The open-source inferencing software increases throughput and reduces the cost of LLM token generation, the chipmaker said.
Amazon’s CEO plans to invest more than $100 billion this year, mostly in AI infrastructure — but could face challenges buying enough hardware and power.
Both OpenAI and Microsoft are reportedly probing the Chinese startup to check if it accessed and used OpenAI’s technology to build its R1 reasoning large language model.
The project is designed to keep the US ahead of competing nations, such as China, in AI.
The acquisition can be seen as part of Lenovo’s strategy to support more AI workloads.
The new data centers are likely to be used to train AI models and deploy AI and cloud-based applications.
Oracle says its new Exadata database compute platform will run generative AI workloads faster for the same price as the old version.
More than half of these data centers are expected to be built in the US, Microsoft President Brad Smith said.
The dashboard is designed to offer insights to make migrating to Graviton-based instances easier.
The new updates include the ability to reach customers for outbound campaigns proactively.
