inf2.xlarge vs p2.xlarge
| inf2.xlarge | p2.xlarge |
| Machine Learning ASIC | GPU instance |
| Machine Learning (ML) inference inf – AWS Inferentia 2 – Generation xlarge – Size | GPU intensive tasks like machine/deep learning, high-performance computing p – Performance / GPU (Graphics Processing Unit) accelerated 2 – Generation xlarge – Size |
| 4 | 4 |
| 16 GiB | 61 GiB |
| x86_64 | n/a |
| AMD EPYC 7R13 Processor | Intel Xeon E5-2686 v4 (Broadwell) |
| 3.6 GHz | |
| 2 | 0 |
| no | no |
| 1 | 1 |
| yes | no |
| Inferentia2 | |
| AWS | |
| 1 | |
| 32 GiB | |
| no | no |
| inf2.24xlarge, inf2.48xlarge, inf2.8xlarge, inf2.xlarge | p2.16xlarge, p2.8xlarge, p2.xlarge |
| inf1.xlarge, inf2.xlarge | |
| ml.inf2.xlarge | |
| no | no |
| no | no |
| no | no |
| 16 | |
| EBS only | EBS only |
| default | |
| supported | |
| required | |
| up to 10240 MBps | 750 MBps |
| 1250 Mbps | |
| 156.25 Mbps | |
| 6000 IOPS | |
| 40000 IOPS | |
| 10000 Mbps | |
| 1250 Mbps | |
| Up to 15 Gigabit | High |
| yes | yes |
| 4 | |
| 1 | |
| 15 | |
| 15 | |
| yes | no |
| required | |
| no | no |
| yes | no |
| no | no |
| 1.00x | 1.30x |
Regional Prices
Geography | Region | inf2.xlarge | p2.xlarge |
|---|---|---|---|
| Hidden | Hidden (hidden-1) | 0.8340 | 0.9720 |
| Hidden | Hidden (hidden-10) | 1.1373 | n/a |
| Hidden | Hidden (hidden-11) | 1.2889 | n/a |
| Hidden | Hidden (hidden-2) | 0.9098 | 1.0800 |
| Hidden | Hidden (hidden-3) | 0.9478 | 1.3260 |
| Hidden | Hidden (hidden-4) | 0.9857 | 1.4650 |
| Hidden | Hidden (hidden-5) | 0.9857 | 1.5420 |
| Hidden | Hidden (hidden-6) | 1.0615 | 1.5420 |
| Hidden | Hidden (hidden-7) | 1.0615 | 1.7180 |
| Hidden | Hidden (hidden-8) | 1.1373 | 1.7180 |
| Hidden | Hidden (hidden-9) | 1.1373 | n/a |
| US | US East (Ohio) (us-east-2) | 0.7582 | 0.9000 |
| US | US East (Virginia) (us-east-1) | 0.7582 | 0.9000 |
| US | US West (Oregon) (us-west-2) | 0.7582 | 0.9000 |
Related Comparisons
Other instances worth weighing against these.