On August 28, the Wall Street Journal, citing sources with inside knowledge, reported that Nvidia (NVDA.O) has put on hold a collaborative initiative with AI cloud service providers. This move came less than two months after the program's initial announcement. Concerns have arisen among some Nvidia staff that the plan could potentially draw antitrust scrutiny, given the company's possible over-involvement in its clients' business operations. The initiative, dubbed the 'AI Computing Power Partnership Program,' entailed Nvidia's commitment to lease GPU computing resources from cloud service providers in instances where they struggled to secure alternative clients. This arrangement was designed to safeguard their revenue streams and aid in financing. Nvidia revealed this week that such agreements, which generally span six years, encompass commitments amounting to $36 billion. As per the agreement, once the cloud service provider's revenue surpasses a foundational threshold (which accounts for expenses like chip depreciation, data center maintenance, and personnel costs), Nvidia is eligible to receive 50% of the surplus. Looking ahead, Nvidia may opt to modify this program or incorporate it into other ventures.