Cloud cost optimisation predictions for 2022

Chip shortages and COVID-19 restrictions continue to slow cloud development.

Cloud spending however continues to expand, according to Canalys, with 35% year on year growth, to $49.4 Billion in Q3 2021. This is driven by more businesses moving to the cloud, and the spending expanding for those already in the cloud.

Jon Bryany, Lead Cloud FinOps Consultant at Strategic Blue

Wednesday, February 02, 2022 | 3 MINS

AWS & Google Cloud predictions

There has been a change in the cloud industry in late 2021. With both AWS and Google now talking much more about price-performance than purely the compute price. This follows the AWS launch of the 6i families and Google’s launch on the T2D compute instances. This means that for the first time there’s no immediate cost benefit to upgrade to the latest generation. For lots of small regular workloads, the performance benefits are not significant enough to justify the effort of the upgrade.

AWS has released its new Graviton 3 processor-based instances, the C7g instance, into preview and we can expect a lot of Graviton 3 based services to arrive throughout the year. Particularly any of the growing serverless solutions which continue to aid cloud adoption.

The AMD Milan-based instance, the AWS M6a, has finally been released in some regions. We expect more families and regions to be available throughout the year. We expect this to be a significant challenge to the 6i families on both performance and price.

Google is working closely with AMD and is pushing the T2D instances for high-performance scale-out workloads, but the choice between the N2D instances and T2D can be complex and depend on performance needs and usage profile.

Last year I predicted an AMD CPU & GPU combo, using the latest generation CPU and GPUs for HPC – it hasn’t happened yet but perhaps 2022 is the year AMD starts making significant in-roads into the accelerated cloud market.