Ren
Head of AI, Ex-Product Manager, AI Supply-Chain Bottlenecks
MaxLinear is being repriced as a scale-up play: it is one of only three companies, with Marvell and Broadcom, making DSPs for 1.6T optical interconnects, yet is worth about $2B against their $50B+. Its Keystone DSP ships to multiple hyperscalers at 400G and 800G, and its Annapurna platform adds 1.6T AEC and 3.2T retimers for scale-up. Infrastructure revenue grew 130%, revenue is projected at $630M by 2028 from about $200M, and Applied Optoelectronics' 800G and 1.6T ramp needs DSPs. x.com
MaxLinear's longer-term DSP demand depends on which near-packaged-optics architecture customers adopt. Keysight's Kazuo Yamaguchi explains that retimed NPO retains a DSP and allows independent engine testing, while linear-drive NPO removes it. Linear-drive designs make failures harder to isolate. The DSP risk remains worth considering: architecture choice and whose DSP customers select both affect how much of this opportunity MaxLinear captures. x.com
MaxLinear's 1.6T roadmap threatens Credo's dominant AEC position with a specific share target. MaxLinear targets 20–30% of the 1.6T active electrical cable cycle, where its management estimates Credo has 93–99% share, initially across 1–2 US data centers. MaxLinear also reports 1.6T DSP and TIA design-win commitments, with initial revenue expected in 2027 and a stronger ramp in 2028. That competitive concern is behind Credo's pre-market drop. x.com
MaxLinear's AI optics ramp is now confirmed in its results and guidance. Q2 revenue rose 55% to $168.8M, beating estimates, operating margin jumped to 22.3% from 7.2%, and infrastructure grew 145% on optical AI data-center adoption. Q3 guidance of $210-220M implies 27% sequential growth. Keystone is ramping for 800G optics at hyperscale data centers, with 1.6T next, and a $19.7M inventory build suggests orders are already visible. x.com
MaxLinear has strong operating leverage as its 800G optical data center ramp runs far faster than expected. Q3 revenue is guided up about 27% sequentially with flat OpEx, lifting operating margin near 28% on the way to 30%+. Weak cash flow reflects prepaying for 5nm wafers to meet a six-month backlog. FY26 optical guidance rose to $210-230M entirely from 800G, and 1.6T Rushmore, retimers and storage accelerators ramp in 2027-2028 with massive upside. The sell-off looks way overdone. x.com x.com
MaxLinear's planned 1.6T AEC entry is too distant to justify the scale of Credo's selloff. Credo still has an estimated 90%+ AEC share and relationships with five of six major hyperscalers; MaxLinear expects initial revenue only in 2027 and a more meaningful ramp in 2028. Competition bears watching, but the market looks to be overreacting and Credo is starting to look ridiculously cheap. x.com x.com
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.com
MaxLinear's post-earnings reaction looks like an overreaction to solid growth. Q2 revenue rose 55% to $168.8M, infrastructure grew 145%, and adjusted operating margin expanded to 22.3% from 15.9%. Q3 guidance of $210-220M implies another ~27% sequential jump, 2026 optical data-center revenue guidance was raised again to $190-210M from $150-170M, and 1.6T products are in customer qualification. x.com
MaxLinear's PAM4 DSPs are crucial for 1.6T optical links. Headwinds remain from plenty of legal noise, such as the Silicon Motion merger that didn't go ahead. x.com
MaxLinear's optical data center business has hit an inflection point, and the stock could reach $70-80 this year. Q1 EPS of 22 cents beat the 18-cent consensus and revenue of $137.2M topped estimates, up 43% year over year as infrastructure grew over 130% to become its largest market. The CEO called Q1 the start of a multi-year growth phase, with a step-function revenue increase expected in Q2 from optical ramps at multiple hyperscale customers. x.com x.com
Unlocking Samsung production capacity would give MaxLinear a real edge in the capacity-constrained active electrical cable market. CEO Kishore Seendripu outlined an ambition for 20-30% share of the 1.6 Tbps AEC generation over the next couple of years, initially targeting one or two major US data-center customers and challenging dominant incumbent Credo Technology. x.com
