPU architecture and instance class are on Azure, B4as v2 (4 vCPU, 16 GiB, $0.15/hour in East US); on Google Cloud, e2-standard-4 (4 vCPU, 16 GiB, $0.134/hour in Iowa); on Akamai, G8 Dedicated 16x4 (4 vCPU, 16 GiB, $0.21/hour in Newark, NJ); on Vultr, vhf-4c-16gb (4 vCPU, 16 GiB, $0.132/hour in New Jersey).
Is t3a.xlarge burstable?
Yes. t3a.xlarge shares physical CPU cores and is designed for workloads that are idle most of the time and occasionally need full CPU. Sustained high CPU use can be throttled or cost extra, so pick a general purpose size with dedicated vCPUs for steady load.