AWS continues its emphatic endorsement of Arm with its Graviton chips
AWS is the dominant deployer of Arm-based instances, with a 57.2% share of Arm instances in the public cloud. While its initial focus was to drive costs down for low-end workloads, AWS has expanded its Arm capabilities over the years, directly benefiting enterprises in multiple ways. Just before Liftr began tracking AWS in 2019, AWS was the first cloud provider to make Arm processors available. But market longevity is only one part of the story.
An introduction to Graviton
Graviton is AWS’ private-label Arm processor that provides the lowest-cost instances for price-sensitive customers. Now in its fourth generation, AWS offers multiple customer options depending on workload, pricing, and performance requirements. Additionally, AWS provides several web-based tools to help customers understand the savings potential and migration requirements and overcome concerns about switching to Graviton. We have been tracking Graviton since 2019 and offer independent data points on its adoption and price benefits to customers as well.
Growth of Arm at AWS
A history of growth
As of December 2024, Graviton represented 25.0% of the total instances deployed by AWS. In January 2025, that percentage share was even higher.
AWS has made the most of its first-mover status in Arm. While growth ramped slowly, Liftr history shows a shift in momentum in mid-2020 that aligns with the release of the second-generation Graviton processors. Graviton has grown steadily over the past four years. As of December 2024, Graviton represented 25.0% of the total instances deployed by AWS. In January 2025, that percentage share was even higher. In fact, Graviton represents a greater share of AWS instances vs. x86 vendor AMD since the first quarter of 2021.
Workload expansion
The first generation of Graviton was targeted at low performance and low-cost instances such as development environments or web servers. However, different Graviton generations and configurations have led to 9 distinct instance types for the widest range of Arm workloads.
Count of Workloads by Graviton Generation
Sample of Workloads by Generation
Deployment of Graviton by Generation and Workload Type as of Jan. 31, 2025
According to Liftr data as of January 2025, generation 2 Graviton is by far the most widespread generation with the most instance types. However, Generation 4 was announced in the 4th quarter of 2024 and supports the widest range of instance types. It's probably not a coincidence that this Generation 4 announcement occurred after AWS deployed more Graviton instances in the 3rd and 4th quarter of calendar 2024 than Intel or AMD (54% and 40% of each quarters new deployments, respectively).
Price / Performance advantage
Like other Arm-based instances from other cloud providers such as Azure, the biggest customer value is lower price vs. comparable Intel- or AMD-based instances. In this case, we can see slight savings (5% to 10%) by comparing the latest generation Graviton, AMD and Intel general purpose instances.
For example:
In other situations, Liftr Insights has seen savings of up to 50% between Arm and x86. While the example above may not be the same differential, it should be noted that the M8G instance is the only Arm instance on AWS available in a 1 VCPU configuration for very low-end workloads.
Additionally, AWS offers some of its cloud-based managed services on Graviton processors, such as databases (Aurora) or Kubernetes (KBS), driving down the total cost of ownership beyond the cost of infrastructure.
The biggest customer value is lower price vs. comparable Intel- or AMD-based instances.
Sustainability Advantage
Arm processors also are known for lower energy consumption and AWS claims up to 60% less energy than comparable instances. Sustainability goals and reporting are important to enterprise ESG metrics.
Conclusion
AWS was the first to deploy Arm in the cloud and prove that lower cost, lower energy cloud instances benefit enterprise customers. But, AWS has also taken its momentum into different directions by offering a diverse range of solutions for Graviton to go beyond the traditional Arm positioning and presenting a viable competitor to x86 architectures across many types of workloads. This combination of initiatives has led to Graviton to represent a significant footprint in AWS and launch competitive Arm solutions from Microsoft and Google, which will be discussed in the next article.
Enterprises relying on compute instances and managing their bottom line and sustainability goals can keep up-to-date with Liftr data for their forecasting and planning needs. A simple solution is one of Liftr's Supporting Data for Global AI & Cloud reports.
Feel current and informed with the first notice of the latest reports like these; subscribe to the Liftr newsletter.