Performance and cost optimization in cloud infrastructures through migration to ARM-based architectures

In the early 2020s, cloud computing entered a phase where rapidly rising demand for compute – driven by data-intensive analytics and the emerging wave of generative AI – collides with tighter constraints on cost, energy efficiency, and sustainability. Gartner forecasts that worldwide end-user spending on public cloud services will grow 21. 7% to 597. 3 billion in 2023 (from 491 billion in 2022), making infrastructure optimization a prerequisite for scalable operations rather than only a competitiveness factor. Against this background, the long-dominant x86 paradigm is increasingly complemented by Arm-based server platforms that emphasize efficiency and scale-out density. This study assesses the feasibility and expected benefits of migrating general-purpose cloud workloads from x86 to Arm using the 2022-2023 commercial landscape: AWS EC2 C7g (Graviton3), Google Cloud Tau T2A (Arm-based), and Azure Arm virtual machines (Ampere Altra). The work synthesizes performance and operational considerations and develops a migration-oriented methodology focused on benchmarking, dependency management, and phased rollout. Provider disclosures indicate meaningful efficiency gains (e. g. , Graviton3 claiming up to 25% higher performance than Graviton2 and up to 60% lower energy for comparable performance), supporting Arm as a practical option for mainstream cloud deployments when migration risks are systematically managed.