m8a.24xlarge

Analysis

The m8a.24xlarge is a general purpose instance in the m8a series with 96 vCPUs and 384 GiB of memory on AMD EPYC 9R45 (up to Up to 4.5 GHz), which makes it balanced at 4 GiB of memory per vCPU.

Within the same generation and size, the m8g.24xlarge on AWS Graviton4 costs $4.31/hour, 26% less ($1,120 a month), though switching means rebuilding software for arm64. The other variants are the m8i.24xlarge at $5.08, the m8gd.24xlarge at $5.54 and the m8id.24xlarge at $6.26.

Running it in N. Virginia instead of Tokyo saves $1,244 a month per instance: $5.84/hour against $7.55/hour, a 29% spread. It is not offered in Canada Central, São Paulo, London, Singapore and Sydney, so plan around that if you need a presence there.

On demand it costs $5.84/hour in US East (N. Virginia), $4,265 a month if left running, which puts it mid-table for general purpose types at the 62nd percentile per vCPU ($0.0609 per vCPU-hour) and the 62nd percentile per GiB ($0.0152 per GiB-hour).

Looking beyond its own line, 15 of 28 current types with 96 vCPUs and similar memory are cheaper. The cheapest of those is the hpc6a.48xlarge on AMD EPYC 7R13 at $2.88/hour.

Compared with the previous generation, it costs 5.0% more than the m7a.24xlarge on AMD EPYC 9R14 ($203 more a month per instance), in exchange for the newer processor.