atlas
p3…
Try

3 of 814 instances · cheapest is p3.2xlarge at $2,457/mo

Instance Workload vCPUs Memory Local disk Network On-demand/mo ↑ Per vCPU
p3.2xlargeGPU & accelerators 8 61 GiB EBS only Up to 10 Gbps $2,457 $307
p3.8xlargeGPU & accelerators 32 244 GiB EBS only 10 Gbps $9,829 $307
p3.16xlargeGPU & accelerators 64 488 GiB EBS only 25 Gbps $19,657 $307