Injective launches perpetual futures for Nvidia GPU rental fees

Trading the cost of AI compute

Injective has launched a new perpetual futures market that lets people trade the price swings of Nvidia graphics processing units, or GPUs. This financial product is tied to the Super Intelligence Compute Index, which is known as the SICI. It provides a way for people to invest in or manage the risks associated with how much it costs to rent out AI computing power.

The new market allows for up to 10x leverage, meaning investors can take larger positions based on their collateral. This setup enables traders to bet on both the increase and decrease of rental fees for these specific hardware components. As AI demand grows, the volatility of these costs becomes a significant factor for many businesses.

The launch builds on a previous product from Injective that was introduced in August 2025. That earlier version was only linked to the rental fees of a single GPU model, the H100. By expanding to five different models, the new product offers a much broader view of the AI computing landscape.

How the SICI index works

The SICI index tracks the rental prices of five specific Nvidia GPU models: the H100, B200, H200, A100, and RTX 4090. To ensure the data is accurate, the index uses pricing information gathered by LO:TECH, a market data firm based in London. This data is pulled from various computing-resource trading platforms.

The index is not weighted equally across all five models. The H100 is the most significant component, making up 40% of the index, while the B200 accounts for 25%. The remaining weight of the index is distributed among the H200, A100, and RTX 4090 models. This weighting reflects the relative importance of these chips in the current market.

To keep the price stable and avoid sudden spikes from single-day anomalies, the index updates every hour. It calculates the average rental fee over the previous seven days. Traders use USDC as the collateral for these trades, providing a stable way to back their positions in the futures market.

Hedging for AI developers

This new financial product is designed to help different types of companies manage their exposure to fluctuating costs. For AI developers, the primary concern is the rising cost of renting the hardware needed to train and run their models. By using these futures, developers can hedge against sudden price hikes in GPU rentals.

On the other side of the market, GPU suppliers can use the product to manage their own risks. These suppliers might worry about a drop in rental income if the market becomes saturated or if prices fall too quickly. The futures market allows them to protect their revenue streams against these downward swings.

The ability to manage these risks is becoming more important as the industry scales. As more companies invest in AI infrastructure, the financial impact of hardware costs becomes a central concern for corporate planning. This tool provides a way to turn those unpredictable costs into manageable financial variables.

Market context and outlook

The launch comes at a time when there are growing concerns about the profitability of AI infrastructure. Gartner recently warned that more than 80% of neocloud providers that specialize in AI cloud services might stop operating independently by 2030. These companies are currently struggling to generate a sufficient return on their massive investments.

Because of these difficulties, financial products that track GPU rental-fee volatility are becoming more common. As the industry faces pressure to prove that AI investments can be profitable, tools that manage the costs of that infrastructure become more relevant. This move by Injective follows a trend of creating more complex financial instruments for the AI economy.

The expansion from a single-model index to a five-model index suggests a move toward a more mature market. By covering a wider range of hardware, the SICI provides a more comprehensive look at the state of AI computing. This could lead to more sophisticated trading strategies for those looking to navigate the high-cost world of AI hardware.