AI DATA CENTER STACK 🧵

Compute is shifting from CPUs to accelerators.
A $1T installed-base replacement is underway.

Every $1 spent on GPUs pulls ~$5 into networking, memory, power, and cooling. AI is a systems build, not a chip cycle.

Training grabs headlines.
Inference becomes the real cost center.

Power is the choke point.
Data centers go up in ~18 months.

If AI monetization slips, hyperscalers pause CapEx immediately.
Hardware cycles always digest.

Below are the key data-center players—by layer 👇 COMPUTE & SERVERS

$NVDA
Controls the accelerated compute layer end-to-end. GPUs are just the entry ticket.
CUDA locks software. NVLink locks topology. Full-rack designs force buyers into Nvidia-defined clusters.
The company sells architectures, not chips. Switching costs compound with scale.

$AMD
The pressure valve.
MI accelerators + ROCm give hyperscalers negotiating leverage. Chiplets enable faster custom variants.
Not winning share outright—but becoming mandatory for second-sourcing in mission-critical clusters.

$INTC
Turnaround optionality.
Gaudi targets inference economics, not bleeding-edge training.
IDM model offers supply-chain security when geopolitics matter more than performance.
Foundry losses still cap the upside.

Neo-Clouds in 2026: This is where the bottleneck moved.

Here’s the bull case by name 👇

$APLD — Applied Digital +56%
Applied Digital isn’t selling GPUs. It sells energized certainty. Polaris Forge proves execution: the first 100MW liquid-cooled building fully energized, moving the story from MOUs to certificates of occupancy. The real moat is leasing shells to AI landlords like CoreWeave, avoiding GPU obsolescence while locking in $11B of contracted lease revenue over 15 years. Tenant fit-out revenue confirms hardware is already being installed. As additional 150MW blocks come online in 2026–27, capex flips to cash flow and valuation logic changes fast. $IREN — Iren Limited +33%
IREN owns the full stack: power, land, cooling, GPUs. No middleman margin. The 750MW Childress campus was designed from day one for liquid-cooled Blackwell racks, not retrofitted later. The inflection is strategic, not cyclical. A $9.7B Microsoft contract with prepayments funds expansion and anchors five years of cash flow. AI Cloud revenue is on track to overtake Bitcoin mining, forcing a multiple reset from crypto proxy to sovereign AI infrastructure.

RBC Capital Top 7 picks for 2026 👇

1) $MSFT — Microsoft
RBC calls Microsoft the best-positioned software giant for generative AI monetization. Azure momentum validated the thesis as commercial RPO backlog surged to $392B, up 51%, signaling multi-year AI infrastructure demand. Copilot adoption exceeded internal expectations across the stack, supported by Microsoft’s unmatched distribution and deep enterprise data integration. Scale turns AI into revenue faster here than anywhere else. 2) $META — Meta Platforms
RBC maintains a Buy with an $810 target, pointing to Meta’s ability to monetize AI through recommendation quality and ad conversion gains. Q3 revenue reached $51.24B, while Llama models enhance engagement across Facebook and Instagram. Heavy Reality Labs investment continues, but pricing power in ads and AI-driven efficiency offset rising capex. User scale remains the advantage.

Top 15 Stocks with Strongest Network Effects 🧵

Network effects one of the most durable moat in markets.
Morningstar calls them one of the strongest economic moats. 👇

1) $META — Meta Platforms
Meta operates the largest direct network effect ever built, anchored by 3.43 billion daily active users across Facebook, Instagram, WhatsApp, and Messenger. Social graphs, messaging history, groups, and content distribution create extreme switching costs that new platforms cannot replicate. $META Advertising performance compounds as more interactions improve targeting precision, reinforcing advertiser demand. With reach extending to 73.7% of global internet users, Meta’s network scale converts engagement into durable pricing power, supporting valuation despite elevated AI and infrastructure investment.

AI isn’t one market. It’s a stacked system. And capital flows to bottlenecks. 🧵

What’s working now isn’t model hype.
It’s control over fabrication, packaging, memory, interconnect, and edge deployment.
Below is the AI stack—layer by layer—and the companies taxing it.

Key players: $ASML $AMAT $KLAC $SNPS $TSM $INTC $NVDA $AMD $AVGO $ARM $MU $MRVL $ALAB 👇 $ASML
ASML is the gatekeeper. EUV is non-negotiable below 5nm. High-NA EUV enables 2nm logic nodes powering next-gen Nvidia and AMD accelerators. €3.6B in EUV bookings confirms demand is structural, not cyclical. No substitutes exist.

$AMAT
Applied Materials monetizes atomic-scale complexity. GAA transistors, HBM, and advanced wiring require precision deposition. Xtera epitaxy tools are built specifically for 2nm wiring. Advanced packaging systems now grow faster than the core business.

$LRCX
Lam dominates etch for 3D NAND and HBM stacks. Cryogenic etching enables extreme aspect ratios. Advanced packaging and GAA shipments exceeded $1B and are projected to triple by 2026. ALTUS Halo integrates molybdenum to reduce resistance at shrinking nodes.

$KLAC
KLA owns yield. One defect can destroy a $30k wafer. As complexity shifts to packaging, KLA’s inspection tools follow. Near-monopoly pricing power persists regardless of capex cycles.

$SNPS
Synopsys is silicon’s operating system. A $2B Nvidia partnership embeds CUDA directly into EDA workflows. Synopsys ai automates verification for massive AI designs, collapsing design timelines.

$CDNS
Cadence anchors system-level design. Certified AI flows for TSMC N2 and A16 lock relevance at the frontier. IP blocks for PCIe and DDR monetize reuse rather than reinvention.

Top Wall Street analysts are most bullish on these 5 S&P500 stocks. 🧵
I’m aligned with them on these names.
Here’s the bull case where fundamentals, positioning, and setup still look asymmetric. 👇

1/ $DDOG | Datadog
Analysts PT $212.3 | +50% upside

Non-AI usage just hit its fastest growth in 12 quarters, signaling core observability demand never broke.
Remaining performance obligations expanded 53% YoY to $2.79B, pointing to long-dated enterprise commitments.
Billings surged 40.8%, FCF margin climbed to 24%, and retention stabilized at 120%.
Expansion is quietly resetting higher. 2/ $TTD | The Trade Desk
Analysts PT $62.3 | +67% upside

Kokai is changing unit economics.
Adjusted EBITDA margin reached 43%, and EPS beat consensus by 125%.
Connected TV continues absorbing budget as streaming shifts to programmatic pipes.
Data clean rooms and identity infrastructure are becoming required plumbing, not optional tools.

$NVDA is currently the most successful chipmaker.

Nvidia is up +50.3% from 52w low, while S&P500 +20.4%.

Is it overvalued? Should you invest now? Let’s examine the valuation, revenue growth rate by segments, profitability, and expenses. A short 🧵👇 @samsolid57 @TheRayMyers @ftr_investors @micscofie1d @Stock_Opine @HedgeVision @FromValue Looking at $NVDA Nvidia's valuation through the Forward EV/Sales multiple, it stands at 15.9, above median of 12.2, and significantly lower than the January 2025 peak, when the multiple reached 20, as well as the 2024 high, when it peaked at 25.

Q4 2024 has concluded the earnings season, revealing how the most popular SaaS stocks have fared. These stocks include $CRWD, $PLTR, $NET, $ZS, $S, $SNOW, $MDB, and $NOW.

Let's compare their key metrics: DBNRR, CAC Payback Period, Margins, valuations, and the level of SBC. 🧵👇 Let's examine how SaaS companies performed relative to their initial revenue forecasts for Q4 2024.

Leading the pack by a significant margin, $PLTR delivered a 7.3% revenue beat, followed closely by $MDB, which also exceeded expectations with an impressive 5.7% beat.

Other strong outperformers included $HUBS, $TEAM, $SNOW, $DDOG, $VEEV, and $IOT, with revenue beats ranging from 4.3% to 3.1%, respectively.

Slightly smaller, but still notable, outperformance came from $GTLB, $MNDY, $CRWD, $ZS, $CFLT, $NET, and $S, which exceeded their forecasts by 2.6% to 1.6%.

The smallest beats were reported by $AXON at 0.9% and $PANW at 0.3%.
Meanwhile, $NOW came in below its own guidance, missing by -0.4%.

$NFLX is one of the key players in Streaming platforms.

Netflix is up almost 83.1% in 2024, while S&P500 up 23.3%.

Should you invest now? Let’s examine the valuation, revenue growth rate by regions, key metrics, profitability, and expenses. A short 🧵👇 $NFLX Netflix is trading at a Forward EV/Sales multiple of 8.7, above the median of 6.2. The company's Forward EV/Sales multiple is near its historical highs.

$SNOW is 8.3% positions in my portfolio.

Snowflake is down almost -22.4% in 2024, while S&P500 up 23.3%.

Should you invest now? Let’s examine the valuation, revenue growth rate by segments, key metrics, profitability, and expenses. A short 🧵👇 After a significant drop, $SNOW Snowflake is trading at a Forward EV/Sales multiple of 12.2—below the levels of 2022-2023 and significantly below the median level of 18.1.

$SNOW Q3'24 Results:
↗️$942M rev (+28.3% YoY, +28.9% LQ) beat est by 5.0%
↘️GM* (72.9%, -2.0 PPs YoY)🟡
↘️Operating Margin* (6.3%, -3.5 PPs YoY)🟡
↘️FCF Margin (8.3%, -5.6 PPs YoY)🟡
↗️EPS* $0.25 beat est by 66.7%
*non-GAAP

Product
↗️Product Revenue $900M (+28.9% YoY)
↘️GM* (76.3%, -2.0 PPs YoY)

Key Metrics
➡️NDR 127% (127% LQ)
↗️RPO $5.70B (+54.1% YoY)🟢
↗️Billings $1,068M (+30.9% YoY)🟢
↗️Data sharing 36.0% (+2.0 PPs QoQ)
➡️2,946 Marketplace Listings (+26.3% YoY, +143)

Customers
➡️10,618 customers (+19.2% YoY, +369)
➡️542 $1M+ customers (+25.2% YoY, +32)

Operating expenses
↘️S&M*/Revenue 36.2% (-0.5 PPs YoY)
↗️R&D*/Revenue 23.9% (+2.2 PPs YoY)
↘️G&A*/Revenue 6.7% (-0.1 PPs YoY)

Quarterly Performance Highlights
↗️Net New ARR $284M (+21.8% YoY)
↗️CAC* Payback Period 17.9 Months (+0.7 YoY)🟡
↘️R&D* Index (RDI) 1.36 (-0.51 YoY)🟡

Dilution
↘️SBC/rev 40%, -2.5 PPs QoQ
↘️Basic shares up 0.7% YoY, -1.3 PPs QoQ🟢
↗️Diluted shares up 0.1% YoY, +1.2 PPs QoQ🟢

Headcount
➡️7,823 Total Headcount (+15.3% YoY, +193 added)

Guidance
↗️Q4'24 $906.0 - $911.0M guide (+23.1% YoY) beat est by 2.1%
↗️$3,430.0M FY guide (+28.6% YoY) raised by 2.2% beat est by 1.7%

👉 Subscribe to my newsletter for the full earnings review. $SNOW Snowflake Customers

$ZS is among the top ten positions in my portfolio.

Zscaler is down almost -25% YTD and a staggering -34.5% from 52-week high.

Should you invest now? Let’s examine the valuation, revenue growth rate by segments, key metrics, profitability, and expenses. A short 🧵👇 $ZS Zscaler is trading at a Forward EV/Sales multiple of 9.5, significantly below the average of 19.9.
The company's Forward EV/Sales multiple is near its historical lows, comparable to the lows of 2020 and January 2023.

When comparing cybersecurity companies with similar revenue growth, they trade at much higher valuations: $CRWD Forward EV/Sales is 15, and $NET Forward EV/Sales is 14.

Even $PANW, with significantly lower revenue growth of 12% YoY for Q2 2024 and an expected slowdown to approximately 11-12% YoY, trades at a Forward EV/Sales of 11.6, higher than Zscaler.

This suggests that Zscaler might be undervalued based on the Forward EV/Sales multiple.

$DDOG is among the top five large positions in my portfolio.

Datadog is down almost -5.2% YTD and a staggering -40% from its 2021 peak.

Should you invest now? Let’s examine the valuation, revenue growth rate by segments, key metrics, profitability, and expenses. A short 🧵👇 $DDOG Datadog is trading at a Forward EV/Sales multiple of 12.6, significantly below the average of 22.3 and lower than the valuations based on multiples before 2020.

Q2 2024 has concluded the earnings season, revealing how the most popular SaaS stocks have fared. These stocks include $CRWD, $PLTR, $NET, $ZS, $S, $SNOW, $MDB, and $NOW.

Let's compare their key metrics: DBNRR, CAC Payback Period, Margins, valuations, and the level of SBC. 🧵👇 Starting with companies beating their own first-quarter revenue forecasts, $ZS leads with a 4.6% beat, which is significant for Zscaler.

$SNOW beat its forecast by 2.4%, which seems minimal given management's statements about a conservative forecast for the next quarter.

Other notable beats include $PLTR, $NOW, $IOT, $DDOG, $GTLB, and $MDB with beats ranging from 3.8% to 3.0%.

The smallest beats came from $VEEV at 1.1%, $S at 1.0%, and $PANW at 0.9%.

$CRWD deserves separate attention as it faced challenges from the July 19 software update incident, which negatively impacted revenue, but still managed to beat its forecast by 0.3%.

$NFLX reported its Q2 earnings.

Netflix beat revenue estimates by 0.3% and EPS estimates by 2.7%, but missed its guidance by -0.7%.

Should you invest now? A short 🧵👇 The revenue growth of $NFLX Netflix has accelerated over the last four quarters to 16.8%.

Revenue growth on an F/X neutral basis is 22%, due to high inflation in Argentina.

Considering the forecast for the next quarter, if the company exceeds its own forecast by 0.7% as in Q2, the growth would be approximately 14.7%, indicating a slowdown in revenue growth.