New revenue-sharing model for AI startups
Nvidia has officially introduced a revenue-sharing and credit support program designed to lower the barrier to entry for AI startups and large model developers. This strategic shift allows emerging companies to access high-performance GPU computing power through partner cloud providers without the immediate capital burden of building their own data centers.
Under this initiative, Nvidia supplies the hardware, software platforms, and credit support, while partners convert these resources into cloud-based capacity. In exchange, Nvidia receives traditional hardware sales revenue while also securing a portion of the cloud service revenue, establishing a recurring income stream. The company has already announced its first two partners: Australia-based Sharon AI, which will deploy up to 40,000 Grace Blackwell GB300 GPUs, and Firmus Technologies, which is constructing a 360-megawatt AI factory in Batam, Indonesia, with plans to house up to 170,000 Nvidia GPUs.
Executive leadership transition
Coinciding with this operational expansion, Nvidia has recruited Nick Parker, a 26-year Microsoft veteran, to serve as its new executive vice president of worldwide field operations. Parker, who previously led Microsoft’s worldwide commercial sales business, will report directly to Nvidia CEO Jensen Huang effective August 24.
The role comes with a significant compensation package, including a $1 million base salary, a $5 million signing bonus, and equity grants valued at $40 million. Parker succeeds Jay Puri, who is retiring after 21 years of leading Nvidia’s global sales operations. The move is viewed as a high-profile shift between two of the most significant players in the global artificial intelligence sector.
Technical infrastructure updates
Beyond its commercial and leadership shifts, Nvidia continues to integrate its latest hardware into open-source ecosystems. The company recently submitted Linux kernel patches for the Device Tree of its Vera Rubin VR-NVL server platform. This effort supports the integration of OpenBMC—an open-source baseboard management controller software—onto its latest rack-scale supercomputers, signaling a continued commitment to industry-standard infrastructure management.