Open in interactive viewer → charts, metric popovers & call review

Peraso Reports $0.3 Million of Revenue, a $10 Million Rescue, and Not One Analyst Shows Up

The fixed wireless floor fell out, the strategic-alternatives review closed without a sale, and the defense-drone story slid another quarter to the right — on a call with an empty question queue
MOSY · Earnings Call · 2026-11-12

An empty queue is the loudest thing on the tape

On November 12, Peraso — trading as MOSY, a roughly $63 million market cap name — reported third-quarter 2026 net revenue of $0.3 million. The year-ago quarter was $3.2 million. That is a 90% collapse in the core business, and it happened in a quarter where the CEO's prepared remarks were almost entirely about drones, jamming, humanoid robots and millimeter-wave. Then the operator opened the floor.

I show there are no questions in the queue at this time. That will conclude today's conference call.

Operator · 2026-11-12
Three months ago this call had two analysts working the line — Jon Hickman of Ladenburg and Kevin Liu — pressing on NATO channels, drone volumes and the Roth equity facility. Now: silence. You don't need to model a sell-side estimate cut to read what a zero-question queue means for a sub-$100 million semiconductor story. The coverage has thinned to nothing at the exact moment management is asking the market to underwrite a defense pivot.

The $2 million order, and the memory squeeze that is starving it

To be fair, there is a real datapoint. In October Peraso booked a production order in excess of $2 million for a 60 GHz fixed wireless deployment, which CEO Ron Glibbery called “the largest fixed wireless order that we've received in six quarters” — Ron Glibbery, CEO · 2026-11-12, with shipments stretching through mid-2027. He also name-dropped Verizon's commentary about giving millimeter-wave "a second life" in fixed wireless — carefully, without claiming the carrier as a customer. But the same remarks explain why the last six quarters were so thin, and it's a deliciously ironic macro linkage. The bottleneck isn't demand for Peraso's chips. It's memory cost. Elevated DDR4/DDR5 prices and constrained component supply — the very shortage that has been minting money for the high-bandwidth-memory complex — is choking the smaller fixed wireless OEMs who buy Peraso's silicon. Glibbery conceded it will take time: “Although we believe it will take at least several quarters for recent market dynamics to fully normalize” — Ron Glibbery, CEO · 2026-11-12. The memory pricing tailwind for AI is a headwind for Peraso's customer base. That's a genuine cross-theme insight buried in a bad quarter. The balance sheet is the other half of the fixed-wireless problem. Cash stood at roughly $0.6 million on September 30, down from $3.3 million at June 30. CFO commentary in the prepared remarks notes the company drew only about $0.1 million from the $25 million Roth committed equity facility in the quarter and about $1 million after it. That is not a company with the luxury of waiting for memory to normalize on its own schedule.

The defense pivot keeps sliding one quarter to the right

This is where the prior transcripts earn their keep. In August, Glibbery was emphatic about timing on ground-based tactical comms: “I can see us starting to ship literally later this year on ground-based applications” — Ronald Glibbery, Chief Executive Officer · 2026-08-11 — "later this year" being Q4 2026, i.e. now. He also told Jon Hickman the company had been “engaging with new drone customers, and I mean design wins on a weekly basis” — Ronald Glibbery, Chief Executive Officer · 2026-08-11. Compare the language this quarter. Ground-based tactical networking is now described as being in "late-stage evaluations" with multiple prospects, and "increasingly viewing ad hoc ground-based networks as a realistic near-term revenue opportunity." Evaluations and near-term opportunity are not shipping this year. The drone business is "expected to unlock future complementary growth." The timeline didn't reverse — it just quietly changed tenses. Meanwhile the new product, the PRM2145 jam-resistant module for UAVs, is a real design win in search of a purchase order. And this theme is not Peraso's alone. Across the last five days of reporters, the Defence Drone Initiative shows up at both DPRO and FLT.TO as a marketplace they are plugging into; PDYN is marketing a low-cost Counter UAS interceptor; INVZ is pitching scalable LiDAR for defense applications. Peraso is riding a broad drone-defense wave, not creating one — and unlike several of those peers, its revenue line hasn't caught the wave yet. Even the hybrid architecture it sketches, pairing 60 GHz links with LEO backhaul, echoes the satellite platform language RKLB used this week. Good positioning; unproven conversion.

The rescue that closed the door on a sale

The most consequential line in the prepared remarks isn't about technology at all. It's Kevin Liu's disclosure that “the company has concluded its previously disclosed review of strategic alternatives” — Kevin Liu, Interim CFO · 2026-11-12. For several quarters Peraso has been openly exploring a merger, asset sale, or "other similar transaction." That process is over. What replaced it is a $10 million strategic investment from NantCapital, earmarked for working capital and drone-communication development, leaving roughly 20.9 million shares outstanding. Read that twice. The strategic-alternatives review — the thing you run when a sale is the best outcome — ended not with a transaction but with a dilutive rescue from a single investor. That is the single biggest change at this company this quarter, and it happened without a single analyst on the line to ask what NantCapital paid per share, what covenants came with it, or whether the runway now extends beyond 2027. The optics inside the quarter are equally unhelpful. The headline 78.5% gross margin looks spectacular against 63.7% the prior quarter — but Liu attributed a meaningful chunk of it to "a more favorable mix of millimeter-wave product revenue and a one-time benefit of approximately $55,000 from the reversal of inventory accrual," on a base of only $0.3 million of revenue. A high margin on almost no revenue is arithmetic, not progress.

The tape is voting on the wrong half of the story

The market's own theme tape has been souring on precisely the adjacent cluster Peraso wants credit for. Over the last 30 days, the AI-networking cohort has bled: co packaged optics is down double digits across a dozen tickers, PCIe Gen 6 and gig per lane names are off, and high bandwidth memory has rolled over after a strong year. Peraso's autonomous edge AI pitch — the humanoid-robot-to-edge-server demo from October — sits one hop away from that de-rating cluster. The company is telling a defense-and-robotics story while the tape is actively marking down the networking and memory-adjacent complex that its fixed wireless customers depend on. So what actually changed? Three things, in order of importance. The strategic-alternatives process ended without a sale and was replaced by a $10 million check and a narrower runway. The fixed-wireless business printed its worst quarter in memory — literally — while management pointed to a single $2 million order as the turn. And the defense narrative, sold in August as imminent Q4 shipments, arrived in November as "late-stage evaluations." The stock is a $63 million call option on 60 GHz millimeter-wave finding a home in a drone. The company just bought itself more time to prove it — and lost the audience that would have asked whether that time is enough.