Temple 8 Research
AI Buildout, Macro & Energy Infrastructure Research
Marvell looks well positioned for scale-up compute domains, where interconnect is the wall that limits performance. Scale-up domains should be thought of differently from simply networking racks together, and Marvell's own commentary now focuses on exactly that opportunity. x.com x.comÂ
Marvell is the first beneficiary of coherent-lite optics adoption, as @tengyanAI argues, a view worth agreeing with. It would be a buy now if the macro situation were not so volatile. x.com x.comÂ
Marvell risks paying a peak-scarcity price for Lumentum if optical supply catches up by 2028-30. The proposed ~$99B deal values Lumentum at roughly 20x guided annualized revenue and 58x estimated after-tax operating income, while Lumentum, Coherent and Chinese suppliers are adding laser capacity. The DSP-and-laser combination has strategic logic, but durable scarcity economics and Nvidia's response remain uncertain. x.comÂ
Owning Lumentum's laser fab could secure Marvell's FY28 supply and hedge its DSP franchise as optical links shift to CPO. The rumored deal's strategy makes sense, but short-term pain is likely: at a 30% premium, roughly $17.5B in new debt and the rest in stock, rough math puts EPS dilution near 15%. It turns accretive only if Lumentum outgrows Marvell at today's guided rates for five or six more quarters, which looks achievable while the laser shortage lasts. x.com x.com x.comÂ
Rising 1.6T switching demand favors optical suppliers, though their execution risks vary. Bank of America says the market could grow ~101% annually through 2030. Marvell's leading DSPs and Lumentum's supply-short 200G EML lasers are lowest risk; Applied Optoelectronics' capacity expansion and Fabrinet's customer concentration make them medium risk. Aeluma remains pre-commercial, while Lightwave Logic's polymer modulators are a high-risk moonshot despite its TSM partnership. x.comÂ
Marvell's management is running the stock as a pump and dump to move shareholders' money into their own pockets. Wall Street prices Marvell as an ASIC design house without realizing that ASIC design is just IT outsourcing. It is not a real CXL play either: CXL IP is sold by Cadence, Synopsys and Rambus with no moat, and Astera Labs, which builds the switches and software around CXL, is the only good CXL play. x.comÂ
Marvell looks interesting at less than 30x calendar-2027 EPS against expected 60-70% annual EPS growth, but overpaying for Lumentum could drive its shares lower. Marvell could pay up to $110B, roughly a 30% premium; the stock-versus-debt mix is unclear, and Lumentum holders might own a third of the combined company. If the deal falls apart, Marvell could rise 10-12%. x.comÂ
Marvell expects its revenue to grow tenfold in five years to $80B, driven by custom chips it builds for hyperscalers. Those companies design their own chips to reduce reliance on Nvidia and cut the long-term cost of off-the-shelf parts. The custom segment had no revenue two years ago and is expected to bring in $30B by 2031. x.comÂ
Marvell looks very expensive on 2027 estimates compared with Nvidia, though not a short at current levels. Memory spot is also a better bet, despite the week's fear. x.comÂ
Marvell's expanded multi-year deal with GlobalFoundries is very bullish, adding silicon germanium capacity at GF's Burlington, Vermont facility to support Marvell's next-gen pluggable, near-packaged and co-packaged optics as AI data centers push toward higher-bandwidth interconnects, per @wallstengine. x.comÂ