General description
High performance at the lowest cost in Amazon EC2 for generative AI inference
-
Elastic Compute Cloud
Service
-
Inferentia 2
Series
-
AI
Type
Hardware description
The inf2.8xlarge shows the following characteristics:
-
Virtual machine
Type
-
32 CPUs
CPU number
-
arm64
CPU architecture
-
131072 MB
RAM
-
None
GPU
-
25000 Mbps
Maximum network bandwidth
-
None
Root volume
-
None
Extra volume(s)
CPU specifications
After analyse of the hardware our systems collected the following data about processors:
Pricing
Here's a cost sample of inf2.8xlarge:
| Zone | Hourly | Monthly | Yearly |
|---|---|---|---|
| Asia Pacific (Mumbai) | 2.558 | 1177.0 | 14121.0 |
| Asia Pacific (Singapore) | 2.755 | 1267.0 | 15207.0 |
| EU (Frankfurt) | 2.952 | 1358.0 | 16294.0 |
| EU (Ireland) | 2.460 | 1132.0 | 13578.0 |
| EU (London) | 2.952 | 1358.0 | 16294.0 |
| EU (Paris) | 2.755 | 1267.0 | 15207.0 |
| EU (Stockholm) | 2.165 | 996.0 | 11949.0 |
| US East (N. Virginia) | 1.968 | 905.0 | 10862.0 |
| US East (Ohio) | 1.968 | 905.0 | 10862.0 |
| US West (Oregon) | 1.968 | 905.0 | 10862.0 |
| Zone | Hourly | Monthly | Yearly |
|---|---|---|---|
| Asia Pacific (Mumbai) | 2.904 | 1429.0 | 17143.0 |
| Asia Pacific (Singapore) | 3.101 | 1519.0 | 18230.0 |
| EU (Frankfurt) | 3.297 | 1610.0 | 19316.0 |
| EU (Ireland) | 2.805 | 1383.0 | 16600.0 |
| EU (London) | 3.297 | 1610.0 | 19316.0 |
| EU (Paris) | 3.101 | 1519.0 | 18230.0 |
| EU (Stockholm) | 2.510 | 1248.0 | 14971.0 |
| US East (N. Virginia) | 2.313 | 1157.0 | 13885.0 |
| US East (Ohio) | 2.313 | 1157.0 | 13885.0 |
| US West (Oregon) | 2.313 | 1157.0 | 13885.0 |
Pricing in USD. Monthly is calculated from "hourly" . Yearly is calculated from "Yearly no upfront" .