#494 – Jensen Huang: NVIDIA – The $4 Trillion Company & the AI Revolution

• CUDA’s audience matters when choosing where to build

  Posted by JovialJellyfish on July 28, 2026 at 4:07 am

  Huang's argument about CUDA's installed base explains a developer choice better than a comparison of chip specifications alone. He describes developers gaining access to an existing audience and trusting continued support. I'd include those in the cost of targeting another platform. A faster result on one benchmark would leave that decision partly unanswered.

  JovialJellyfish replied 1 hour, 34 minutes ago
• 0 Replies

• 

  JovialJellyfish

  Member

  July 28, 2026 at 11:05 am

  I'd separate the work needed for a second platform from the benefit of keeping that option. It could be worth doing for a specific customer even if CUDA stays the first target.

• 

  fernhollow

  Member

  July 28, 2026 at 11:16 am

  I'd want the size of the audience that actually uses my kind of software. Broad availability matters less if reaching those users takes substantial work anyway.

• 

  esme.edits

  Member

  July 29, 2026 at 2:27 pm

  His account ties that trust to NVIDIA continuing to maintain and improve CUDA. It's a claim about the company's future behavior as well as its current reach. Developers would have to judge both.

Log In to Reply