Quiet Giants & Mega-Caps: Top 8 Stocks to Buy in September
Balancing AI infrastructure growth with value to beat the index without chasing hype.
Quick Takeaways
- AppLovin up 2,000% in three years; 70% YTD—quiet compounding beats hype.
- Vertiv Q2 revenue $2.6B, up 35% YoY; backlog $7.9B.
- Arista Q2 revenue $2.2B, up 30% YoY; gross margin >60%.
- ASML high-NA EUV tools process ~60% more wafers/day—sticky service revenue.
- Broadcom AI chip/networking revenue $5.2B, up 63% YoY; backlog $110B.
- Super Micro shipping up to 5,000 AI racks/month; liquid-cooled by default.
- TSMC controls ~70% foundry share; CoWoS capacity a key bottleneck.
- Alphabet at ~23x earnings; FCF yield ~4.5%—durable growth at a discount.
- MongoDB revenue up 24% YoY; Atlas up 29% YoY—watch for margin inflection.
Vertiv (VRT)
Liquid cooling & high-efficiency power systems enabling dense AI racks; PEG ~0.8 offers growth-at-a-reasonable-price.
- Q2 revenue $2.6B (+35% YoY)
- Backlog $7.9B; cooling market set to more than double by 2030
Arista Networks (ANET)
Ethernet’s AI moment; EOS Smart AI turns Ethernet into AI-grade fabric.
- Q2 revenue $2.2B (+30% YoY)
- Gross margin >60%; 800G shipping, path to 1.6T
ASML (ASML)
High-NA EUV is the tollbooth for next-gen chips; services/upgrade flywheel.
- Throughput ~+60% vs prior EUV gen
- Installed base expansion = recurring cash flows
Broadcom (AVGO)
AI networking + custom accelerators + VMware software = diversified AI cash engine.
- AI chip & networking $5.2B (+63% YoY)
- Backlog ~$110B; Jericho 4 & Tomahawk Ultra cycles
Super Micro (SMCI)
Rack-scale integrator with speed moat; Blackwell-class liquid-cooled racks.
- Capacity ~5,000 racks/month (2,000 liquid-cooled)
- Q4 FY25 revenue +25% QoQ
Taiwan Semiconductor (TSMC)
Dominant foundry & packaging (CoWoS) capacity; pricing power intact.
- 3nm ~24% of wafer revenue; 2nm roadmap 2026
- Aug revenue +33.8% YoY; YTD >37%
Alphabet (GOOGL)
Under-valued mega-cap growth with 4.5% FCF yield; diversified AI leverage.
- Trades ~23x earnings (below 5-yr median and peers)
- Search, Cloud, YouTube compounding
MongoDB (MDB)
Atlas simplifies AI app data; watch for sustained margin turn to trigger re-rating.
- Revenue $591M (+24% YoY)
- Atlas +29% YoY; tight hyperscaler integrations
Strategy & Catalysts
Playbook: Pair under-the-radar growth (Vertiv, Arista, Super Micro) with durable toll booths (ASML, TSMC) and cash-rich mega-caps (Alphabet, Broadcom). Keep optionality with a data-layer compounder (MongoDB).
- Macro: Sticky inflation > higher for longer; a later pivot could re-rate both mega-caps and small caps.
- Earnings tape: Data-center suppliers’ prints and orders are the near-term truth serum.
- Tech cycle: Ethernet AI fabrics, high-NA EUV rollout, Blackwell shipments, CoWoS throughput.
- Risk: Supply bottlenecks (packaging, HBM), enterprise spend digestion, geopolitics.
Disclosure & Notes
This report is for informational purposes only and does not constitute investment advice or a solicitation to buy or sell any securities. Figures are drawn from company commentary referenced in your source script and may be rounded.
Here’s a full finance and investment blog based on your script. It’s structured with engaging storytelling, market context, and analysis for global investors and readers. Word count is ~1,050 (within your 1,000–1,200 target).
Quiet Giants and Mega-Caps: My Top 8 Stocks to Buy in September
Key Takeaways
- AppLovin is up 2,000% in three years, showing how quiet performers can crush both small- and large-cap peers.
- Vertiv’s Q2 revenue grew 35% YoY to $2.6 billion, supported by a $7.9 billion backlog.
- Arista Networks reported $2.2 billion in Q2 revenue, up 30% YoY, with gross margins above 60%.
- ASML’s new high-NA EUV tools process 60% more wafers per day, reinforcing its chip monopoly.
- Broadcom’s AI chip and networking revenue hit $5.2 billion, up 63% YoY, with a $110 billion backlog.
- Super Micro ramps to 5,000 AI racks per month, making it the fastest integrator in the Blackwell era.
- TSMC controls 70% of global foundry share and dominates advanced packaging demand.
- Alphabet trades at 23x earnings, cheaper than Microsoft (36x) and Nvidia (40x), with 4.5% free cash flow yield.
- MongoDB Atlas revenue rose 29% YoY, positioning it as a flexible backbone for AI applications.
The Edge That Most Investors Miss
Three years ago, one of my early stock picks was AppLovin (APP). Back then, it was an under-the-radar player in mobile advertising. Fast forward: the stock is up over 70% year-to-date and more than 2,000% in just three years.
This performance quietly outpaced both small- and large-cap peers—without ever becoming a household name. That’s exactly the kind of edge I look for: companies with strong fundamentals, enough “heat” to prove their strength, but not so much hype that the upside is already priced in.
At the same time, I balance these bets with mega-cap giants, which offer durability and steady returns. In today’s market—where inflation lingers and interest rates hover—the sweet spot is finding names that combine growth momentum with reasonable valuations.
Here are the eight stocks I’m buying in September.
1. Vertiv Holdings (VRT) – Powering the Data Center Shift
If you’ve tracked the AI boom, you know data centers are at their physical limits. Air cooling can’t keep up with today’s rack densities.
That’s where Vertiv shines. They manufacture the liquid cooling and power systems—rear door heat exchangers, direct-to-chip loops, prefab cooling modules—that keep hyperscale AI clusters stable.
- Q2 revenue: $2.6 billion, up 35% YoY.
- Backlog: $7.9 billion (near-guaranteed cash flow).
- PEG ratio: 0.8 vs. S&P 500’s ~2.0.
With the cooling market expected to double by 2030 and liquid cooling growing 20%+ annually, Vertiv is positioned as a sleeping giant.
2. Arista Networks (ANET) – Ethernet’s AI Moment
Modern AI clusters need ultra-fast, lossless networking. The industry is standardizing on Ethernet, and Arista is at the center of it.
- Founding member of the Ultra Ethernet Consortium (Spec 1.0 published June 2025).
- Rolling out EOS Smart AI—software for congestion control and AI-grade traffic engineering.
- Q2 revenue: $2.2 billion, up 30% YoY.
- Gross margins: over 60%.
Arista is essentially building the digital roads AI workloads will travel on, making it a core AI infrastructure play.
3. ASML (ASML) – The Gatekeeper of Chips
Every leap in semiconductors runs through ASML’s extreme ultraviolet (EUV) lithography machines. Their newest high-NA EUV tools process 60% more wafers per day than the last generation.
Intel has already deployed two; TSMC is lining up. Each installation locks in service contracts and upgrade revenue, making ASML’s growth both sticky and recurring.
ASML isn’t just a supplier—it’s the toll booth of the global chip industry.
4. Broadcom (AVGO) – From Chips to Software Strength
Broadcom has transformed itself into an AI-first company:
- Q2 AI chip & networking revenue: $5.2 billion, up 63% YoY.
- Backlog: $110 billion, offering multi-year visibility.
- New products like Jericho 4 and Tomahawk Ultra replace legacy gear.
- VMware acquisition broadens recurring software revenue.
Broadcom straddles both hardware and software, making it a rare hybrid with consistent cash flow and AI upside.
5. Super Micro (SMCI) – Speed Is the Moat
In AI infrastructure, delivery speed matters as much as demand.
Super Micro has become the go-to integrator for rack-scale AI clusters, already shipping Blackwell-class racks with liquid cooling.
- Capacity: 5,000 racks/month (2,000 liquid-cooled).
- Annual output: ~$25 billion, up from $20 billion last year.
- Q4 FY25 revenue: up 25% quarter-over-quarter.
When hyperscalers like Microsoft and Amazon need clusters fast, they call Super Micro.
6. Taiwan Semiconductor (TSMC) – The Ultimate Toll Booth
TSMC controls 70% of the global foundry market. If Apple, Nvidia, or AMD wants advanced chips, they go through TSMC.
- 3nm nodes: Already 24% of wafer revenue.
- Roadmap: 2nm in 2026.
- Advanced packaging (CoWoS) demand so high that Nvidia has secured 70% of 2025 capacity.
- August revenue: up 33.8% YoY, with 37% YTD growth.
TSMC is both the foundry and the packaging leader, cementing its role as the heartbeat of AI compute.
7. Alphabet (GOOGL) – Undervalued Mega-Cap
While Microsoft and Nvidia dominate AI headlines, Alphabet looks like the most undervalued big tech growth play.
- Valuation: 23x earnings (vs. MSFT at 36x, NVDA at 40x).
- Free cash flow yield: 4.5%, above the S&P 500 average.
- Cloud, Search, and YouTube growth continue to compound.
Alphabet offers investors a chance to own durable growth at a discount compared to peers.
8. MongoDB (MDB) – Databases for the AI Era
AI apps need flexible, scalable databases. MongoDB’s Atlas platform handles both traditional structured data and messy AI data seamlessly.
- Q2 revenue: $591 million, up 24% YoY.
- Atlas growth: up 29% YoY.
- Strong partnerships with AWS, Azure, and Google Cloud.
MongoDB still runs negative margins, but the moment it hits consecutive quarters of profitability, its stock could re-rate explosively.
Conclusion: Balancing Growth and Value
This September, my focus is on names that combine AI infrastructure growth with reasonable entry points.
- For high growth with value, I lean on Vertiv, Arista, and Super Micro.
- For durable toll booths, ASML and TSMC are untouchable.
- For cash-rich mega caps, Alphabet and Broadcom provide stability.
- For a speculative upside, MongoDB fits the mold.
The market is still digesting inflation and waiting for rate cuts. When rates eventually come down, I expect both mega-caps and small caps to re-rate higher. By blending these plays, I aim to outperform the index—without chasing hype.
✅ Investment Implications:
- Policy tailwind: A Fed pivot in 2025–26 could reignite tech multiples.
- Near-term catalyst: Q3 earnings from data center suppliers will confirm demand trends.
- Strategy: Blend mega-cap reliability with under-the-radar growth stories.
Navigating the AI Data Center Boom: Eight Stocks Poised for Inflation-Resilient Growth in September 2025
In the ever-shifting landscape of global markets, where inflation lingers like a stubborn fog and interest rate cuts whisper promises of relief, investors crave that rare blend of reliability and rocket fuel. Three years ago, a quiet pick like Apploving—up 70% year-to-date and a staggering 2,000% over the period—taught me the power of under-the-radar gems. These aren’t the screaming headlines of Nvidia’s latest chip; they’re the sturdy plumbing and wiring making the AI revolution possible. Mega-caps offer the bedrock of steady returns, but when undervalued, they pair beautifully with small-cap dynamos chasing explosive upside.
As September 2025 unfolds, with the Federal Reserve signaling potential rate reductions amid cooling inflation, I’m doubling down on eight stocks that straddle this sweet spot. These picks—spanning data center infrastructure, semiconductors, and software—aren’t hype chasers. They’re high-growth engines with value baked in, designed to outpace the S&P 500’s sluggish 8-10% annual returns. Drawing from recent earnings and market forecasts, this analysis uncovers why these companies are primed to weather economic headwinds and surge when liquidity flows freer.
The AI arms race is the backdrop: hyperscalers like Amazon, Microsoft, and Google are pouring billions into data centers, projected to consume 8% of global electricity by 2030. But power grids strain under the heat—literally. Rack densities have skyrocketed from 5kW to 50kW+, rendering air cooling obsolete. Enter liquid cooling: a $5.38 billion market in 2024, exploding to $17.77 billion by 2030 at a 21.6% CAGR, per Grand View Research. Overall data center cooling? A $22.13 billion behemoth in 2024, doubling to $56.15 billion by 2030 at 16.4% CAGR. These stocks aren’t just riding the wave; they’re building the surfboards.
Vertiv Holdings: The Unsung Hero of Cooling Efficiency
Picture a data center as a blazing furnace: AI servers guzzle power, generating heat that could melt steel. Vertiv Holdings (VRT) supplies the lifeline—coolant distribution units, rear-door heat exchangers, and prefabricated modules that keep the inferno in check. As operators pivot to liquid cooling for densities exceeding 100kW per rack, Vertiv’s backlog ballooned to $8.5 billion in Q2 2025, up 21% year-over-year.
Financials tell a sleeping giant’s tale. Q2 revenue hit $2.64 billion, surging 35% YoY, with adjusted EPS at $0.95—beating forecasts by 42%. Margins expanded as operating cash flow exploded post-2023, and EPS forecasts have consistently outperformed. Analysts peg 12-15% upside over the next year, fueled by a PEG ratio of 1.2—far below the S&P 500’s forward 2.0. At 59x trailing P/E, it looks pricey, but DCF models suggest 15.7% undervaluation amid AI’s 30% CAGR through 2030.
Vertiv’s edge? It’s not flashy—it’s essential. With book-to-bill at 1.2x and organic orders up 15%, this mid-cap ($35B market cap) crushes small-cap volatility while dodging mega-cap scrutiny.
Arista Networks: Ethernet’s AI Backbone
Data centers aren’t silos; they’re interconnected webs demanding blistering speeds. Arista Networks (ANET) weaves them with high-speed Ethernet switches and EOS software, tuned for AI workloads like traffic engineering and congestion control. As the Ultra Ethernet Consortium’s Spec 1.0 drops in June 2025—with Arista as a founder—Ethernet cements its role in AI backends, paving paths to 1.6T ports.
Q2 revenue roared to $2.205 billion, up 30.4% YoY, smashing estimates by $100 million. Gross margins held firm at 65.2% GAAP (65.6% non-GAAP), with operating cash flow hitting a record $1.2 billion. EPS beat by 12%, earning a 15%+ analyst upside. Full-year guidance jumped to $8.75 billion (25% growth), eyeing $10 billion by 2026.
At a $129B market cap, Arista blends mega-cap stability with 30%+ growth. Its 48.8% operating margin underscores efficiency, positioning it to snag share from legacy players as AI clusters scale.
ASML: The Monopoly on Chip Precision
No advanced chips without ASML’s EUV machines—the tollbooth every fab must pass. Holding 90%+ market share, ASML’s high-NA EUV tools shrink features by 30%, packing 60% more wafers daily than predecessors. Intel’s two units are in production; TSMC lines up for 2028 mass adoption, despite initial cost hesitations.
2025 sales forecast: €30-35 billion, up 7-25% YoY, with high-NA growing 30%. Installed base expansion locks in recurring service revenue, yielding durable free cash flow. Analysts see sticky growth as AI nodes shrink to 2nm by 2026.
ASML’s €350B market cap screams mega-cap fortress, but at 40x forward P/E, it’s undervalued versus Nvidia’s frenzy—pure value in a growth wrapper.
Broadcom: AI’s Custom Chip Architect
Broadcom (AVGO) didn’t just join the AI boom; it remade itself around it. Q3 FY2025 revenue: $16 billion, up 22% YoY, with AI semis at $5.2 billion (63% surge). Backlog? A record $110 billion, half AI-tied. VMware software added double-digit recurring flow, with Jericho 4 and custom accelerators for Amazon/Microsoft boosting margins to 78.4%.
Q4 guidance: $17.4 billion revenue, AI at $6.2 billion (66% YoY). Free cash flow hit $7 billion, up 47%. At 40x forward P/E, Broadcom’s $800B cap offers mega-cap ballast with 50-60% AI growth in FY2026.
Super Micro Computer: Rack-Scale Speed Demons
Super Micro (SMCI) builds the racks AI dreams run on—Blackwell-class with liquid cooling as default. Capacity: 5,000 racks/month, including 2,000 liquid-cooled, pushing annual output toward $25 billion from $20 billion last year.
Q4 FY2025 revenue: $5.8 billion, up 25% QoQ and 8% YoY, though EPS missed at $0.41. FY2026 target: $33 billion minimum. Modular designs and cooling expertise make it the integrator of choice for hyperscalers racing Blackwell clusters.
This $50B mid-cap volatile? Yes. But delivery speed trumps all in AI’s sprint.
TSMC: The Foundry Foundry
Taiwan Semiconductor (TSM) manufactures 70% of global chips—Apple, Nvidia, AMD all funnel through. 3nm wafers: 24% of revenue; 2nm roadmap for 2026. CoWoS packaging? Nvidia snagged 70% of 2025 capacity for Blackwell.
August 2025 revenue: up 33.8% YoY; YTD 37%+. Advanced packaging to exceed 10% of revenue in 2025. As logic shrinks and HBM stacks grow, TSMC’s pricing power endures.
$900B mega-cap at 25x P/E? The ultimate tollbooth.
Alphabet: Undervalued AI Powerhouse
Alphabet (GOOGL) flies under AI radars, but at 23x earnings—below its 25x 5-year median and peers like Microsoft’s 36x or Nvidia’s 40x—it’s a steal. Free cash flow yield: 4.15%, topping S&P’s 3%. Q2 FCF: $24.3 billion.
$2T cap, 26x trailing P/E—durable growth at a discount.
MongoDB: Flexible Data for AI Apps
MongoDB (MDB) stores unstructured JSON data, easing AI’s messy inputs. Atlas cloud: 74% of Q2 revenue, up 29% YoY to $436 million; total revenue $591 million (24% growth). Non-GAAP margins: 15%, with positive FCF at $70 million.
Customers: 59,900, up 18%. FY2026 guidance: $2.35 billion revenue. At $20B cap, it’s the small-cap spark.
Financial Snapshot: Growth Meets Value
| Company | Market Cap ($B) | Q2/FY Q4 2025 Rev ($B) | YoY Growth | Gross Margin | Backlog/Other | Analyst Upside (12-mo) | PEG Ratio |
|---|---|---|---|---|---|---|---|
| Vertiv (VRT) | 35 | 2.64 | 35% | N/A | $8.5B backlog | 12-15% | 1.2 |
| Arista (ANET) | 129 | 2.205 | 30% | 65% | N/A | 15%+ | N/A |
| ASML | 350 | N/A (FY est.) | 7-25% | N/A | High-NA ramp | N/A | N/A |
| Broadcom (AVGO) | 800 | 16 (Q3) | 22% | 78% | $110B backlog | N/A | N/A |
| Super Micro (SMCI) | 50 | 5.8 (Q4) | 8% | 9.5% | 5K racks/mo | N/A | N/A |
| TSMC | 900 | N/A (Aug) | 34% | N/A | 70% CoWoS for Nvidia | N/A | N/A |
| Alphabet (GOOGL) | 2000 | N/A | N/A | 59% | 4.15% FCF yield | N/A | N/A |
| MongoDB (MDB) | 20 | 0.591 | 24% | 74% | 59.9K customers | N/A | N/A |
Sources: Company filings and analyst consensus as of Sept 2025. Figures approximate for comparability.
This table highlights alignment: revenue surges (avg. 25% YoY) with improving margins, backlogs signaling multi-year visibility.
The Road Ahead: Charts of Compounding Power
Visualizing the opportunity, consider Atlas growth for MongoDB versus S&P— a line chart underscores the divergence.
MongoDB’s Atlas line steepens to 29% YoY in Q2 2025, dwarfing the S&P’s 2% quarterly creep—proof of AI’s pull.
For broader context, a bar chart of YoY revenue growth across picks reveals the pack’s punch.
Bars tower 20-35% for most, with Super Micro’s QoQ focus tempering its YoY—yet all eclipse the index.
Why Now? Inflation-Proofing Your Portfolio
Inflation at 2.5% in September 2025 erodes cash, but these stocks thrive: AI demand is inelastic, with hyperscalers budgeting $200B+ annually. Rate cuts? They juice capex, supercharging backlogs. Risks linger—supply snarls, geopolitics—but diversified exposure (60% mega/mid-cap) mitigates.
My edge: blending Vertiv’s value (PEG 1.2) with MongoDB’s growth (29% Atlas). Portfolio tilt: 40% mega (TSMC, Alphabet, Broadcom), 40% mid (Arista, Super Micro, Vertiv), 20% small (MongoDB). Expected return: 20-25% annualized, beating S&P by 10-15%.
As Moore’s Law bends but doesn’t break, these eight aren’t just buys—they’re bets on the digital decade. In a world of noise, quiet strength wins. What’s your next hidden gem?