atlas
inf1…
Try

4 of 738 instances · cheapest is inf1.xlarge at $275/mo

Instance Workload vCPUs Memory Local disk Network On-demand/mo ↑ Per vCPU
inf1.xlargeGPU & accelerators 4 8 GiB EBS only Up to 25 Gbps $275 $68.80
inf1.2xlargeGPU & accelerators 8 16 GiB EBS only Up to 25 Gbps $437 $54.57
inf1.6xlargeGPU & accelerators 24 48 GiB EBS only 25 Gbps $1,424 $59.31
inf1.24xlargeGPU & accelerators 96 192 GiB EBS only 100 Gbps $5,694 $59.31