dl2q.24xlarge vs p4de.24xlarge
| dl2q.24xlarge | p4de.24xlarge |
| Machine Learning ASIC | GPU instance |
| Deep learning / Machine Learning (ML) inference and training dl – Deep Learning 2 – Generation q – Qualcomm inference accelerator 24xlarge – Size | GPU intensive tasks like machine/deep learning, high-performance computing p – Performance / GPU (Graphics Processing Unit) accelerated 4 – Generation d – Instance store volume e – Extra storage or memory 24xlarge – Size |
| 96 | 96 |
| 768 GiB | 1152 GiB |
| n/a | x86_64 |
| Intel Xeon Platinum 8259 (Cascade Lake) | Intel Xeon Platinum 8275L |
| 3 GHz | |
| 0 | 2 |
| no | yes |
| A100 | |
| NVIDIA | |
| 8 | 8 |
| 80 GiB | |
| 640 GiB | |
| no | no |
| 128 GiB | 6402 GiB |
| no | no |
| dl2q.24xlarge | p4de.24xlarge |
| p4d.24xlarge | |
| ml.p4de.24xlarge | |
| no | no |
| no | no |
| no | no |
| 345 | |
| EBS only | 8 x 1000 SSD |
| default | |
| supported | |
| required | |
| 19000 MBps | 19000 MBps |
| 19000 Mbps | |
| 2375 Mbps | |
| 80000 IOPS | |
| 80000 IOPS | |
| 19000 Mbps | |
| 2375 Mbps | |
| 100 Gigabit | 400 Gigabit |
| yes | no |
| 60 | |
| 4 | |
| 50 | |
| 50 | |
| no | yes |
| required | |
| no | yes |
| 4 | |
| no | yes |
| no | no |
| 1.00x | 3.17x |
Regional Prices
Geography | Region | dl2q.24xlarge | p4de.24xlarge |
|---|---|---|---|
| Asia Pacific | Asia Pacific (Singapore) (ap-southeast-1) | n/a | 32.9365 |
| Europe | Europe (Frankfurt) (eu-central-1) | 11.5952 | n/a |
| Hidden | Hidden (hidden-1) | n/a | 34.2912 |
| Hidden | Hidden (hidden-2) | n/a | 35.5735 |
| Hidden | Hidden (hidden-3) | n/a | 37.6223 |
| US | US East (Virginia) (us-east-1) | n/a | 27.4471 |
| US | US West (Oregon) (us-west-2) | 8.9194 | 27.4471 |