Google's $750M partner push

Google Cloud is putting $750 million behind partners that build and deploy artificial intelligence agents for big companies. (cloud.google.com) Google announced the fund on April 22 at Cloud Next 2026, saying it will apply across consulting firms, software partners and channel partners. The company said the money will support agent development, deployment incentives, workshops and teams of Google “forward deployed engineers” working with major systems integrators. (cloud.google.com) Bloomberg reported the push is aimed at firms including McKinsey, Accenture and Deloitte, while Google’s own partner blog highlighted deeper technical work with global consulting firms and new ways to surface partner-built agents inside Gemini Enterprise. (bloomberg.com) (cloud.google.com) An AI agent is software that can carry out multistep tasks with limited human prompting, and the expensive part for corporate customers is often not the model but the integration work. Google cited a partner study last year saying its partners can capture up to $7.05 in extra revenue for every $1 a customer spends on Google Cloud. (cloud.google.com) That helps explain why Google paired the fund with new infrastructure announcements. In its Cloud Next 2026 roundup, Google said the event centered on the Gemini Enterprise Agent Platform and its eighth-generation Tensor Processing Units, or TPUs, which are custom chips for training and running AI models. (blog.google) Google said its new TPU line now splits into TPU 8t for training and TPU 8i for inference, the step where a trained model generates answers in real time. The company said both chips are part of its AI Hypercomputer system and are slated for general availability later in 2026. (blog.google) (cloud.google.com) That chip split follows last year’s Ironwood launch, when Google introduced its seventh-generation TPU as its first design built specifically for inference. Google said on April 23, 2025 that Ironwood could scale to 9,216 chips and was designed for what it called the “age of inference.” (blog.google) Google is making the case that the market is shifting from selling raw model access to selling full systems that include chips, cloud software and consulting help. Google said nearly 75% of Google Cloud customers now use its artificial intelligence products, and 330 customers processed more than 1 trillion tokens each over the past 12 months. (blog.google) The backdrop is a cloud market still led by Amazon Web Services and Microsoft, with Google trying to gain share while enterprise spending rises. Synergy Research data published in February showed AWS at 29%, Microsoft at 22% and Google at 12% of the global cloud infrastructure market in the fourth quarter of 2025. (crn.com) (datacenterdynamics.com) For Google, the bet is that the winners in enterprise AI will not just supply chips or models. They will also pay the firms that install the agents inside finance, customer service and operations systems. (cloud.google.com)