Energy demand is skyrocketing

Superintelligence will require the equivalent of thousands of nuclear plants to power it

That’s why energy stocks are surging

Here are 14 energy companies benefiting from the AI data center boom that you should consider

A thread 🧵

1. Bloom Energy Corporation

Market cap: $31B

Bloom Energy develops solid-oxide fuel cells that generate electricity on-site using natural gas, biogas, or hydrogen. The company also builds high-temperature electrolyzers that produce hydrogen more efficiently than conventional systems, making it a strong candidate to supply clean, reliable power for the AI infrastructure boom.

Recent agreements reinforce its momentum: a $5B partnership with Brookfield to supply power for AI data centers, over 100 MW with Equinix across U.S. sites, and a 500 MW commitment from SK ecoplant linked to hydrogen projects. CoreWeave is also deploying Bloom units at its Illinois AI facility, showing growing adoption across major hyperscale and industrial clients. 2. Hyliion Holdings Corp

Market cap: $420M

Hyliion develops the KARNO linear generator, a sealed unit that uses high-temperature oxidation to produce electricity and can run on a wide range of fuels including natural gas, hydrogen and raw well-head gas. The company acquired the technology from GE Additive in 2022 and a customer unit has logged 100 days of continuous operation.

Compared with Bloom Energy, which relies on a mature solid-oxide fuel-cell platform scaled across data centres, Hyliion emphasises fuel flexibility and use of field-gas. Bloom is actively signing large AI-power deals and has a long record with colocation operators.

Hyliion secured a $6 M grant from the U.S. Department of Energy to deploy up to 2 MW of KARNO units on Permian well-head gas and achieved the 100-day runtime milestone. Bloom signed an “up to $5 B” partnership with Brookfield Asset Management for AI data centres and surpassed 100 MW with Equinix, Inc. across U.S. sites.

X is full of amazing stock ideas

Here’s a thread with the summarized bull case for 8 X darlings

Let’s start

1. Kraken Robotics | $PNG

Battery and sonar systems producer for subsea defense and infrastructure.

Already profitable with 30%+ CAGR long-term growth guided. A close supplier to Anduril, rapidly expanding its manufacturing footprint with operations across Canada, the US, and Europe.

Trading at just a $1B enterprise value, despite strong tailwinds from defense modernization and maritime security. 2. Eos Energy | $EOSE

Utility-scale battery manufacturer using zinc instead of lithium, making its systems safer and capable of longer discharge durations, ideal for data center backups and grid support.

The company is vertically integrated with a US-based supply chain, benefits from strong federal subsidies, and is rapidly expanding its manufacturing capacity to meet rising energy storage demand.

$AMD has 10x potential from here

They have 10 years of hypergrowth ahead

GPUs, CPUs, FPGAs, APUs, NPUs, DPUs, model deployment, photonics, gaming, VR...

This company will be at the core of the AI era

Let’s dive deep into it 🧵 1. Introduction

“10-bagger” is a term often used carelessly. Any stock with momentum gets hyped as the next one, but $AMD is different. Despite strong results and cheap fundamentals, especially given its expanding TAM, the stock remains 47% below its all-time high.

$NVDA is stealing the spotlight, and the market is missing how $AMD is quietly taking over the CPU market once led by $INTC.

While it may seem like the underdog in AI, $AMD is clearly laying the foundation to become a major player in both hardware and software, in a world where compute power is the most valuable resource.

The way a $180 billion company becomes a 10-bagger is by targeting a multi-trillion dollar market. $AMD is in the middle of a transformation that could take it to a $2 trillion valuation and beyond.

$MSFT is collaborating with $AMD to bring best-in-class DeepSeek inference performance.

Their benchmarks showed that $AMD's MI300X can vastly outperform $NVDA's H200.

_________________________________________

Why did they collaborate to improve $AMD's hardware performance?

They don’t like $NVDA's high prices and low supply.

That’s exactly what I’ve been saying.

Squeeze your customers too much, and they’ll pour money into the second-best option, even if it’s worse.

And I have no problem admitting $AMD has been worse, primarily because of their software.
But now that $AMD is investing hard in that department.

And that other companies and developers are contributing to improving the performance of these chips.

The result is that catching up with $NVDA isn’t all that impossible.

And investors need to remember, this is just the beginning.

- ROCm is improving extremely fast.

- Silo AI and Nod.AI were acquired precisely to improve the software, and they’re working on it.

And in a couple of months, $AMD will launch its new rack system featuring the MI355X, a next-gen AI accelerator with a brand new architecture.

$ORCL has already bought 30,000 units.

Investors tend to assume $NVDA will remain the leader forever

The reality is that $AMD is coming for the first place

If you are an investor in these stocks, you can't be taken by surprise

You need to understand what's happening under the surface

Let's explain it

1/

$NVDA dominated AI for years

• 90% share in data center GPUs

• CUDA as the de facto AI software stack

• NVLink and NVSwitch for scaling infrastructure

But 90% market share doesn’t last forever

Just ask Intel. 2/

NVIDIA’s dominance hasn't been good for the industry

• Companies hate overreliance

• Prices remain prohibitively high

• Lock-in has frustrated customers

And when a company squeezes its customers too hard for too long, they eventually start looking for alternatives.

That’s where $AMD came in

$META just presented their opening statement in their trial for monopoly against the Federal Trade Commission.

These are the main arguments:

1.Competitors offer similar products 2. $META doesn’t charge anything for their apps

$NBIS is one of the most exciting companies in the stock market

It has a high chance of becoming a 10x stock within the next 10 years

Or even just 5 years

Let's me explain why: 🧵👇 1.

Nebius Group comprises the core business, Nebius, and three additional primary business units: Toloka, Avride, and TripleTen.

Nebius, the core business, operates a proprietary data center in Finland and utilizes colocation arrangements in Paris, Iceland, and Missouri. Nebius owns only the Finland facility outright; the other locations are colocations, meaning Nebius rents external companies’ data centers.

Let's first explore the technology and business model of Nebius, followed by its financials.

𝐌𝐲 𝐢𝐧𝐢𝐭𝐢𝐚𝐥 𝐭𝐡𝐨𝐮𝐠𝐡𝐭𝐬 𝐨𝐧 $ASTS

🧵

Let’s begin with the costs necessary to start continuous operations.

$ASTS needs 60 Block 2 satellites to provide continuous service coverage, with each satellite launch costing around $20 million. This results in a total launch cost of approximately $1.2 billion for the constellation.

Additionally, capital expenditures and other operational costs must be considered. Capex is expected to increase to about $100 million per quarter. Assuming operational spending slightly above current levels, total cash burn for 2025 is estimated at around $150 million per quarter, plus launch costs.

Assuming capex spending declines in 2026 but is offset by an expected increase in operational expenses as coverage expands, total cash burn over the next two years is projected to be around $2.4 billion. 𝐒𝐨 𝐰𝐡𝐞𝐧 𝐰𝐢𝐥𝐥 𝐭𝐡𝐞 𝐦𝐨𝐧𝐞𝐲 𝐬𝐭𝐚𝐫𝐭 𝐜𝐨𝐦𝐢𝐧𝐠 𝐢𝐧?

One of their main partners, Rakuten, expects to initiate nationwide coverage in 2026.

This suggests that $ASTS won’t generate recurring revenue until 2026.

Currently, they have $1 billion in cash on the balance sheet. While this isn’t enough to complete the constellation, it provides enough liquidity to stay on track with their roadmap.

However, more dilution or debt offerings should be expected this year. Although they may secure some non-recurring revenue from government contracts and initial revenue from basic coverage services, they won’t generate meaningful revenue until the constellation is fully deployed and continuous coverage begins in 2026.

That said, this doesn’t mean one should refrain from buying shares. While I expect further dilution, if the company continues hitting key milestones, the stock price could rise significantly. This would allow them to raise capital at better valuations. It wouldn’t be surprising to see $ASTS reach $50 per share in 2025, followed by a 10% dilution, which could bring the stock down to $40—but that would still be much higher than today’s price.

IS $AEHR THE NEXT 10x AI MICRO-CAP OPPORTUNITY AFTER A 75% FALL?

$AEHR tests chips for Tesla $TSLA and is positioning itself to become a potential 10x stock by expanding into AI chips.

Let’s see 🧵 $AEHR has dropped 75%, which could make it a decent bottom-fishing opportunity, especially given its connections to the fast-growing AI sector. But before jumping in, there are a few things to keep in mind.