Uber Bets Big on Amazon's Custom AI Chips to Power Faster Rides

Uber is deepening its cloud partnership with Amazon Web Services (AWS) by deploying Amazon's custom-designed silicon to power its real-time infrastructure and artificial intelligence initiatives. Seeking a competitive edge in speed and efficiency, the ride-hailing giant will use AWS Graviton4 processors to handle its "Trip Serving Zones," reducing latency and energy consumption while matching drivers and riders in mere milliseconds. Additionally, Uber is piloting AWS Trainium3 chips to train advanced AI models for route optimization, dynamic pricing, and personalized user experiences. This move marks a significant victory for Amazon's proprietary hardware business, highlighting a broader industry trend where major tech companies are adopting highly specialized, cost-effective cloud silicon over traditional general-purpose computing.

Read the original article on TechCrunch here