Electricity is the hard currency of AI! Morgan Stanley targets four core investment tracks— who is likely to take the lead?

When the global technology world's polymers outshine NVIDIA, a much more aggressive industry outlook has been reduced to the depths of stagnant water flow.

Morgan Stanley's latest in-depth Research Reports on Powering AI in the US make it clear that AI is more than just a story, AI-driven infrastructure investment has moved from concept verification to the modeling stage, and electricity is becoming the most important resource for determining AI's progress.

Is electricity the bottleneck for AI?

According to Duma, from 2025 to 2028, the new power demand for US AI data centers will reach an astonishing 65GW. However, the reality is that only 21GW of incremental power is available in the US grid in the short term (15GW can be added immediately to +6GW in the build up), which means that the US will face a power gap of nearly 45GW in the next 3-4 years.

According to Mass Research Reports, there are currently four methods: 1 to modify the original Bitcoin Miner2 and the new Henry Hub Natural Gas Generating Plant 4 to use Distributed Fuel Batteries.

Notably, Dama said, even if all four approaches are applied, the U.S. grid is only too strong to remain “roughly balanced,” and risks to the power system remain.

Therefore, Morgan Stanley has put forward a new valuation concept for the supply of electricity in time, and the huge value of power solutions is not foreseen: the founder expects to raise the value of 500 billion dollars in 3-5 years by 2030, and the total value will rise to 1.1 trillion US dollars.

An exchange of words will be the key to solving the problem of AI in the next 3-5 years and will be the key to the decision of the Industry Chain.

Four Industries, Who Can Be the Most Powerful Co-ordinator?

Bitcoin Mining, Nuclear Power Stations, Henry Hub Natural Gas Power Plants, and Distributed Fuel Batteries are the next investment paths, according to the report. How are these Industries? How many company values are noted? An analysis will be provided to our fellow investors.

1. Reuse of Bitcoin mining

Using one Indicator, Masha compares the Bitcoin Mining and High Performance Computing (HPC) data center using one of the Indicators.

Oma discovers the value of Bitcoin investor's digital power assets, but realizes that converting Bitcoin mining to high-performance computing (HPC) data centers is a viable solution to eliminate bottlenecks, but a rise in Bitcoin price may result in lower operational capacity and increase transfers Exchange cost.

Before“Crypto Miners Lead an Agenda Revolution! Who has the potential to be the next Coreweave?、A NEW AGE OF GOLD! Fund Development Becomes the Next Major Battleground for Investors to Become New Gold”There are many articles that point to AI, and interested fellow investors can check it out.

In addition, IREN will undertake a key strategic transformation from a single Cryptocurrency Mining Business to Face-to-Face Artificial Intelligence (AI), High Performance Computing (HPC) and providing data center solutions and services.

The company expects Bitcoin mining revenue to reach US$1 billion for the year, while the annual revenue of AICloud Business is already close to US$0.25 billion. IREN has already replaced ASICs for Bitcoin mining in multiple mining centers with GPUs to support AI cloud business, invest in building a liquid-cooled AI data center Horizon, and another facility called Sweetwater is expected to be deployed in mid-2027.

According to Samsung Research Reports, the Company is More Bullish: $Riot Platforms(RIOT.US)$ 、 $Cipher Mining(CIFR.US)$ 、 $Hut 8(HUT.US)$ 、 $Bitdeer Technologies(BTDR.US)$ und $IREN Ltd(IREN.US)$ 。

As a whole, Bitcoin mining sites typically have a large amount of power infrastructure that can be reused to support data center operations, such as AI and big data processing in energy-dense clusters.

With the development of AI technology, the demand for electricity increases dramatically, and these energy infrastructure facilities primarily used for Bitcoin mining can be transformed into Precious Assets that support AI data centers. This kind of conversion not only increases the economic efficiency of these sites, but also reduces the time and cost of building a new data center.

2. Henry Hub Natural Gas is the bridge of the present, and nuclear power is the cornerstone of the future

As demand for AI data centres increases, Morgan Stanley has announced that the Henry Hub Natural Gas supply chain, including upstream suppliers and offshore pipeline enterprises, will benefit from long-term pricing agreements.

The analysis shows that if the 20-year supply agreement was signed, the Enterprise Value/Tax Reduction (EV/EBITDA) of the Henry Hub Natural Gas Supply Business Division would be above $10, suggesting that the value of the Assets would be subject to a wave revaluation.

Beyond that, Maha recognizes that the super-sized Technology Companies (Microsoft, Amazon, and Aquarian) are achieving three goals simultaneously through the Bug core: adding power to monitor requirements, supporting reduced power, and obtaining stable, controllable base power.

This year,《Technology giants compete with each other! 2025's nuclear stocks are on the horizon, or it will be AI's next battleground”、Is there a new solution to the bottom of the world's most powerful performance? This Industry Becomes a Secret Winner, Growing by 220% in the Last Year”Several articles have suggested that AI data centers will become the next generation of “electric tigers”, the share of power consumption in data centers will increase even more, and nuclear power will be the next battle for AI.

IN CONTRAST, UPSTREAM DESIGN AND MANUFACTURING HAVE A HIGHER THRESHOLD FOR PROFESSIONALISM AND TECHNICAL BARRIERS, SO UPSTREAM VENDORS NEGOTIATE HIGHER PRICES. DOWNSTREAM MAINTENANCE CYCLES CAN LEAD TO LONG-TERM CASH FLOW AND A MORE PROFITABLE CYCLE. THE DOWNSTREAM PROJECT CONSTRUCTION LOOP PROFIT SPACE IS LIMITED BY CONSTRUCTION. Due to factors such as project cycle and engineering risk, the profitability ratio is relatively stable upstream or downstream.

Upstream: Raw materials and processing

The supply of basic raw materials, key equipment and fuels required for upstream Industry Chain and nuclear development, mainly includes mining and condensing.

Specifically, uranium mining includes $Cameco(CCJ.US)$ 、 $Uranium Energy(UEC.US)$ 、 $Energy Fuels(UUUU.US)$ 、 $Denison Mines(DNN.US)$ 、 $NexGen Energy(NXE.US)$ ; nuclear fuel processing $Centrus Energy(LEU.US)$ 。

Midstream: Design, R&D and Construction

Mid-stream includes design, R&D, and construction. In which,

$NuScale Power(SMR.US)$ is the first listed SMR nuclear power manufacturer, the company's core product is SMR power modules;

$Oklo Inc(OKLO.US)$ focused on the development of Small Modular Reactors (SMRs), which were invested by Altmann, the father of ChatGPT;

$NANO Nuclear Energy(NNE.US)$ Specializing in the development of Small Modular Modular Resonators, the main 4-block SMR-related content, including Manufacturing, Fuels, Transportation, etc., in the manufacture of multi-component vertically integrated Industry Chains;

$BWX Technologies(BWXT.US)$ Focusing on nuclear reactor component manufacturing and nuclear technology, the biggest difference from SMR/OKLO is that BWXT is a large equipment supplier and technology service provider primarily providing nuclear reactor components, nuclear fuel, and defense-related nuclear technology to government and commercial sectors. Customers include the U.S. Government (such as the Navy) (Nuclear submarines provide nuclear reactors).

Downstream: Operations, Sales and Waste Disposal

Downstream mainly for nuclear power plant operations and energy supply. Participants include $Constellation Energy(CEG.US)$ 、 $Vistra Energy(VST.US)$ 、 $American Electric Power(AEP.US)$ 、 $Southern(SO.US)$ 、 $Exelon(EXC.US)$ 、 $Duke Energy(DUK.US)$ 、 $Entergy(ETR.US)$ 、 $Public Service Enterprise Group(PEG.US)$ etc.

In addition, electrical and electrical equipment includes $GE Vernova(GEV.US)$ 、 $Eaton(ETN.US)$ 、 $Honeywell(HON.US)$ 、 $Emerson Electric(EMR.US)$ 、 $Graham(GHM.US)$ etc.

Of more concern is that, as analysts have pointed out that the US government is driving up stakes in various industrial companies, the next target to dive may be the performance industry.

According to Compass Point Analysts, the US Energy Department recently formed a new Nuclear Fuel Alliance, a key fuel needed for nuclear power generation in the country.

In addition, the latest news reports that facing the huge electricity demand of AI data centers, the largest renewable energy developer in the United States $ NextEra Energy (NEE.US)$ Recently received approval from the Federal Energy Management Commission to allow Iowa's Duan Arnold nuclear power plant to be reconnected to the grid. It is also the third nuclear power station to enter the restart process after the Palisez Nuclear Power Station in Michigan (expected to resume operation this year) and the third facility to enter the restart process after Three Mile Island Nuclear Power Station in Pennsylvania (2027). It may also be the last archival reactor to meet the conditions for reactivation.

However, investors trying to bet on an “AI+ nuclear power plant restart” also need to know that the path will not be easy. NextEra told regulators that restarting nuclear power plants is a “highly capital-intensive process,” with CPC planning to invest up to $0.1 billion in the project by 2025.

Morgan Stanley also pointed out that SMR may have to wait for the next decade if it is to really collapse on a massive scale.

3. Find new models to solve power shortages: microgrid, mobile power supply

— Microgrid

Against the backdrop of the continued development of data centers, enterprises and investors should pay more attention to the “microgrid”, Dama said. In which, $Bloom Energy(BE.US)$ Considered by Morgan Stanley to be a key beneficiary of the provision of microgrid-based solutions.

Microgrid providers credit the advantages of high power density, survivability and high stability as a significant addition to data center reliability. We allow keyloads and facilities to continue running throughout the Community when network coverage is unavailable.

According to Ma, Bloom Energy's Fuel Battery Microgrid solution Always On has the following advantages: (i) High quality power (ii) equipment can be highly reliable (iii) the size of the unit is comparable to the size of the unit, which can provide a high density (iv) high power density in case of failure of a single unit As well as the (v) output power is highly dynamic, causing Bloom's power requirements to vary.

In addition, according to China Securities, AI has led to rapid growth in global data center power usage. In areas with lower redundancy of power systems such as North America, power supply issues are increasingly pressing, and new power access in regions such as PJM in the United States requires an average of 3-5 years to wait. Traditional grid-dependent power supply models are difficult to meet the needs of new data centers.

Gas Wheel Generator Supply Chain, GEV payment cycle length up to 3-5 years, Solid Oxide Fuel Cell, simply known as SOFC is designed for modular quick couplers, high efficiency, high robustness and low discharge characteristics to be an alternative power source for offshore data centers One way in diameter. The SOFC rapid component is the core advantage, the forward term capital of the pipeline supports the higher core components in the short term, the high cost of replacing the short term core components, considering the addition of ITC and the dips that do not come, other long-term economic considerations may be important.

Bloom Energy is a global SOFC leader, with a market share of around 4 in the Solid Fuel Battery market, and has secured large orders for data center power supply systems from industry leaders such as Oracle, Equinix, AEP, and more.

By 25Q2, the Company expects to pay SOFC orders over 1GW, covering data center public, utility and industrial customers. The company announced a partnership with Oracle in July 2025 to provide power for its AI data center, with a 90-day payment/2024 partnership with Coreweave that will be deployed in Illinois in 2025Q3, signaling that SOFC is quickly moving into the AI market. In the same year, Bloom and AEP signed a 1GW order (100MW in total), the largest purchase ever for Fuel Batteries, the Equinix Collaboration Model with over 100MW, covering 19 data centers, showing SOFC's capacity to recover and modulate submersible power in the top IDC customer group.

Bloom currently produces 1GW and is expected to double to 2GW by 2026 to meet the rapidly growing IDC market. In the future, system costs are expected to continue to decline as the effects of scale become apparent. Emphasizing the advantages of fuel flexibility and modular deployment, the company believes that SOFC is expected to gradually replace traditional diesel and gas turbines in the coming years as the data center's core power choice.

— Mobile Power

To address the stringent regulatory requirements of Additionality in some regions, Morgan Stanley has adopted a bridge-style development strategy.

The strategy recommends that mobile power sources, such as Henry Hub Natural Gas Generators or Fuel Batteries, be extended again to the large data centre model when they are allowed to access the grid properly. In the long run, this type of expansion can be implemented in a horizontal manner, which can provide pass-through protection for the final access to the network.

In areas such as Texas that enforce strict “additionality” policies, data center developers will not be able to get access to the grid if they fail to synchronize the construction of new generation facilities. This regulatory trend is mainly based on two aspects:

Prevent electricity demand from data centers from driving up electricity prices for residents;

Ensure the security and resilience of the running of the entire power grid.

Against this backdrop, companies that can provide mobile power or fast power generation equipment—such as GE Vernova, Caterpillar, are expected to benefit.

summed

Morgan Stanley pointed out that AI investments are moving forward in the near future, the US AI Industry Chain will welcome a key shift: that power supply bottlenecks will be at the heart of reshaping the value structure of the industry.

During this window, energy and computing infrastructure companies with Early Power, Fast Track and High Add-on capabilities are expected to see an increase in valuation first.