In a recent analysis, Carmen Li explains why GPUs, due to their term structure, are well-suited for futures contracts. The structured nature of GPU supply and demand aligns with the fixed-term nature of futures, making them a natural fit for such financial instruments. Conversely, large language model (LLM) tokens are highlighted as more appropriate for perpetual contracts. The continuous and evolving nature of LLM tokens aligns with the perpetual market's need for flexibility and ongoing adjustments. This distinction underscores the importance of matching asset characteristics with the appropriate financial instruments for optimal trading strategies.