F1 instances support field programmable gate arrays (FPGAs). There are two models of F1, as described in the following table:
Model |
GPUs |
vCPU |
Memory(GB) |
SSD Storage (GB) |
f1.2xlarge |
1 |
8 |
122 |
470 |
f1.16xlarge |
8 |
64 |
976 |
4 x 940 |
FPGAs with high memory and computing power are very promising in the field of deep learning. However, TensorFlow and other popular deep learning libraries don't support FPGAs. Therefore, in the next section, we will only cover the prices of P2 and G2 instances.