Week of July 6, 2026 · 9 stocks
Nvidia's picks-and-shovels crowd runs the board while an earnings-heavy week loads the chamber
From Nvidia's specialist partners to the power, cooling and networking names its build-out depends on, AI plumbing dominated the feed — with a Buffett-blessed airline and a pile of earnings-week hopefuls filling out the rest.
This week the crowd talked itself hoarse about AI plumbing. Of the 1,281 names and roughly 808,000 mentions we track, the genuine trending equities clustered hard around the build-out — the specialist partners (Penguin Solutions), the power-and-cooling incumbents (Vertiv), and the networking-and-silicon layer (Arista, Wolfspeed). Threaded through it: an unusually heavy July earnings calendar and Warren Buffett's improbable return to airlines. One housekeeping note — we stripped out a half-dozen "trending stocks" that were really memecoins wearing a listed company's cashtag (Axie Infinity under $AXS, a Pumpfun token under $PUMP, a browser-MMO coin under $KINS, and friends). What's left is what people are actually saying about actual stocks.
Two days out from its July 7 print, Penguin Solutions is the loudest AI-infrastructure name almost nobody had heard of — 1,192 posts and a stock that ran roughly 20% in three days to new highs. The hook: management pre-flagged on June 1 that both revenue and EPS would land at the high end of guidance, and the company carries an invitation-only Nvidia "AI Factory" specialist badge.
The crowd is unabashedly long, anchoring $75–$100+ into the print, pointing to call open interest up 10x and a position up ~3x since they found it. The only bear voice is technical-pullback caution.
The fundamentals argue for two-sided risk: 57x trailing earnings on a systems integrator whose last quarter actually shrank 6% to $343M at a 27% gross margin. Management has effectively underwritten the beat — the print decides whether the tape agrees.
Vertiv cut the ribbon on a new Johor, Malaysia plant this week to feed Asian AI-datacenter demand, days after closing its June 12 acquisition of Italian liquid-cooling maker ThermoKey. That's 1,214 posts and an 11% jump in daily mentions for the name traders call the power-and-cooling bottleneck the whole Nvidia build-out runs through.
The conversation treats the day's ~3.5% dip as a coil, not a crack — calendar-spread bets targeting an all-time-high reclaim, a big call sweep framed as a buyable pullback.
Underneath, the growth is real: Q1 revenue up 30% to $2.65B at a 36% gross margin. The catch is the ~74x trailing multiple, which prices in flawless execution and punishes any whiff of hyperscaler capex slowing. The July 29 report is the next referee.
Arista's week ran on a strategy headline: a formal collaboration with AMD and the Ultra Ethernet Consortium to define scale-out AI Ethernet — the open alternative to Nvidia's InfiniBand that hyperscalers actively want to fund. Cue 1,156 posts and chart-watchers marking a bull flag near $177 with a wave-3 target north of $280.
Bulls lean on the pristine balance sheet — zero debt, growing net cash — and outsized call buys as a pre-breakout tell.
Here the data cuts against the mood. Business is compounding at 35% (Q1 revenue $2.71B, 43% operating margin), but co-founder Andy Bechtolsheim sold about $77M of stock across June 9 and June 17, and CTO Ken Duda added a $7.4M sale on June 24. Founder-scale selling into the euphoria is the loudest counter-signal a compounder offers.
A curiosity, this one: Better Home & Finance drew only 76 posts and its daily mentions fell 45%, yet it cleared the trending cut on intensity rather than volume. The spark is a partnership with the ChatGPT team pitching "digital native homeownership," bundled with the falling-yields trade that also lifts Opendoor.
The thin chatter is uniformly constructive — insider buying flagged despite heavy stock comp, topline accelerating, no visible bear.
The books explain the caution the crowd is skipping. Revenue reaccelerated 53% to $53M, but operating margin sits at –85% and the company burned $77M of cash in Q1. The Street doesn't model a profit until FY27 (EPS –$6.27 this year, +$0.24 next). It's a turnaround narrative that still has to become a business; August 6 earnings is the checkpoint.
Up roughly 250% in three months and ripping ~15% on the day to about $52, Wolfspeed is running on a story change. GE Aerospace signed a memorandum to build high-voltage silicon-carbide power modules around Wolfspeed's 10 kV die, and China's Yangjie Tech said SiC capacity is fully booked — a read-through that the EV-glut hangover is finally clearing. That's 1,103 posts and a 27% pop in daily mentions.
The crowd has latched onto the aerospace-and-defense pivot as the escape from the consumer-EV price war, tossing around "$120 by end of July."
The mechanics make it violent: post-bankruptcy fresh-start accounting, a float near 6M shares and a beta above 8, offset by a 24M-share resale overhang from June's S-1. August 19's print tests whether the pivot is revenue or narrative.
Eight green days in a row and roughly +43% off the June lows put Shift4 near the top of the board with 1,157 posts. The fundamental peg is the 2026 World Cup, where Shift4 runs payments across US and Canada stadiums including MetLife, host of the final — a live showcase of high-volume reliability.
With short interest near 40%, the crowd reads the melt-up as mechanical short-covering stacked on a genuine growth story, citing 51% international growth behind the Smartpay Australia rollout.
The valuation is what keeps the bulls credible: Q1 revenue grew 32% to $1.12B, yet the stock trades under 8x forward earnings — unusually cheap for that pace in fintech. August 4 earnings tests whether World Cup volume shows up as real 2026 revenue.
The oddity of the week: Pure Storage cleared the trending screen on just 19 posts and essentially no crowd conversation at all — no bull thread, no bear thread, unusual for a stock up 62% over the trailing year. (Our data bundle even mislabels it "Everpure," thanks to an entity-name collision.)
So there's little crowd texture to report — which is itself the point. This is a name the chatter is ignoring while the fundamentals do the talking.
And they're talking: Q4 FY26 revenue up 20% to $1.06B at a 70% gross margin, with sell-side models pointing toward ~$4.37B next year on the hyperscale AI-storage tailwind. A quietly strong operator riding the same capex wave the loud names are chasing — just without the noise.
Buffett bought an airline again — the same Warren Buffett who swore them off in 2020 — and the disclosed $2.65B Berkshire-adjacent stake in Delta lit up the feed with 635 posts. Goldman piled on, lifting its target to $116 from $80, as falling oil pushed DAL, AAL and UAL to multi-year highs.
Traders are framing it as boring value turned momentum, a legacy carrier suddenly wearing a growth costume into its July report.
The numbers back the re-rating better than most airlines can: 12x trailing earnings, 23% ROE, 8.3% return on capital and $1.2B of quarterly free cash flow. The wrinkle is the tape itself — sitting at 94% of its 52-week range on light volume, extended enough that the July 9 print either extends the melt-up or reverses it fast.
The most boring name on the board is coiled for a reason: PepsiCo reports July 9, and the stock has gone nowhere for a year. Dividend accumulators are quietly adding under $136 while swing traders nurse a roughly –5% five-year total return — and, fair warning, a slice of the "$PEP" chatter is actually Pepecoin, not the soda.
The conversation is genuinely split, with no clean catalyst thesis on either side.
The fundamentals read defensive-and-cheap: Q1 revenue up 8.5% to $19.4B, a 44% return on equity funding the 3.5% dividend, and 22x trailing earnings — undemanding versus Coke and Mondelez on the same growth. The swing factor is the July 9 print against a persistent GLP-1 snacking-headwind narrative that keeps the volume story honest.
As always, the live mention counts, sentiment and underlying data are on TickerTalks — this is what the crowd is saying, not what you should do about it.
Want this live, not weekly?
See what's trending right now →TickerTalks is a research tool, not financial advice. This recap describes what accounts on X are discussing — it is not a recommendation to buy or sell any security. Social attention is not predictive of future price moves; investing involves risk of loss.
