c5a.16xlarge

Analysis

The c5a.16xlarge is a compute optimized instance in the c5a series with 64 vCPUs and 128 GiB of memory on AMD EPYC 7R32 (up to 3.3 GHz), which makes it compute-heavy at 2 GiB of memory per vCPU.

Running it in Mumbai instead of São Paulo saves $1,659 a month per instance: $1.50/hour against $3.78/hour, a 151% spread.

It is the cheapest of the 2 processor variants AWS sells at this size and generation. The c5ad.16xlarge on AMD EPYC 7R32 costs 12% more, and no others also sit above it.

At $2.46/hour on demand in US East (N. Virginia), about $1,799 a month, the c5a.16xlarge is one of the cheaper ways to buy compute optimized capacity: only 14% of the 291 current compute optimized types cost less per vCPU-hour than its $0.0385.

Reserving it for a year with no upfront payment cuts the Linux rate to $1.55/hour, which beats on demand once it runs more than 63% of the time; three years brings it to $1.06/hour, 57% below on demand. Windows Server roughly doubles the bill at $5.41/hour, because the licence is charged per vCPU.

Looking beyond its own line, 5 of 28 current types with 64 vCPUs and similar memory are cheaper. The cheapest of those is the hpc7g.16xlarge on AWS Graviton3 at $1.68/hour.