How to Trade Bitget Compute Perpetuals?

As the AI wave sweeps across the world, high-end GPU compute has become known as the "new oil." Whether for large language model training or inference, rental prices for NVIDIA H100 and B200 GPUs directly affect costs and profitability across the industry. In the past, price discovery for these resources largely took place between institutions and cloud providers, making it difficult for retail traders to participate directly.

Now, Bitget is opening the door. Bitget has officially launched the industry's first pre-market compute perpetuals for retail traders: H100USDT and B200USDT. These futures track the H100 and B200 GPU rental price indices published by Silicon Data. Quoted in USD per GPU hour, the indices reflect standardized average rental costs across multiple cloud service and compute providers. Without purchasing GPUs or renting servers, users can use USDT as margin to go long or short on global AI compute prices. The futures offer up to 10x leverage, 24/7 trading, and funding fee settlements every 8 hours.

This means you can trade the cost of compute itself and participate in price discovery across the global GPU rental and AI infrastructure markets.

Why are compute perpetuals worth watching?

1. Lower barriers to entry: Trade compute prices directly

Traditionally, gaining exposure to GPU rental price movements meant becoming a customer of a cloud provider or major AI company, or using complex over-the-counter agreements for hedging. On the physical side, it also involved high barriers such as hardware procurement, data center hosting, and operations management. Bitget's pre-market compute perpetuals use cash settlement, with all PnL settled in USDT. You do not need to take delivery of GPUs or manage servers in a data center. Opening and closing positions, as well as settlement, can all be completed in the app, with a trading experience similar to crypto perpetuals. This design makes compute price exposure much more accessible, allowing retail traders to express their views on H100 or B200 rental price movements with relatively small amounts of capital.

2. Exposure to a frontier sector: Direct access to a core AI infrastructure cost

The H100 remains a mainstream accelerator for large language model training and inference. The B200, based on NVIDIA's Blackwell architecture, is emerging as a benchmark for the next generation of AI compute. Bitget's decision to launch these two futures first is no coincidence. They sit at the center of global AI compute supply and demand, with their price movements attracting significant market attention.

The futures indices are published by professional data provider Silicon Data. They are based on actual quotes from multiple cloud service and compute providers, with factors such as specifications, rental terms, and regions standardized to reflect actual market supply and demand as closely as possible.

3. A crypto-native experience: Built for crypto traders

These futures were designed with the trading habits of crypto users in mind. USDT margin, up to 10x leverage, 24/7 trading, and funding fee settlements every 8 hours are all highly compatible with existing USDT-M perpetual futures. You do not need to open a traditional brokerage account or adapt to the trading-hour restrictions of forex or commodity markets. Funds can be flexibly allocated across spot, futures, and compute products.

For users already familiar with crypto derivatives trading on Bitget, the learning curve is low, while the range of tradable markets expands beyond crypto into the emerging AI compute sector.

Start trading compute perpetuals in three steps

Getting started is simple:

1. Go to TradFi: Open the Bitget app and tap TradFi in the bottom navigation bar.

2. Search for H100 or B200: Enter "H100" or "B200" in the drop-down list or search bar to access the corresponding futures trading page.

3. Start trading: Enter the quantity and price, then select a market or limit order. You can open either a long or short position. Confirm your order to place the trade. Transfers from your spot account to your futures account are typically instant, with all transactions settled in USDT.

Note: These futures are currently in the pre-market phase. Price movements are primarily driven by internal matching between market makers and users on the platform and are not yet fully connected to broader external liquidity. Quotes are directly affected by buying and selling activity on the platform, and prices may experience relatively high volatility.

Conclusion

Compute is becoming one of the most important factors of production in the AI era. As institutions begin using futures to hedge rental costs, retail traders now have a way to participate as well. Open the Bitget app and search for H100 or B200. This could be your starting point for trading price movements in AI infrastructure.

Risk warning: Pre-market perpetual futures indices may experience significant volatility, and leverage amplifies both potential returns and risks. Manage your positions according to your own risk tolerance.