SAN FRANCISCO — Uber is deepening its reliance on Amazon Web Services by expanding a contract to run additional ride-sharing features on AWS’s custom AI chips. The agreement signals a strategic shift as the ride-hailing giant moves more of its real-time systems from traditional processors to specialized hardware designed for machine learning workloads.
The deal covers expanded use of AWS Trainium and AWS Inferentia chips. These chips are built to accelerate training and inference tasks, which are critical for Uber’s dynamic pricing, route optimization and fraud detection systems. By integrating them into its infrastructure, Uber aims to reduce latency and improve efficiency in services that depend on rapid data processing.
The decision comes as Uber looks to reduce costs and improve performance across its platform. The company has been migrating parts of its backend to AWS over the past two years, but this expansion marks a stronger commitment to Amazon’s AI ecosystem. It also represents a direct challenge to competitors like Oracle Cloud and Google Cloud, both of which have been vying for Uber’s business in recent years.
Analysts say the move reflects a broader trend among large enterprises to adopt purpose-built chips for AI workloads. AWS launched Trainium in 2021 and Inferentia in 2019, targeting companies with heavy machine learning needs. Uber’s adoption could influence other tech firms to follow suit, especially in industries where real-time decision-making is essential.
Uber has not disclosed the financial terms of the expanded contract. A company spokesperson confirmed the partnership but declined to provide further details on which services will be migrated first or how the chips will be deployed.
Source: techcrunch.com