Ride-sharing company Uber is scaling up its use of Amazon Web Services (AWS), including adopting the latest generation of its Graviton and Trainium chips.

Under the expanded agreement, Uber will now use AWS Graviton4-powered instances to support its "Trip Serving Zones" and has started pilot training some AI models on the Trainium3 AI chips.

Uber's Trip Serving Zones are part of its system that handles millions of predictions and processes location data for users in milliseconds. By expanding its use of AWS compute, storage, and networking, Uber is aiming to help improve real-time operations for this while reducing energy consumption and reducing latency, to help match drivers to riders faster.

"Uber operates at a scale where milliseconds matter," said Kamran Zargahi, vice president of engineering at Uber. "Moving more Trip Serving workloads to AWS gives us the flexibility to match riders and drivers faster and handle delivery demand spikes without disruption.”

The use of AWS' Trainium3 chips, meanwhile, remains somewhat experimental, trialing the training of some of Uber's AI models that analyze data from billions of rides and deliveries to determine which driver or courier to send, calculate arrival times, and recommend the best delivery options to the customer.

As the model learns from more trips, this will continue to improve.

"Uber is one of the most demanding real-time applications in the world, and we're proud to be an important part of the infrastructure powering their global operations,” said Rich Geraffo, vice president and managing director of North America at AWS.

“We're helping Uber deliver the reliability hundreds of millions of people count on today—and the AI-powered experiences that will define ride-sharing and on-demand delivery tomorrow.”

While Uber is a known customer of AWS, the company has a multicloud strategy, also using Google Cloud and Oracle Cloud Infrastructure. In 2023, the company signed two seven-year deals with Google and OCI, revealing that it would be closing its own data centers.

Prior to that migration, Uber housed some 95 percent of its IT in its own data centers. Among those was a Colorado facility the company bought from Microsoft in 2015.