Post

More from @HyperTechInvest

Daniel Romero

@HyperTechInvest

Nov 5

The market is in a correction

Growth is in a downturn

Here are 5 stocks you should consider:

1. Galaxy Digital | $GLXY

Galaxy Digital is a full-stack crypto and infrastructure platform built for institutions, covering Markets, Asset Management, Banking, GK8 custody, and a fast-growing AI/HPC data center business.

In crypto, Galaxy operates GalaxyOne, offering trading, financing, custody, staking, and research, with GK8 enabling institutional-grade self-custody within the platform.

Through Ventures, Galaxy invests across core infrastructure and DeFi opportunities, recently co-leading $1.65B into Forward Industries $FORD to build the largest Solana treasury strategy.

At Helios in Texas, Galaxy secured $1.4B in project financing and a multi-year lease with CoreWeave, now covering nearly 800 MW of approved capacity. With additional land and interconnection requests, the site could reach around 3.5 GW of long-term potential, pending further approvals and possible rezoning or replatting.

2.Ondas | $ONDS

Ondas Holdings is building a multi-domain autonomy stack through targeted acquisitions across aerial, ground, sensing, and counter-UAS systems.

•American Robotics (2021): fully automated BVLOS drone platform for continuous industrial data capture, forming the backbone of the U.S. drone unit.

•Airobotics (2023): urban-scale drone-in-a-box systems and operations software, expanding into security and critical-infrastructure missions.

•Iron Drone (2023): autonomous interceptor for counter-UAS, integrated with Airobotics for detect-to-defeat missions.

•Ardenna (2022): rail inspection analytics pairing drones with AI defect detection across rail corridors.

•Apeiro Motion (2025): ground robotics and fiber-optic tether technology, adding land systems and contested-environment communications.

•Insight Intelligent Sensors (2025): AI electro-optical modules for surveillance, early warning, and counter-UAS cueing.

Recent deals in progress further extend the stack, including Sentrycs for cyber-over-RF counter-UAS takeover, and 4M Defense for smart demining and subsurface intelligence.

3.Bitfarms | $BITF

Bitfarms is transitioning from pure Bitcoin mining to a U.S.-focused energy and compute infrastructure company, targeting HPC and AI data centers while maintaining Bitcoin exposure. The pivot is anchored in Pennsylvania and the PJM grid.

The Stronghold acquisition provides 1.1 GW of growth pipeline in Pennsylvania and control of legacy power sites, including the upcoming Panther Creek AI campus, projected at around 350 MW. Financing is secured through a $300M Macquarie project facility, with groundbreaking planned for Q4 2025 and initial energization by year-end 2026.

Execution partners are confirmed. T5 Data Centers is leading development at Panther Creek, leveraging its experience with major U.S. hyperscalers. Bitfarms has also appointed external advisors to market its sites to HPC clients and refine the compute strategy.

The portfolio is being rebalanced toward North America. Bitfarms sold its 200 MW Yguazú site in Paraguay to HIVE, planning to reinvest in its U.S. pipeline and reach ~80% North American power by end of 2025. Management now reports a 1.3 GW total energy pipeline, with over 80% based in the U.S.

Read 6 tweets

Daniel Romero

@HyperTechInvest

Nov 2

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.

3. Fluence Energy Inc

Market cap: $4B

Fluence Energy designs and sells utility-scale battery storage systems built from modular units with integrated controls, supported by cloud-based software that optimizes how those batteries and renewable assets operate in power markets. Its ecosystem combines Gridstack hardware for physical storage, Fluence OS for system management, and Mosaic software for AI-driven market bidding and dispatch.

The company emerged as a Siemens–AES spin-out and has become one of the largest pure-play energy storage providers globally, with multi-year deployments across major utilities and independent power producers. Recent wins include a 160 MW / 640 MWh project with Torch Clean Energy in Arizona and 200 MW of systems for Ukraine’s DTEK to strengthen grid reliability during outages.

Financially, Fluence has scaled into the multi-billion-dollar annual revenue range, though results have fluctuated due to manufacturing cycles and project timing. Management continues to highlight a large backlog and expects growth to stabilize as new U.S. production capacity comes online.

Read 15 tweets

Daniel Romero

@HyperTechInvest

Oct 26

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.

3. AST SpaceMobile | $ASTS

Building a satellite constellation to enable direct to smartphone connectivity without the need for specialized hardware.

The long term TAM is massive, the company aims to provide reliable high speed mobile coverage anywhere in the world, unlocking a global market that traditional networks can’t reach.

Read 9 tweets

Daniel Romero

@HyperTechInvest

Jun 11

$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.

2. The Need for Compute

I believe the AI era represents a $195 trillion opportunity over the next decade. This revolution will be bigger and more important than most expect.

Companies that position themselves well in this hypergrowth cycle will see massive stock returns. $AMD is one of the best-positioned players, with the potential to sit at the core of the AI world.

— They offer the top general-purpose processors (CPUs) and likely the strongest chip-level AI accelerators for inference.

— Soon, they’ll launch rack-scale AI systems.

— They also lead in FPGAs, highly programmable chips used in AI, robotics, aerospace, and more.

— Through APU and NPU innovation, $AMD is reshaping the small-scale AI space.

But it’s not just hardware. $AMD is turning a former weakness, software, into a strength.

— ROCm is improving incredibly fast.

— They acquired Europe’s largest AI lab and reshaped their software team through Nod AI and Mipsology, making $AMD a key player in the open-source AI ecosystem.

Read 26 tweets

Daniel Romero

@HyperTechInvest

Apr 25

$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.

Now imagine what happens when these chips deliver the performance they’re supposed to.

And hyperscalers realize they can get a lot more bang for their buck.

Investors are extremely impatient.

When everything goes up except your stock, losing faith becomes the default reaction.

But the reality is, things take time.

$AMD completed the largest acquisition in semiconductor history in 2022, and they’ve also acquired ZT Systems and Pensando.

You think these companies get integrated overnight?

It’s a hard and slow process.

Read 5 tweets

Daniel Romero

@HyperTechInvest

Apr 23

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

3/

$AMD was late to the AI race.

• NVIDIA had H100 and then H200

• AMD had nothing remotely competitive until the MI300X

• Even when MI300X launched, ROCm wasn’t ready

• $AMD didn't make ROCm open source until last year

• The AI software stack wasn’t usable, or at least not at the level of CUDA

The result? A distant second, or even third place.

Read 22 tweets

Share this page!

Enter URL or ID to Unroll

Daniel Romero

Try unrolling a thread yourself!

More from @HyperTechInvest

Daniel Romero

Daniel Romero

Daniel Romero

Daniel Romero

Daniel Romero

Daniel Romero

Did Thread Reader help you today?

Don't want to be a Premium member but still want to support us?

Send Email!