Google Cloud Provides Dedicated Nvidia GPU Cluster for Y Combinator Startups

Google Cloud is partnering with Y Combinator to offer startups a dedicated and subsidized cluster of Nvidia GPUs and Google Tensor Processing Units (TPUs) to facilitate AI model development.

This initiative aims to build strong relationships with early-stage AI startups, hoping these startups will grow into substantial, compute-intensive businesses.

James Lee, General Manager for Google Cloud’s Startups and AI business, emphasized the strategy: "We want to offer early support to these startups so they become accustomed to building and working on the Google Cloud Platform. As they grow, we grow alongside them, becoming a crucial partner throughout their lifecycle."

For the Summer 2024 batch, Y Combinator startups will receive priority access to this cluster and $350,000 in cloud credits over two years. Diana Hu, YC Group Partner, highlighted that this investment is a strategic bet, given that 5% of Y Combinator startups over the past 18 years have achieved unicorn status.

Hu added, “There’s significant excitement about the next generation of startups potentially becoming decacorns. Cloud providers are recognizing the need to engage early to support their growth.”

This partnership also aims to attract more AI startups to Y Combinator by addressing a common challenge: limited compute resources. Large enterprises can secure extensive GPU access through long-term deals, but small startups often struggle. The dedicated cluster provided by Google Cloud will be particularly beneficial for training AI models, which require high-performance computing workloads in bursts.

Google Cloud will further support YC startups with $12,000 in Enhanced Support credits, a free year of Google Workspace Business Plus, and access to Google’s AI experts through monthly office hours. The exact number of GPUs in the deal was not disclosed, but it is designed to meet the needs of YC’s foundation model companies.