Every once in a while, the market hands you one of those moments where a company flies completely under the radar, posts ridiculous numbers, and suddenly every analyst on Wall Street scrambles to play catch-up.
That moment just happened with Credo Technology (CRDO).
While everyone has been obsessed with the usual suspects in the AI trade, this lesser-known player in high-speed connectivity just delivered a fiscal second-quarter beat so big it left analysts speechless, then followed it up with guidance that essentially says “hold my beer.” The stock is already up 155% year-to-date and outperformed even Nvidia and Broadcom, yet I still feel like most retail investors have never heard the name.
The Quarter That Broke the Internet (Literally)
Let us start with the numbers, because they are almost comical in how much they smashed expectations.
Adjusted earnings came in at 67 cents per share. Street was looking for 49 cents. Revenue hit $268 million against the $235 million whisper number. Fine, solid beat, happens every day, right?
Then came the guidance bomb: $335 million to $345 million for the current quarter when analysts were modeling $247 million. That is not a beat. That is an annihilation.
Shares ripped 18% higher in Tuesday premarket trading, and honestly, that felt conservative given the numbers.
Why Credo Actually Matters in the AI Buildout
Here is the part most people miss: GPUs are sexy, but without insane connectivity between them, the whole AI cluster falls apart.
Modern AI training clusters now have tens of thousands of GPUs that need to talk to each other at mind-bending speeds with almost zero latency. We are talking 800G, 1.6T, and soon 3.2T connections. The copper cables that worked fine five years ago? They are now the bottleneck.
Credo makes two things the hyperscalers desperately need right now:
- Active Electrical Cables (AECs) — basically supercharged copper cables with built-in retimers that can push 800G+ over longer distances at lower power and lower cost than optical solutions.
- Optical DSP chips — for when you absolutely have to go optical (the highest-end stuff).
Management now says their long-term total addressable market has tripled in the last year to $10 billion. That is not marketing fluff — that is new products and new customers coming online faster than anyone modeled.
Wall Street Finally Wakes Up
The analyst reaction was immediate and overwhelming. Almost every major firm hiked their price target — many by 40-50% in a single morning.
“Significant beat and raise with new customer ramps and new product solutions layering on to growth in FY27 and beyond… CRDO is turning into a more well-rounded interconnect play.”
Barclays — raised target from $165 → $220
“We see CRDO F26E revenues ~$1.19B still small compared to a ~$22B scale-out cable market… long-term TAM of $10B, triple last year with new products.”
Mizuho — $165 → $225
“No Thanksgiving hangover here — strong results and stronger guide likely cleared even the highest of buyside bars… path to $5B in revenue in the ‘next several years’ begins to address the next leg of the Credo story.”
TD Cowen — $190 → $240, Top Pick
Bank of America and Stifel both went to $240 and $225 respectively, with BofA specifically calling out that AECs are now ramping at five large hyperscalers — up from basically one major customer a year ago.
In my experience, when you see this kind of synchronized target inflation the morning after earnings, it usually means the Street was materially behind the story.
The Customer Diversification Story Is Just Starting
For the longest time, the bear case on Credo was “it’s a one-customer story” (read: too much exposure to a single hyperscaler). That thesis just died.
Management guided to mid-single-digit sequential growth every quarter through fiscal 2027 on the AEC side alone. That is code for “we have multiple new hyperscalers ramping in parallel and none of them are cutting back.”
Add in PCIe retimers, optical DSPs, and 1.6T ethernet solutions that are all hitting the market over the next 12–18 months, and you start to understand why the company feels comfortable throwing out a $10 billion long-term TAM figure.
Valuation: Expensive or Actually Reasonable?
Yes, the stock trades at roughly 65–70x forward earnings at $170–190. That is not cheap.
But let us run the math the way growth investors do:
- Consensus for FY26 (ending Apr 2026) is now ~$1.15–1.2 billion revenue → ~170–180% y/y growth.
- Management basically endorsed that number and said FY27 will grow mid-single-digits sequentially each quarter off that base → call it another 30–40% y/y.
- That puts Credo on a path to $2 billion+ revenue in calendar 2027 with 50%+ gross margins and improving operating leverage.
Suddenly 25–30x 2027 earnings does not look insane for a company growing revenue at 50–100% clips with a $10 billion TAM in sight.
Or to put it more bluntly: if the AI buildout is real (and every hyperscaler capex guide says it is), then the connectivity layer is going to be a multi-decade compounding story. Credo is one of maybe three public pure-play ways to invest in that theme.
It's going to be a year of volatility, a year of uncertainty. But that doesn't necessarily mean it's going to be a poor investment year at all.
