AI BriefingNvidia and Uber Warn of Surging AI Compute Costs
Key point
Executives at Nvidia and Uber pointed to the problem of rapidly rising AI compute and token costs.
Details
Nvidia's VP of Applied Deep Learning revealed that his team's compute costs have now surpassed labor costs. This suggests that even a company leading the AI chip market is seeing infrastructure operating costs exceed human resource costs.
Uber's case starkly illustrates this cost pressure. According to Uber's CTO, the 2026 AI coding budget was already fully depleted by this past April.
The main sources of cost are as follows:
- Token costs alone amount to $500 to $2,000 per engineer per month (excluding licensing and hardware costs).
- Massive inference costs arising from simple prompt inputs alone.
This phenomenon is raising fundamental questions about whether the current token-based pricing model is sustainable for enterprises adopting and operating AI at scale.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.
Related
Finnish AI Cloud Startup Verda Raises $189 Million in Series B
NVIDIA Releases Open-Source LLM Router 'Switchyard'… Up to 30% Cost Savings Compared to OpusNVIDIA Expands AI Infrastructure Financing Guarantees... Strengthening Role as 'Central Bank of AI'