Understanding AWS Graviton Playbook

I was watching AWS Re-invent Keynote yesterday, and one of the things that stood out to me was the below quote:

“For the third year in a row, more than half of the CPU capacity that we’ve added to the AWS cloud comes from Graviton”

While this isn’t a newly disclosed info (they shared the same data in last year’s keynote), the continued success of Graviton deserves a deeper attention. To contextualize how big of a success Graviton has been, let me share a quote from 2024 keynote:

Graviton is growing like crazy. Let’s put this into context. In 2019, all of AWS was a $35 billion business. Today, there’s as much Graviton running in the AWS fleet as all compute in 2019.

It is indeed kind of crazy to think that Graviton was only launched in November 2018, and in just 5-6 years, it has become such an integral part of AWS offerings! In some estimates, Graviton was ~20% of AWS CPU usage by mid-2022, and given the added capacity in the last three years, it may be close to ~30% of CPU usage today. You can tell how AWS feels quite emboldened by Graviton’s success and they very much want to replicate this success beyond the realm of CPUs.

I will explore a brief history of Graviton and elaborate on Amazon’s desire to replicate and constraints it may face in repeating Graviton’s success behind the paywall.

In addition to “Daily Dose” (yes, DAILY) like this, MBI Deep Dives publishes one Deep Dive on a publicly listed company every month. You can find all the 65 Deep Dives here.