c8a.16xlarge

Analysis

The c8a.16xlarge is a compute optimized instance in the c8a series with 64 vCPUs and 128 GiB of memory on AMD EPYC 9R45 (up to Up to 4.5 GHz), which makes it compute-heavy at 2 GiB of memory per vCPU.

Within the same generation and size, the c8g.16xlarge on AWS Graviton4 costs $2.55/hour, 26% less ($654 a month), though switching means rebuilding software for arm64. The other variants are the c8i-flex.16xlarge at $2.85, the c8i.16xlarge at $3.00 and the c8gd.16xlarge at $3.14.

Running it in N. Virginia instead of Tokyo saves $652 a month per instance: $3.45/hour against $4.34/hour, a 26% spread. It is not offered in Canada Central, São Paulo, London, Mumbai, Singapore and Sydney, so plan around that if you need a presence there.

On demand it costs $3.45/hour in US East (N. Virginia), $2,518 a month if left running, which puts it mid-table for compute optimized types at the 69th percentile per vCPU ($0.0539 per vCPU-hour) and the 73rd percentile per GiB ($0.0269 per GiB-hour).

Looking beyond its own line, 20 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.

Compared with the previous generation, it costs 5.0% more than the c7a.16xlarge on AMD EPYC 9R14 ($120 more a month per instance), in exchange for the newer processor.