Layer-1 Blockchain Injective has launched a perpetual futures market that tracks the leasing prices of five NVIDIA GPU. Traders are allowed to bet with up to 10 times leverage to AI the changes in costs.

This index, named Super Intelligence Compute Index ( SICI ), tracks the rental prices of NVIDIA H100, B200, H200, A100, and RTX 4090 chips. The data is sourced from the computing power market platform compiled by LO : TECH, a London-based market data company.

According to the announcement on Monday, this index assigns the highest weight to H100, which accounts for 40%, followed by B200 at 25%. The index is updated every hour and uses the moving average of rental prices over the past seven days. Traders can use USDC as margin to bet on whether the index will rise or fall.

This launch is based on the market initiative introduced by Injective in August 2025, which linked the rental prices with NVIDIA H100. The product has been expanded from a single GPU model to a five-chip index.

Companies that develop AI models usually rent computing power from cloud service providers such as CoreWeave and Nebius, rather than purchasing NVIDIA GPU directly. As rental prices fluctuate with demand, Injective indicates that this futures market can help developers hedge against higher computing costs, while also allowing GPU providers to protect themselves from declines in rental income.

AI Infrastructure boom raises financial concerns

As companies that have rented AI computing power begin to face doubts about whether they can recoup the costs of infrastructure construction and operation of GPU, this launch comes at just the right time.

In October, research institution Gartner warned that by 2030, over 80% of professional AI cloud service providers, namely the so-called neoclouds, may cease to operate independently due to the difficulty of obtaining sufficient returns from their investments.

Elon Musk also echoed these concerns on Sunday, stating that over 200 neocloud providers may face failure or integration, as lending institutions are becoming reluctant to provide financing to companies that rely on revenue from AI data centers.

Elon Musk holds financial interests in this field through SpaceX; the “Super Intelligence” business of SpaceX recently announced third-quarter revenue of $7.6 billion, a year-on-year increase of 830%, partly due to new cloud service agreements.

At the same time, major technology companies are also reorganizing their businesses to control costs, as they are investing billions of dollars in AI infrastructure.

Bloomberg reported on Monday that Microsoft expects its capital expenditures in 2026 to be around $175 billion and is currently offering voluntary resignation programs to about 60,000 eligible employees in order to cut costs as infrastructure spending increases in AI.

Meta has also invested over $100 billion in the AI sector and cut 8,000 jobs in May as part of a reorganization aimed at reducing costs.