Chris Rolland

CFO tied both to the September mobile ramp and internal capacity expansion, noting $1 billion inventory is within normal range.

Deflected to closing the current deal first and explicitly declined to name target categories or discuss valuations.

Phil described very dynamic supply environment with broad-based input price increases. Sharing costs selectively with customers in a balanced way. Philip Carter added they can increase prices on long-life products and reaffirmed long-term 50-55% gross margin target post-Qorvo.

Phil described offering a technology-advantaged solution with multigenerational commitment that enables engineering focus. Win based on technology differentiation and sustained value generation. Will continue pursuing economically attractive opportunities but remains financially disciplined.

Philip Carter confirmed Android down sequentially from prior quarter's ~$100M level. Expects double-digit growth from Q1 to Q2 but growth to moderate through the year as company remains selective on devices.

Philip Carter attributed Q2 margin decline of 160bps to typical mobile seasonality, lower volume, and slightly higher Android mix. Also noted three consecutive quarters of exceeding guidance has driven some expedite fees. Nothing abnormal from typical seasonality.

Phil clarified Android strength is concentrated in Google where Skyworks has a strong footprint. Strategy remains focused on premium segment that values integration and performance. Deliberately defocusing from lower-end commodity space.

Phil pushed back on characterization of Qorvo deal as doubling down rather than diversifying. Argued deal reduces customer concentration and represents concentration in wireless not just handsets. Would not expect major transformational activities while deal is pending.

Phil confirmed Woburn closure is a tailwind across utilization, CapEx, OpEx, and fixed costs but declined to provide specific numbers. Rob noted 16% FY25 OpEx growth reflects timing and targeted R&D investments, expects discipline in FY26.

Phil acknowledged customer concentration suppresses valuation multiples. Priorities for diversification: gross margin accretive, stickier revenue, longer time constant to balance handset volatility. Must walk and chew gum, maintain focus on largest customer while pursuing growth.

Phil declined to comment on specific adjacencies but highlighted wide range of core technologies. Confirmed BAW remains a critical cornerstone of investment with robust roadmap.

Kris confirmed all content loss stems from the single socket share loss. Won back slots elsewhere. 20-25% range depends on mix between new vs legacy product and SKU mix.

Hassane attributed the AI DC TAM jump to gigawatt-scale data center buildout projected to 2030 plus new products expanding content per rack.

El-Khoury said trends are market-dependent by region. China strong in auto, North America strong in AI and industrial. Europe auto going sideways. Industrial better than expected. Energy storage resuming aggressive growth in NA and China.

El-Khoury said wins span from SPS at XPU level through medium/high-voltage discrete at rack, to SiC MOSFET/JFET at power boxes. Hyperscaler architecturally defined with power rack suppliers like FlexPower across all hyperscalers.

El-Khoury said ON does M&A for strategic portfolio reasons, not fab fill. Treo already provides embedded compute capability. Always looking for value and complementary opportunities for margin or market expansion.

El-Khoury said SiC share strong in China 800V market, gaining in North America, proliferating in Europe. SiC JFET growing in AI DC. Not converting to 300mm. 8-inch in production at 350-micron.

El-Khoury said SiC expanding into new applications like AI data center and plug-in hybrids. Geographic outlook matches EV and AI deployment: strong in China and US. 300mm too early. 8-inch in production at 350-micron, shipping in 2026.

El-Khoury confirmed ON addresses all mentioned applications today. The ability to address the full power tree is the differentiation. Will continue to expand portfolio as share is gained.

El-Khoury said weeks are at target sweet spot of 9-11. Dollar amounts will increase as sustainable recovery materializes. Inventory management is disciplined, not strategic drain.

El-Khoury said ON has manufacturing in China plus foundry relationships. 19 global factories with most products qualified at multiple sites. No need for reactive footprint changes.

El-Khoury said LTSA renegotiations are not new, have been ongoing throughout 2024. Not related to the Q1 demand drop.

El-Khoury said image sensors are stable, refocusing on machine vision where image quality adds value. Ramping 8-megapixel sensor.

Bill confirmed the two initial Dust design wins do NOT yet include the Credo DSP, framing DSP+PIC co-attach as long-term upside.

Bill said Credo is strongest in scale-out, with front-end coming along and a couple of customers now deploying in switch racks and disaggregated chassis.

Dan reiterated the mid-single-digit sequential growth framework through fiscal 27, set 3-4 quarters ago and consistently outperformed since.

Dan said Q3 OpEx of ~$70 million sets a new base (up 22% sequentially), reflecting project spend and hiring including Hyperlume, for ~50% year-over-year OpEx growth.

Asked whether emerging 1.6T optical and laser supply constraints hurt the DSP outlook or accelerate AECs; Brennan said constraints do not slow optical growth and customers will use AECs wherever they can.

Asked when a fifth hyperscaler could become a 10% customer; Brennan said first traction is expected toward the end of the fiscal year and all hyperscalers have the ability to reach 10%, but declined to commit on branded cables for the large GPU vendor.

Brennan said copper stays large but the biggest investments are in optical, expects optical above 10% over time, and sees a growing IC opportunity in AI inference.

Bill pushed back on the air-pocket/handoff framing, describing diversification of shipments across customers through fiscal '26 rather than project cliffs.

Michael confirmed a first ELS module order at meaningfully higher ASP but slightly lower (above-corporate) margins, trading some margin for revenue.

CEO acknowledged OCS ramp as their tightest rope alongside pump lasers and CPO. Assessed competitive position as very strong for the next year but noted it will not last forever. Working on cost reduction and innovative MEMS solutions.

CEO confirmed road maps contemplating PICs, photodiodes, laser drivers, and vertically integrated ELS modules. Described ELS as a shorter path to market than supplying lasers alone.

Michael said capacity improvements came from yield/throughput on 4-inch wafers, not heavy CapEx. Both CW and EML will ramp with 1.6T. Completely sold out for next 6 quarters with LTA protection. Wupen added that fab capacity is interchangeable between lasers and EMLs.

Michael explained they are consolidating customer base rather than expanding it, signing multi-year LTAs with strategic partners. Wupen noted 200G EML applications exist today with 2 customers not dependent on 102.4T switches.

Alan cited 18-month lead times for capital equipment and qualification. Repurposed Japanese wafer fab that was slated for closure. Adding capacity beyond 40% after June quarter. Barriers to entry protect long-term position.

Alan explained vertical integration deferred because EML capacity is more valuable serving external chip customers than converting to CW lasers for internal use. EML-based transceiver designs coming in second half CY2025.

Alan noted different transceiver architectures suit different applications. EML advantage most pronounced at 1.6T and higher speeds with multi-wavelength. 200G EMLs will differentiate from competitors and drive higher margins.

Steve ruled out large M&A but cited the small Hailo acquisition (AI at the edge) as advancing the roadmap by about five years.

Steve said constraint means they could use more capacity than allocated, not that growth is capped; ruled out an internal 300mm fab.

Rich noted FPGA usage in AI vision/detection and factory inspection is the fastest growing AI area. Declined to provide specific AI revenue breakdown. AI coding assistant launched in February showing 40% productivity improvement.

Eric corrected market estimate: typical December seasonality is down a couple percent, March up 2-3%. Steve expects above-seasonal performance in both quarters without quantifying.

Eric confirmed AI grew from 4% to over 6% of sales. Created a dedicated AIML product group to coordinate AI accelerators across eight business units.

Steve expressed strong conviction in MCU share gains, citing customers returning from inventory depletion and June quarter guidance at upper end of industry.

Ganesh noted AI-driven data center (~30% of total DC revenue) is growing while non-AI budgets are squeezed. Confident in long-term opportunity.

Eric said FY25 CapEx ~$150M with FY26 even lower. Ganesh noted CHIPS Act negotiations progressing, expects final agreement by end of December.

Jim confirmed Vietnam facility for materials/SiC business providing geographic flexibility. SiC demand stabilized after FY25 headwind, expecting some growth in FY26.

Jim confirmed share gains in FY25 per third-party data, with pricing as expected and no unusual pressure. Ramping capacity across assembly, InP, EMLs, and CW lasers.

Jim confirmed U.S. hyperscaler customers served almost entirely from Malaysia today. Internal EMLs are majority of EML-based transceivers but external partners also used.

Jim reported stable transceiver pricing, confirmed more product cost work ahead including yields and input costs across the company.

Jim confirmed no changes in customer orders or forecasts from DeepSeek, views efficiency improvements as long-term positive for AI adoption. No impact on 800G/1.6T timeline.

Jim acknowledged limited pricing upside in transceivers vs industrial, but confirmed sizable product cost opportunity particularly through yield improvements.

Jim confirmed dual-source EML strategy (internal + partners) for 800G/1.6T, noted no major capacity constraints so far thanks to production team's strong execution.