g6.48xlarge

Analysis

The g6.48xlarge is a gpu instance instance in the g6 series with 192 vCPUs and 768 GiB of memory on AMD EPYC 7R13 (up to 2.6 GHz), which makes it balanced at 4 GiB of memory per vCPU.

It is 18% cheaper than the older g5.48xlarge on AMD EPYC 7R32, so staying on the previous generation costs more as well as being slower. The newer g7.48xlarge on Intel Xeon Scalable (Granite Rapids) is 114% dearer at $28.51/hour.

Running it in N. Virginia instead of São Paulo saves $6,820 a month per instance: $13.35/hour against $22.69/hour, a 70% spread. It is not offered in Ireland and Singapore, so plan around that if you need a presence there.

It carries 8 GPUs with 192 GB of GPU memory, so most of the price is the accelerator rather than the CPUs. It has local instance storage (8 x 940 GB NVMe SSD), which is fast but lost when the instance stops. Network: 100 Gigabit.

It is the cheapest of the 2 processor variants AWS sells at this size and generation. The g6e.48xlarge on AMD EPYC 7R13 costs 126% more, and no others also sit above it.

It is the largest of the 8 g6 sizes. Dropping to the g6.24xlarge saves $4,873 a month. Unit price is not flat across the series ($0.0531 to $0.2012 per vCPU-hour), so the size changes the rate you pay.