Source item

Jensen brings in the Groq acquisition related to higher priced tokens. Says the economics of generating tokens is not a race to the bottom. He argues it’s a race to the top, the smartest models users will pay the most for. Token pricing will be increasingly segmented into tiers. Tier 1: High throughput and low latency (often free or low-cost). Tier 2: Larger models, specialized pricing. Tier "N": The largest, smartest models. The more input token context, the smarter the models, the price goes up. This will be Nvidia's most advanced GPU ,Groq 3 into the Vera Rubin platform My take: Makes sense. Not all AI is created equal, and premium tokens will in part be powered by Groq. Application providers will pay for different tiers depending on the task. Groq should be 25% of data center.