Companies like Uber, Spotify, and Datadog are already migrating their workloads to Arm-based instances to achieve greater efficiency, performance, and sustainability.
Arm has announced an ambitious initiative aimed at facilitating and accelerating the migration of developers and organizations to cloud platforms based on its architecture. This decision comes against a backdrop of the radical transformation that Artificial Intelligence is imposing on computing infrastructures, where factors such as performance, energy efficiency, and total cost of ownership have become key priorities.
With its Neoverse processor family, Arm has established itself as a benchmark platform for leading cloud providers, such as AWS (Graviton), Google Cloud (Axion), Microsoft Azure (Cobalt 100), and Oracle Cloud Infrastructure (Ampere). All these Arm-based solutions have demonstrated significant improvements in the price-performance ratio.
Leading Companies are Already Migrating to Arm
Numerous companies have begun the process of migrating their services and applications to Arm environments. For example, Uber is moving over 5,000 services to instances based on this architecture, seeking not only to optimize costs but also to reduce its carbon footprint and benefit from new silicon optimized for AI.
Spotify, on the other hand, has reported improvements of up to 250% in the performance of its workloads running on Axion, compared to previous solutions on Google Cloud. Datadog has already migrated 70% of its workloads to AWS Graviton, highlighting a notable improvement in efficiency and scaling capability.
New Resources for Cloud Developers
To facilitate this transition to Arm architecture, the company has introduced a range of new tools and support channels, including:
- Arm Cloud Migration Resource Hub: a centralized portal offering over 100 step-by-step guides for migrating common workloads across multiple cloud platforms. An updated dashboard of the software ecosystem compatible with Arm is also included.
- Community and Expert Support: access to discussion forums, tutorials, webinars, and technical sessions within the Arm Developer Program. Additionally, developers can interact with a specialized cloud migration team focused on Arm.
- Personalized Assistance for Enterprises: teams of Arm engineers available to assist large organizations in their evaluation, migration, and optimization processes.
A More Friendly Environment for AI and Cloud Native
The company has also emphasized the advantages that Arm brings to executing modern workloads, particularly those related to artificial intelligence and microservices. Recent integrations of Arm Kleidi with frameworks like PyTorch allow developers to automatically benefit from significant performance improvements in AI on Arm hardware without the need for additional steps.
Furthermore, Arm has launched a specific extension for GitHub Copilot in order to enhance accessibility to its technology, accelerate development times, and reduce adoption costs.
With a community of over 20 million developers and an expanding ecosystem, Arm is laying the groundwork for a new generation of cloud workloads, encompassing everything from web servers and databases to analytics systems, machine learning, and distributed services.
The company concludes its announcement by highlighting that this is just the beginning of a broader strategy to make Arm the preferred platform for cloud and AI developers. “More updates are on the way,” states the release, in a clear commitment to the sustained growth of the Arm ecosystem in the data centers of the future.
via: ARM blog