NVIDIA Swaps ‘Compute Resources’ for ‘Future Profits’! Launches New Startup Support Strategy ‘Revenue Share’
What Happened? Overview of the News
- Revenue Sharing Partnerships: NVIDIA is entering into revenue share agreements with rapidly growing startups, offering them development “token credits (compute resources)” in exchange for a portion of future product profits and cloud revenues.
- Securing Large-Scale Infrastructure: Two initial partners announced are Australia’s Sharon AI (which will deploy up to 40,000 GPUs) and Singapore’s Firmus Technologies (which plans to build a 360MW data center in Indonesia capable of housing up to 170,000 GPUs).
- NVIDIA as a Mediator: NVIDIA is positioning itself as a mediator, ensuring model builders and cloud AI companies can directly access full-stack computing powered by NVIDIA chips.
Why Is This Important? Key Points to Note
- ‘Compute Resources as Oil’: For AI startups, access to scarce GPU resources is crucial. Previously, securing these resources was often a challenge due to liquidity issues. This new structure represents an innovative form of funding that trades “future success” instead of cash.
- Accelerating Vertical Integration: NVIDIA is transforming from a mere chip supplier into a platform that digs deep into the revenue streams of startups.
- Addressing Market Liquidity Issues: With companies like OpenAI already beginning to establish equity and revenue-sharing agreements with chip manufacturers, NVIDIA formalizing this large-scale program is a significant development.
🦈 Shark’s Eye (Curator’s Perspective)
The reality is that GPUs are becoming a currency in today’s world! If you’ve got the confidence that “our model will hit big!” you can tap into NVIDIA’s powerful compute resources without cash. This setup is a game-changer for startups. The scale of Firmus Technologies’ plan to build a “170,000 GPU data center” in Indonesia is simply mind-blowing! With such firepower, NVIDIA aims to solidify its status as both the “landlord” and “investor” for startups. I can’t wait to see innovative AI models that have been held back due to liquidity issues finally hit the market because of this scheme!
What’s Next?
- Following Suit from Other Manufacturers: Competitors like AMD and Amazon are likely to ramp up similar revenue-sharing schemes, intensifying the competition for promising startups.
- Geographic Expansion of Infrastructure: NVIDIA’s moves, aiming for $20 billion in funding, should accelerate the establishment of large-scale AI infrastructure in new hubs like Batam Island, Indonesia.
A Final Word from Haru Shark
“Borrowing chips with future profits” is exactly the kind of dynamic business that screams 2026! Now’s the time to bite! 🦈🔥
Terminology Explained
-
Token Credits: Virtual rights provided by NVIDIA to utilize GPU compute resources. Startups will consume these credits for AI development and operations.
-
Revenue Share: A system where multiple companies collaborate on a project and share the resulting profits according to a pre-determined ratio.
-
Full-Stack Computing: Providing an integrated solution that encompasses everything necessary for computation, from chips (hardware) to software and networking.
-
Source: Nvidia offers startup customers chance to swap compute power for revenue share