Ryan Koontz

CEO noted scale-across constraints exceed even EML constraints. Pump laser output expected to increase faster than EMLs due to fewer fab constraints.

CEO provided specific metrics on supply-demand gap (>30%) and 50% YoY supply increase target. Acknowledged still significantly lagging demand despite active ramp.

Michael confirmed initial 1.6T wave is heavily EML-driven with 200G EML a quarter ahead of plan. SiPho expected to be majority at 1.6T but absolute EML volumes will grow appreciably. Differential 200G EMLs offer another ASP tailwind. Wupen distinguished WDM applications (EML dominant) from parallel fiber (SiPho gaining).

Michael confirmed 2-3 month pushout on internal CW laser integration to late Q2/early Q3. On CPO, cited 400mW power level few can match plus subsea-proven reliability as key competitive advantages driving broader customer engagement than market appreciates.

Michael cited very high market share with limited competitive landscape. Customer base is traditional telecom OEMs pivoting toward hyperscalers. Wupen added 10-15 year heritage in this technology with mastered laser capacity.

Michael outlined CW laser roadmap: 70mW shipping now as pipe cleaner to external customers, 100mW sampling this month for internal transceiver use, full production expected mid-2026.

Michael said despite noise about silicon photonics, EML demand is unaffected, completely sold out. At 1.6T speeds, equal EML and SiPho designs exist. Supply constraints will determine technology adoption.

Wupen cited roughly 30% volume annual growth for ZR pluggable modules. Sold out on narrow linewidth lasers for balance of FY2026. Focusing on component sales rather than ZR module sales due to margin considerations.

Michael confirmed customer interest in non-China supply but demand exceeds collective non-China capacity. Some customer design insourcing but outsourcing remains dominant. Lumentum expects to become more meaningful participant starting next quarter.

Wajid confirmed DCI products accretive even above 40% GM target. Combination of secular demand, internal Thailand manufacturing, and pricing power drives margin confidence.

Alan sees DCI pluggable growth as additive to traditional coherent, not cannibalistic. Adding factory capital to meet demand from ZR module customers. Both markets growing.

Alan admitted being late to 400G ZR module market with small share, but making good progress on 800G ZR/ZR+. Biggest DCI revenue still from tunable laser components sold to other module makers.

Wajid expects sequential gross margin improvement through FY but muted at operating level due to increased R&D spending ahead of revenue and Thailand capacity overhead costs.

Gary said Ciena is largely immune to data-center construction pacing for a couple of years; backlog reflects existing data centers and GPU refresh.

Marc said plugs and RLS line systems are growing well above corporate average, with the DCOM ramp accretive to margins going forward.

Gary cited multiple defensibility layers: deep customer collaboration originating with Meta, vertical integration of core technology, software integration capability, and installation services. Views the confluence of these elements as highly defensible.

Gary confirmed three hyperscalers deploying scale across with additional sites from the first customer. All currently North America-centric. Plans are large and expansive but each hyperscaler has different architectures, distances, and clustering approaches.

Gary confirmed tight relationships with fiber providers like Corning and wholesale carriers like Lumen. Ciena's services business is a major growth vector, with a cloud provider becoming largest service customer for the first time in FY2025.

Gary described shift toward higher proportion of line systems (scale across track-laying phase), rapidly ramping 800G plugs, increased line system intelligence needs, and DCOM as an unplanned addition that changed the cloud portfolio mix significantly.

Scott confirmed Ciena is the most vertically integrated optical modem supplier in the industry, using both own components and third-party suppliers for strategic flexibility.

Gary confirmed both dynamics: competitive takeaways driven by 18-24 month technology lead, plus broader SP infrastructure spending recovery after five years of underinvestment.

Scott confirmed ZR pluggable volume is dominated by cloud customers. Service providers use pluggables for different modularity reasons but at lower volumes.

Jim outlined multiple mitigation levers (manufacturing moves, supply chain adjustments, price increases) but declined to detail specifics. Expects net cost to be immaterial going forward.

Gary confirmed both breadth (expanding set of cloud customers) and depth (deeper relationships with existing ones across submarine, metro, pluggables, and services).

Gary attributed NA SP strength to three factors: Ciena-specific design wins maturing, inventory normalization completing, and return to infrastructure investment after 5G focus.

Spice described a wide Space Systems margin range, with component solutions such as solar toward the low 30s and some products north of 70%, while the larger SDA Tranche 2 and Tranche 3 satellite platforms sit in the mid-30s, so quarterly margin depends on the mix of high-end components versus programmatic work.

Beck said acquisitions resolve chronic supply pain like electric propulsion, and vertical integration strengthens cost and delivery competitiveness.

Spice said the mix is trending toward government, which he prefers for reliable payment, while a large commercial pipeline also remains.

Beck credited strong commercial and, for the first time, sovereign space-agency bookings standardizing on Electron.

Spice said SolAero gross margins exceeded the 30% long-term target in the first half of 2025, up from high-single-digits at acquisition.

Beck said STARRAY modular solar arrays address quick-turn customer demand for complete arrays not readily available in the U.S.

Jim identified indium phosphide as the primary constraint, noting assembly and test capacity is available so the growth limiter is the 6-inch ramp.

Jim reiterated Apple partnership with revenue starting second half of calendar 2026. 6-inch gallium arsenide VCSEL technology manufactured at Sherman TX. Another example of customer leveraging US manufacturing footprint.

Jim confirmed multi-rail revenue contribution starting second half of calendar 2026. Communications segment grew 44% YoY with participation at system, module (ZR/ZR+), and component (pump lasers) levels.

Jim reiterated OCS revenue expected calendar 2025 with existing customer orders. Digital liquid crystal technology offers higher reliability vs MEMS solutions.

Jim confirmed significant DCI opportunity both as module maker and component supplier, noted good demand and reasonable gross margins, deferred detail to Investor Day.

Francois noted sovereignty dialogue ongoing for several quarters but accelerated this quarter as NIST 2 and DORA enforcement deadlines approach. Mix of new projects and deferred refreshes now happening under compliance urgency. More regulation coming (Cyber Resilience Act, 2027). F5's deployment flexibility uniquely suited for regulatory uncertainty.

Cooper confirmed no changes to end-of-support dates. Still early in refresh with well over 50% of installed base on legacy platforms. Working with customers for orderly refresh path.

Francois noted 4G-to-5G transition continuing with capacity growing in early-adopter geographies. However, expected 5G revenue uplift for carriers hasn't materialized, constraining CapEx. New use cases being found but not enough to drive significant segment growth.

Cooper confirmed January price increase is flowing through. Customer reception reasonable. Competitors' more aggressive pricing practices are driving business to F5. F5 focused on delivering value commensurate with price increases.

Francois described two displacement categories: 1) new accounts where F5 wins less than half of installed base initially, still early innings; 2) consolidating point products onto F5 platform within existing accounts, accelerating with ADSP launch.

Ken Duda described extending EOS state-streaming architecture for AI with in-network RDMA counters, buffering, congestion data plus host-level NIC telemetry unified in CloudVision.

Todd highlighted two-way synergy: cross-selling SD-WAN into campus accounts and bringing Arista's full portfolio through VeloCloud's strong MSP channel.

Jayshree noted NeoCloud customers more adventurous with alternative AI accelerators and Ethernet, citing Ultra Ethernet consortium spec progress.

Jayshree explained CPO has weak adoption due to field failures and serviceability concerns, with Arista favoring co-packaged copper and pluggable optics approach.