atlas
p2…
Try

3 of 1348 instances · cheapest is p2.xlarge at $657/mo

Instance Workload vCPUs Memory Local disk Network On-demand/mo ↑ Per vCPU
p2.xlargeGPU & accelerators 4 61 GiB EBS only High $657 $164
p2.8xlargeGPU & accelerators 32 488 GiB EBS only 10 Gbps $5,256 $164
p2.16xlargeGPU & accelerators 64 732 GiB EBS only 20 Gbps $10,512 $164