Bit Brainiac

Ex buy/sell-side analyst | Threading overlooked compounders | Not your daddy's finance guy

[Bit Brainiac @BigBitBrainiac]

@TLEROY5 Idk you tell me :)

[Bit Brainiac @BigBitBrainiac]

$SIVE Earnings Takeaways: Why The Story May Be Bigger Than The Headline Numbers

Been holding Sivers Semiconductors for a little over a month now and still a big fan after going through the latest updates. A few things stand out that I think the market may be underappreciating:

• The adjusted financials are tied to PCAOB compliance - a necessary step for a potential future Nasdaq listing. This is a structural upgrade, not a red flag.

• A meaningful portion of revenue wasn’t “lost,” it was deferred under stricter accounting rules and is expected to be recognized in 2026+.

• The sharp increase in Finland-related revenue is extremely interesting. In this sector, it strongly suggests growing exposure to Nokia ($NOK), which would be a major validation of Sivers’ RF technology.

• CEO commentary continues to point toward larger-scale customer ramps beginning into late 2026, which fits the broader timeline management has been communicating.
This still feels like a company laying infrastructure for a much bigger scaling phase rather than optimizing for short-term optics.

[Bit Brainiac @BigBitBrainiac]

@aleabitoreddit Can’t help you for a 10x but for a 20x $BOT is imho the best option

[Bit Brainiac @BigBitBrainiac]

$OKLO One of the more interesting AI infrastructure themes isn’t compute itself: it’s power generation.

As hyperscalers continue scaling AI workloads, electricity demand is becoming a strategic bottleneck. That’s why small modular reactors (SMRs) are getting increasing attention.

Why Oklo keeps showing up in discussions:

• Focused on advanced nuclear reactor technology

• Positioned around long-term clean baseload power demand

• Strong narrative exposure to AI-driven energy consumption

• Backed by growing political and industrial support for nuclear innovation

The market still seems early in understanding how important reliable energy generation could become in the AI era

[Bit Brainiac @BigBitBrainiac]

$LEU

Been buying nuclear stocks since mid-March and I still think the market is massively underestimating what’s coming in H2 2026 and 2027.

The AI power race is real. Data centers are going to need reliable baseload energy at a scale renewables alone can’t provide. Nuclear is becoming strategic again, not optional.

Why Centrus?

• Only U.S.-owned, publicly traded uranium enrichment company

• Critical player for HALEU fuel needed for next-gen reactors & SMRs

• Direct leverage to U.S. energy security and reshoring of the nuclear fuel cycle

• Tiny float + strategic importance = explosive upside if sentiment shifts

People are still treating nuclear like a trade. I think it becomes a national priority. $LEU is one of the purest ways to play it.

[Office of Nuclear Energy | US Department of Energy @GovNuclear]

Don’t blink. The nuclear energy renaissance is underway.

[Bit Brainiac @BigBitBrainiac]

Glass substrate stocks going limit-up across Asia on news of Intel x Apple cooperation talks

The angle: Intel’s leadership in advanced packaging (EMIB) + push into next-gen substrates is pulling Apple closer into its ecosystem.

This is bigger than it looks - packaging is becoming the new frontier of semis.

As Moore’s Law slows, performance gains shift to:
→ advanced packaging
→ heterogeneous integration
→ substrate innovation

If Apple leans into this stack, it’s a major validation signal.

Names to watch:
$INTC $AAPL $TSM $AMKR $KLAC $LRCX

Quietly, the battleground is shifting beyond the node.

[Bit Brainiac @BigBitBrainiac]

$YSS is increasingly positioning itself as a scaled, low-cost satellite platform - but the real question isn’t launch, it’s throughput.

As constellations expand, the bottleneck shifts toward data transmission, latency, and ground connectivity.

That’s where second-order names like Sivers Semiconductors ($SIVE) become critical - enabling the mmWave and RF layer that actually moves the data.

Market is still focused on $YSS execution (capacity, contracts, ramp).
Less attention on whether the underlying comms stack can scale alongside it.

If $YSS delivers on deployment, the upside likely propagates across the stack - and $SIVE sits right in that path.

[Bit Brainiac @BigBitBrainiac]

The most under-discussed infrastructure bottleneck is not power.

It’s water.

AWWA estimates U.S. drinking-water infrastructure needs of $2.1tn to $2.4tn through 2050, with current annual capex of ~$33.6bn versus required annual investment of ~$90.2bn. That implies a ~$56.6bn annual funding gap.

This is not just an “old pipes” problem. The cost stack now includes aging assets, regulatory compliance, climate resilience, cybersecurity, more complex treatment needs and rising O&M costs.
The equity angle is not to buy every water stock.

The best setups are companies with:
- pricing power
- recurring replacement demand
- exposure to water safety / quality
- low project concentration
- strong margins and FCF

My shortlist:

$WTS: best stable compounder. Water safety, flow control, plumbing, institutional buildings and data centers. FY25 sales +8%, adj. EPS +19%, FCF $356m.

$XYL: best scaled pure-play water platform. Pumps, treatment, smart metering, analytics and services. Q1 backlog $4.7bn, smart metering orders +15%, largest service order ever at $850m.

$ZWS: best underappreciated retrofit / water-quality play. Drinking water, filtration, safety, flow systems. Q1 core sales +11%, adj. EBITDA +18%, margin 26.8%.

$WMS: best stormwater / resilience angle. Drainage, stormwater capture, onsite wastewater. Q3 adj. EBITDA margin 30.2%, FY26 EBITDA guide midpoint ~$945m.

The market is focused on AI power.

But the next decade of infrastructure spend will also require water quality, stormwater resilience, leak reduction, filtration and distribution upgrades.

Boring sector. Long runway. Real capex need.

[Bit Brainiac @BigBitBrainiac]

$IREN: Sweetwater 1 energized (1.4GW)
→ Connected to ERCOT
→ Part of 2GW campus
→ Power ramp coming online
AI infra isn’t about chips first… it’s about power.

[Bit Brainiac @BigBitBrainiac]

🇪🇸 Indra momentum continues:

• $50M U.S. expansion (new Kansas radar factory, 200+ jobs)

• Secured amphibious vehicle contract for Spanish Marines
• Expanding defence footprint internationally (Thailand, naval deals)

Defence + ATC = strong structural tailwinds 🚀

[Bit Brainiac @BigBitBrainiac]

One of the biggest “edges” in trading?

Having someone in your life who keeps your head clear when the markets try to take it.

What’s your edge?

[Bit Brainiac @BigBitBrainiac]

$IDR Indra & Hanwha’s move in Chile could be just the beginning. 🇨🇱 What starts as a single contract may evolve into a strategic foothold, positioning Indra as a key defense & tech provider across South America. Early signal of a much bigger cooperation ahead.

[Bit Brainiac @BigBitBrainiac]

Wednesday is going to tell us everything we need to know about the real economy.

7:15am. ADP jobs report. Last month: 63K private jobs added. Forecast: 41K.

That's not a rounding error. That's a 35% drop in expected hiring, and it comes the day before a jobs report Wall Street is already bracing for.

10:00am. ISM Manufacturing PMI. Two months of expansion after a year of contraction. The question is whether $120 oil just killed it.

The detail nobody's talking about: the jobs report itself lands Friday April 3rd. Markets are closed for Good Friday.

That means whatever drops, whether it's bad (+57K expected vs -92K last month) or catastrophic, nobody can react until Monday April 6th.

A 72-hour window where professional traders know the number and retail doesn't have an exit.

Watch Wednesday. That's your warning shot.

[Bit Brainiac @BigBitBrainiac]

Nike just beat earnings and the stock fell.
Everyone's calling it a disaster.
Look closer.
24 consecutive years of dividend increases. Unbroken.$8.1 billion in cash on the balance sheet.
Wholesale, the channel they almost killed, up 5%.
The Sabrina line is the second most-worn sneaker in the entire NBA.
The stock is at a 9-year low because the market has given up.
That's usually when the interesting money starts paying attention. $NKE

[Bit Brainiac @BigBitBrainiac]

$MU - Micron down 30% in 8 trading sessions.
Was up 60% mid-March on memory shortage thesis. Now barely positive for the year.

Everyone's panicking over Google's new AI model. "Chips are dead!" "Memory demand collapsing!"

Here's what they're missing:

The sell-off started AFTER a blowout earnings report. The company delivered, the stock got destroyed anyway.

Memory shortage didn't disappear. Google's breakthrough doesn't eliminate the need for HBM (high-bandwidth memory) - it changes the architecture, not the demand.

In the last 8 sessions: down 30%. Sandisk and Western Digital also down 9%+.

This is maximum fear on a sector that's structurally undersupplied.

AI training still needs memory. Inference still needs memory. Data centers still needs memory.

What changed? Sentiment. Not fundamentals.

$MU just posted record earnings and guided strong. The stock sold off because traders don't understand that Google's model efficiency INCREASES total compute demand long-term, not decreases it.

I'm not buying yet. But everyone who wanted out just got out. That's usually when things get interesting.

[Bit Brainiac @BigBitBrainiac]

Agentic AI needs payment rails, not speculation tokens.

The question isn't "will crypto be used" - it's "who captures the value when AI agents transact billions of times per day?"

Right now we have:

- Bitcoin: too slow, too expensive per transaction
- Ethereum: better, but still $2-5 per transaction in gas fees
- Stablecoins: solving the volatility problem but still riding expensive L1s

The real opportunity: whoever builds instant, sub-cent settlement rails that AI agents can use programmatically without human intervention.

This isn't about pumping coins. It's infrastructure.
Solana comes closest (fast, cheap), but even at 400ms block times and $0.00025 per transaction, you're still dealing with occasional congestion when AI usage scales.

The winner will be whoever solves:
- Instant finality (no waiting for confirmations)
- Programmable money (AI can custody and transact autonomously)
- Sub-millisecond latency at billions of transactions/day
Regulatory compliance built-in

Traditional banking rails can't do this. Crypto rails aren't there yet either.

First to solve it captures trillions in AI-to-AI commerce.

Everyone else is just exit liquidity.