CoreWeave became the first cloud provider to deploy Nvidia's GB300 NVL72 platform this week, marking the latest milestone in the chipmaker's push to expand AI computing infrastructure beyond traditional hyperscale customers. The deployment represents Nvidia's continued effort to diversify its customer base as demand for AI processing power spreads across industries and geographies.
The move comes as Nvidia approaches a $4 trillion market valuation, with its stock trading at $159.34 as of Thursday. The company's infrastructure expansion strategy has accelerated in recent months through partnerships spanning cloud providers, enterprises, and national governments.
Nvidia has forged multiple partnerships in recent weeks to broaden its AI infrastructure reach. In June, Oracle expanded its collaboration with Nvidia to help customers develop production-ready AI applications1. Hewlett Packard Enterprise launched new AI factory offerings at its Discover conference, combining Nvidia Blackwell accelerated computing with HPE's infrastructure stack2.
The company also unveiled NVLink Fusion in May, allowing industries to build semi-custom AI infrastructure with partners including MediaTek, Marvell, and Qualcomm Technologies3. "A tectonic shift is underway: for the first time in decades, data centers must be fundamentally rearchitected," said Jensen Huang, Nvidia's founder and CEO3.
According to the Financial Times, these partnerships reflect Nvidia's strategy to attract customers beyond "hyperscalers" like Amazon and Google, which currently make up more than half its data center revenue4. The company is targeting emerging cloud providers such as CoreWeave, Nebius, and Crusoe, as well as enterprise customers managing their own IT infrastructure4.
Nvidia announced major European deployments in June, partnering with governments across France, Italy, Spain, the UK, and Nordic countries to build sovereign AI infrastructure12. The initiatives will deliver more than 3,000 exaflops of Nvidia Blackwell compute resources1.
In France, Nvidia is working with Mistral AI to develop a cloud platform powered by 18,000 Grace Blackwell systems2. Germany will host what the company calls the world's first industrial AI cloud for European manufacturing, while Italy is collaborating with Nvidia to build a sovereign AI platform through supercomputing company Domyn2.
The infrastructure expansion reflects Nvidia's evolution from a chip company to what Huang described as "an essential infrastructure company" at Computex 20251. The company is partnering with telecom operators including Orange, Telefónica, and Swisscom to support national AI infrastructure efforts2.
"I'm more certain about the business opportunity beyond the big cloud providers today than I was a year ago," Huang told the Financial Times in March3.
