power is the new compute.⚡️

the more AI scales, the more electricity has to move from the grid to the GPU die.

power transformer lead times pushing 2.5 years. generator step-up transformers averaging 144 weeks.

here are the 25 power semi names on my radar.🧵 why power semis, why now:

- $770B hyperscaler capex through 2027 (Jefferies)
- 12 GW data center supply gap in 2025 alone
- NVIDIA 800V HVDC mandate forces wide-bandgap from 2027
- power content per AI rack up 5-10x vs the H100 generation
- NVIDIA named 14 silicon partners for 800V

silicon physically cannot switch 800V at AI density.

SiC and GaN are mandatory now, not optional.

AI infrastructure has 3 layers:

- Compute
- Data movement
- Power

Everyone understood compute first. Then the market started pricing optics/networking.

The next major bottleneck? Grid infrastructure.

Names + thesis below 🧵 The US is already seeing record electricity demand, with AI/data centers becoming one of the key drivers. The EIA expects US power use to hit new records in 2026 and 2027.

That tells you where the next capex cycle has to go.

Not just chips.
Not just servers.
Power infrastructure.

I ran William O’Neil’s CANSLIM screen across the entire US market.

Out of thousands of stocks, these are the top 50.

The #1 name is up +400% YTD, just hit an all-time high, and EPS is growing 400%+ YoY - $SNDK

Here are all 50 names, grouped by sector (many you’ve likely never heard of).

Follow this thread 🧵 The 7 rules every CAN SLIM stock must pass:

C - current quarterly EPS up 25%+ YoY, accelerating
A - 3 year annual EPS growth above 25%
N - new product, new highs, or new management
S - supply tight (buybacks, insider holds, low float)
L - leader in its industry
I - institutional sponsorship rising QoQ
M - market in confirmed uptrend

Only 2% of stocks pass all 7.

That’s where the real leaders come from.

Semiconductor sub-themes still in Stage 1 bases

10 names building massive Stage 1 bases

as semis pull back in the coming weeks, these are the names worth keeping on your watchlist

thread below 🧵 (a few names you’ve probably never heard of) 1/ $ENTG- supplies specialty materials, filtration, and contamination control solutions critical for semiconductor manufacturing

5 year base with strong accumulation, now pushing into new highs

I ran Peter Lynch's 6-rule stock screen across the entire US market.

Out of thousands of stocks, only 16 passed every filter.

The top name trades at 4x forward earnings with a PEG of 0.04 - you either own it or you've been watching it. $MU

Here are all 16 names - some will surprise you.

Follow this thread 🧵 Lynch's 6 rules for finding multi-baggers:

• Trailing P/E < 25
• Forward P/E < 15
• Debt/Equity < 35%
• EPS Growth > 15%
• PEG Ratio < 2
• Market Cap > $5B

All six. At the same time. That's the screen.

Watched the full 3-hour $NVDA GTC keynote on my flight.

Here are 8 takeaways that matter and the names positioned to benefit:

🧵 1/ The Token Factory Era is Here

Jensen made it crystal clear: data centers are now "token factories." Revenue = tokens per second x price per token. The entire Vera Rubin platform is engineered to maximize tokens per watt. This is the new metric that matters.

$NBIS positioned to win - their 5GW+ deep engineering partnership with NVIDIA to power inference and agentic AI factories makes them the go-to infrastructure play for this shift.

As mentioned, I’ve been tracking a fundamental + technical setup with 2–3x upside.

Time to reveal it: $INOD - Innodata.

Here’s why I think it’s just getting started 🧵👇 Innodata $INOD is a data engineering and AI company that helps enterprises build, train, and validate large language models through data collection, annotation, and validation.

Ticker: $INOD
Market Cap: $1.85 Billion

1/ Fundamentals:
Income Statement

Revenue Growth

2020: $58M
2023: $86.8M
2025E: $247M
2026E: $298M

That’s nearly 5x growth in 6 years → CAGR 28% post-2020.

EPS (Earnings Per Share)

2020: $0.6
2021–23: negative EPS
2025E: $0.8
2026E: $1.0+

Clear transition from losses to sustainable profitability.

Net Income
2022: -$12M
2023: -$0.9M
2025E: +$42M
2026E: +$60M

This is a textbook turnaround: scaling revenues are finally hitting the bottom line.

How to build a 3-digit growth portfolio:

my framework, rules, and process 1/ Framework - Finding the right setups

I start with stocks trading above the 50/200-day moving averages to confirm the uptrend.

I look for names that have been basing 6–12 months and reclaiming the 200MA on both daily & weekly charts.

I also want to see revenue and EPS growing 20%+ year over year.

Before pulling daily charts, I check monthly/weekly to see how the stock behaved in past bull & bear markets.

Here is the freebie for my twitter followers 🙏🏻

$NBIS is the stock I bought in my long term portfolio last week.

I don’t usually do write-ups on stocks unless I have strong conviction — and Nebius checks all the boxes.

Why I’m Bullish on Nebius?

Here’s a clear, visual breakdown of my investment thesis: 1️⃣ Nvidia-Backed AI Powerhouse

• Investment: Raised $700M from Nvidia, Accel, and Orbis.
• Cash Position: $3B in cash (50% of its $6B valuation).
• Why It Matters: Nvidia’s backing gives Nebius priority access to AI hardware, a critical edge in the AI boom.