Nvidia to Launch New AI Inference Chip at GTC 2026
Nvidia is set to unveil a new AI 'inference' chip at its GTC 2026 conference in San Jose, California, as part of a strategy to address increasing competition in the AI sector. CEO Jensen Huang's keynote is expected to highlight the shift in industry focus from training to deploying AI models, with growing demand for cost-effective chips. The company has also formed a nonexclusive agreement with Groq to utilize its inferencing technology, which is designed to enhance efficiency in running AI models.
US Withdraws Proposed Rule on AI Chip Exports
The U.S. Department of Commerce has withdrawn a proposed rule that would have required global permits for the export of artificial intelligence chips, as indicated by a notification on a government website. The draft, which aimed to regulate worldwide access to these technologies, was initially circulated for feedback in February but has now been retracted without further explanation.
Uber Relaunches Robotaxi Service in Las Vegas
Uber Technologies Inc. has relaunched its commercial robotaxi service in Las Vegas in partnership with Motional, a Hyundai Motor Co.-backed company, starting March 13, 2026. Riders requesting UberX, Uber Electric, or Uber Comfort may be matched with a Motional Ioniq 5 autonomous vehicle, which will initially have a safety driver present, with plans to remove the operator by the end of the year. The service will operate in designated areas along the Las Vegas Strip and aims to expand its coverage in the future.
Amazon to Enhance AI with Cerebras Chip Integration
Amazon.com Inc. will integrate chips from Cerebras Systems Inc. with its own Trainium processors to enhance AI software performance, as announced by Amazon Web Services (AWS). This collaboration aims to deliver faster AI inference solutions for generative AI applications and large language model workloads, with the new service expected to launch in the second half of 2026. The solution will utilize AWS's Elastic Fabric Adapter networking to optimize the computational processes involved in AI inference.
