Futures
Hundreds of contracts settled in USDT or BTC
TradFi
Gold
Trade global traditional assets with USDT in one place
Options
Hot
Trade European-style vanilla options
Unified Account
Maximize your capital efficiency
Demo Trading
Futures Kickoff
Get prepared for your futures trading
Futures Events
Participate in events to win generous rewards
Demo Trading
Use virtual funds to experience risk-free trading
Launch
CandyDrop
Collect candies to earn airdrops
Launchpool
Quick staking, earn potential new tokens
HODLer Airdrop
Hold GT and get massive airdrops for free
Launchpad
Be early to the next big token project
Alpha Points
Trade on-chain assets and enjoy airdrop rewards!
Futures Points
Earn futures points and claim airdrop rewards
Investment
Simple Earn
Earn interests with idle tokens
Auto-Invest
Auto-invest on a regular basis
Dual Investment
Buy low and sell high to take profits from price fluctuations
Soft Staking
Earn rewards with flexible staking
Crypto Loan
0 Fees
Pledge one crypto to borrow another
Lending Center
One-stop lending hub
VIP Wealth Hub
Customized wealth management empowers your assets growth
Private Wealth Management
Customized asset management to grow your digital assets
Quant Fund
Top asset management team helps you profit without hassle
Staking
Stake cryptos to earn in PoS products
Smart Leverage
New
No forced liquidation before maturity, worry-free leveraged gains
GUSD Minting
Use USDT/USDC to mint GUSD for treasury-level yields
Dongwu Securities: AI Data Centers Worsen Power Shortages, Gas Turbines Are the First Choice, AI Energy Storage Has Promising Potential
Soochow Securities released a research report stating that the explosion of AI computing power is widening the gap in U.S. electricity supply, leading to a surge in power demand. Power plant manufacturers are building their own power sources, primarily gas turbines, followed by solar storage. Chinese companies will benefit from the high boom in U.S. power infrastructure construction. Key recommendations include core targets in gas turbines such as Dongfang Electric, and also focus on Hailianxun, Harbin Electric, Shanghai Electric, etc. Additionally, highlight leading solar storage companies Sunlight Power, CATL, EVE Energy, etc.
Full text below
【Industry * Zeng Duohong】In-depth: AI Data Centers Intensify Power Shortages, Various Power Sources Demand Soars — U.S. Power Research Series Two
Under the explosion of AI computing power, the U.S. electricity supply gap has significantly widened, and power demand has surged. We forecast that by 2030, the U.S. will have a cumulative AI computing power of 153 GW, with peak electricity demand reaching 963 GW. Considering a peak load rate of 55%-60%, the power generation capacity needed in 2030 will be 1,751 GW, requiring an average annual addition of 100 GW in power capacity over the next five years. From 2026 to 2030, the U.S. has only approved a total of less than 200 GW of new capacity, averaging just 50 GW per year, only meeting 50% of the demand. Coupled with poor interconnection among the three major U.S. grids, long blackout durations, and ongoing decommissioning of thermal power plants, the rigid power supply gap has become a core industry contradiction.
The preferred power supply method for AI data centers is gas turbines, followed by solar storage, with self-built power sources by CSP manufacturers becoming a trend. AI data centers demand highly stable power supply. Gas turbines, with their stable power output and low cost per kWh, are the top choice for CSP manufacturers. Heavy-duty turbines are suitable for large-capacity scenarios, and modified aircraft turbines, which deliver quickly, are more aligned with project delivery deadlines. IEA estimates that from 2024 to 2035, an additional 80 GW of gas-fired power will be invested in U.S. data centers. Solar storage, due to its green energy attributes and economic benefits, is an important supplement, with mainstream manufacturers adopting a “gas turbine + solar storage” hybrid power supply scheme. Although SOFCs (Solid Oxide Fuel Cells) are quick to deploy, efficient, and low-carbon, their cost per kWh is relatively high, making them suitable only for small-scale distributed scenarios. Currently, North America’s market is dominated by Bloom Energy, which has secured over 2 GW of confirmed orders.
Chinese supply chains have achieved technological breakthroughs in complete gas turbines and core components, fully benefiting from overseas order spillovers. The global gas turbine industry is experiencing both volume and price increases. In 2024, new orders for turbines over 10 MW are expected to reach 58 GW, and are projected to exceed 90 GW in 2025 (average price per MW at 1.5 million USD, totaling over 130 billion USD). GE, Mitsubishi, and Siemens together hold about 85% of the market share, with capacity scheduled through 2029-2030. Overseas product prices have increased by over 30%. Chinese companies have achieved domestication of F-class heavy-duty turbines, with Dongfang Electric, Shanghai Electric, and Harbin Electric Group as core manufacturers. Companies like Yili Co., Ltd. have made breakthroughs in turbine hot-end components, and Jereh has entered the North American aircraft modification turbine market as an authorized complete set supplier for Siemens. The domestication rate of core components continues to rise, and under the high industry boom globally, there are significant opportunities for order spillovers.
Investment advice: The explosion of AI computing power, the widening U.S. power supply gap, and the surge in power demand have led CSP manufacturers to build their own power sources, primarily gas turbines and secondarily solar storage. Chinese companies will benefit from the high boom in U.S. power infrastructure construction. Key recommendations include core targets in gas turbines such as Dongfang Electric, and also focus on Hailianxun, Harbin Electric, Shanghai Electric, etc. Additionally, highlight leading solar storage companies Sunlight Power, CATL, EVE Energy, etc.
Risk tips: Intensified competition, policy changes exceeding expectations, slower-than-expected promotion of low-voltage DC energy storage solutions in AI data centers, and raw material supply shortages.
(Source: Cailian Press)