Google Cloud has announced a new initiative to assist startups and small businesses in creating generative AI applications, leveraging partnerships with Nvidia's Inception program and Google for Startups Cloud Program.

The initiative offers a combination of cloud credits, go-to-market support, and technical expertise specifically for AI-focused startups.

Google unveiled Axion, a custom Arm-based chip for data centers, which promises up to 30% improved performance and 60% greater energy efficiency over similar x86-based chips.

Axion is based on Arm's Neoverse V2 design and will be deployed in large-scale pods of 8,960 chips, available to Google Cloud customers later in 2024.

Google's AI software services have been updated with the introduction of Gemma, a suite of open models, and enhanced support for Nvidia's NeMo framework to facilitate building custom generative AI applications.

Companies such as Broadcom, CrowdStrike, Snap, Datadog, and Elastic have shown interest in Axion, indicating industry support for Google's latest technology.

These announcements highlight Google's efforts to mitigate challenges such as chip shortages, high energy costs, and the financial demands of AI, placing it in direct competition with industry leaders like NVIDIA, Intel, and AMD in the data center processor market.