Neutral
MRVLMarvellMarvell
LITELumentum Holdings Inc.Lumentum Holdings Inc.
MSFTMicrosoftMicrosoft
ORCLOracleOracle
PLTRPalantirPalantir Weekly|Anthropic S-1, CSP Capex & FCF, Cybersecurity, MSFT & NOW, OpenAI Ads, Gemini 4 Carbon, MRVL & LITE
Over the past few weeks, we have been highlighting a market that is becoming more demanding about capex and free cash flow. This quarter’s CSP results reinforced that shift: promises of incremental capex are receiving less credit, and increasingly raise concerns about when the spending will generate returns. Strong cloud growth alone is no longer enough to settle the debate. Investors want a clearer path from capacity commitments to utilization, earnings and cash generation. Oracle’s guidance for peak capex in FY27–FY28, followed by a decline in FY29, added to infrastructure investors’ anxiety about a spending peak, even as the prospect of better cash flow offered reassurance to Oracle’s own investors. Anthropic’s S-1 is especially relevant in that context. Roughly $132b of cash provides substantial funding, but about $800b of compute commitments makes the timing of revenue and costs central to the story. Our first financial model works through those obligations, margins and customer concentration. For CSPs, the same commitments can support growth while increasing dependence on a small number of large AI customers; the scale of a contract does not by itself establish its eventual cash return. The operating picture remains uneven. Microsoft reported 48% Azure growth and accelerating Copilot revenue, while Google Cloud grew 102.6% alongside greater infrastructure spending pressure. ServiceNow’s smaller-than-expected beat also showed how much work can sit between an AI order and production revenue. We underestimated the deployment support required. Across these businesses, we are watching how quickly demand converts into profitable revenue, and how much additional spending is needed along the way. Cybersecurity could become a much bigger focus from here. The roughly $6.3b theft from Binance on October 29, reported as the largest crypto theft on record, makes the scale of the security risk tangible, although AI’s role in this particular attack has not been established. AI is lowering the barrier to attacks and accelerating vulnerability discovery, creating a growing need for remediation and continuous protection. But a sharp increase in AI-enabled attacks may not translate quickly into revenue for listed security vendors. Many products are sold through subscriptions rather than consumption-based pricing, and customers still need to approve budgets and make purchasing decisions. Some of the incremental spending may also accrue to AI labs. We see a potentially significant longer-term demand shift, while looking for evidence in new budgets, contracts and guidance before assuming a near-term revenue acceleration. This Week’s Reports Anthropic: the S-1 clarifies cash, commitments and margins. Our first financial model examines roughly $132b of cash, about $800b of compute commitments, customer concentration and the assumptions still needed to estimate future earnings. MRVL and LITE: securing the laser bottleneck. We examine the strategic fit of a potential acquisition, including scarce laser supply and a broader optical stack, against dilution and margin risks; no final agreement has been announced. Review | MSFT FY27Q1: Azure grows 48%, with Copilot revenue accelerating. Large deployments and better fleet efficiency supported results, while delayed third-party capacity shifted capex timing and keeps delivery execution in focus. Review | NOW 26Q3: Healthy checks did not translate into the expected beat. We underestimated the deployment support needed to bring AI orders into production, with weaker cRPO and margin guidance pointing to slower monetization and greater upfront costs. Gemini 4 Carbon: Google returns to the top of the benchmark table. Carbon leads the updated AA index by a narrow margin, while faster AI-assisted development intensifies frontier competition and access remains limited. OpenAI advertising: hiring extends into acquisition and measurement. Our latest column connects the expansion of its commercial organization with early channel evidence of pressure on Google advertising, without treating hiring as proof of revenue scale. SaaS tracker: AI is changing which software bills grow. Our new quarterly series draws on 31 companies to examine rising spending on systems of record and data platforms alongside the replacement of some specialist tools. Cybersecurity: faster remediation can create new AI demand. We examine how more frequent vulnerability discovery and continuous protection could expand model usage and the revenue opportunity for security platforms. Preview | DDOG 3Q26: Enterprise AI expands core observability demand. Our checks point to stronger infrastructure, tracing and log consumption, with platform consolidation supporting a forecast of 36% revenue growth in 3Q26. Preview | PLTR 26Q3: Deployment expertise is becoming a constraint on adoption. Our preview examines scarce FDE capacity and the work required to demonstrate ROI, with total revenue forecast about 7% above the high end of guidance. Premium Report Snapshot Subscribe now Read more
View original →Neutral
MRVLMarvellMarvell
LITELumentum Holdings Inc.Lumentum Holdings Inc. Research|MRVL and LITE: Owning the Laser Bottleneck and the Optical Side of Interconnects
On October 28, Bloomberg reported that Marvell was in advanced talks to acquire Lumentum, with an announcement possible as soon as the following week. No final agreement had been reached. Bloomberg valued Lumentum at about $83b, compared with roughly $221b for Marvell. With just $3.92b in cash on its balance sheet, Marvell would likely need to fund a deal of this size largely with stock. At about 27 times Lumentum’s FY2026 revenue of $3.0b, the transaction would dwarf Marvell’s previous optical acquisitions: Inphi (about $10b in 2021) and Celestial AI ($3.25b, announced in late 2025). Our view is that Marvell would be accepting substantial shareholder dilution—and some pressure on gross margin—to secure three assets: a scarce component in AI optics, the photonics half of a more fully integrated electro-optical stack, and an optical-switching business that complements its electrical switches. We examine the implications below. Marvell Is Paying to Own the Bottleneck Subscribe now Read more
View original →Neutral
MRVLMarvellMarvell
AMD
AVGOBroadcomBroadcom
MSFTMicrosoftMicrosoft
NVDANVIDIANVIDIA
ORCLOracleOracle Research|Optics: OCP 2026 Points to NPO Adoption and New Demand from Copper Replacement
We attended the 2026 OCP Global Summit in San Jose from October 12 to 15, including keynotes from NVIDIA, Broadcom and AMD, and more than twenty optical-interconnect sessions. We also met with Marvell, Credo, Ciena, Cisco, Amphenol, Lightmatter, MACOM and others. Our week in San Jose yielded four key themes, discussed below. This note concludes with detailed notes for each of the companies we engaged with. Contents 1. Key Themes 2. NVIDIA: a CPO version of NVL576 but keeps NPO on the table 3. NPO: How everyone else responded 4. NPO: What users want 5. Pluggables strike back: XPO and USFP and their impact on NPO and CPO adoption 6. Light sources, fiber and test: the constraints in focus at OCP 7. Standards and the ecosystem: MSAs landing in quick succession; an argument against NPO multi-sourcing? 8. Copper’s evolving role 9. Company notes 1. Key Themes Theme 1: Scale-up is still a copper/optics mix; the debate is between NPO and CPO and how outstanding issues with both are resolved Nobody at OCP said copper is finished. Meta will not replace the copper backplane in scale-up for now, while Broadcom sees in-rack copper staying viable for two more years. The role of optics grows with scale: Astera Labs says pods beyond 144 GPUs need optics, with copper in the rack and optics between racks from 72 to 576 accelerators, and optical only above 576 accelerators. Image 1: Astera Labs’ hybrid copper/optical pod: hybrid from 72 to 576 accelerators, all-optical above 576. Source: FUNDA; Astera Labs slide. Where optics is needed, the debate is about where the optical engine sits: NVIDIA showed it soldered onto the switch package (CPO) while keeping NPO on the table. Meta, Microsoft, Oracle, Alibaba and Huawei lean toward board-mounted, socketed NPO. Alibaba also proposed IPO (Interposer-based optics), a solution in between NPO and CPO. A key issue is serviceability. The NPO camp sells “replaceable” while NVIDIA counters that NPO is assembled in clean rooms with woven pigtail fibers, so failures mean a factory return and not replacement in the field. Microsoft and Oracle conceded that without a pluggable module, a single failure has a wider impact. Whoever brings FIT rates down (failures in time, per billion device-hours) and improves testing gets closer to volume orders. Theme 2: XPO and USFP and their impact on NPO and CPO adoption Neither form factor changes the NPO or CPO picture until 400G per lane silicon is tested and proven. XPO is closer to market. It already runs at 200G per lane, and Google and Microsoft are the hyperscalers most interested in the form factor. USFP at 400G per lane will require DSPs, TIAs and drivers that suppliers do not have yet. Once these components are available, both keep pluggable optics in scale-out for another generation and could slow NPO and CPO there. Scale-up stays NPO versus CPO (see the high-density pluggables section). Theme 3: Micro-VCSEL and its impact on NPO and CPO Micro-VCSELs do not compete with NPO or CPO as they are mainly seen as a replacement for copper, and the first products shown at OCP already replace copper rather than optics. Credo’s ZeroFlap ALC replaces AECs, not pluggables or NPO sockets. Later, as scale-in comes in, micro-VCSELs have the potential to replace the links inside the tray that still use copper, since many slow lanes work better there than a few fast ones. ELSFP with micro-ring is another solution for those links, so Lumentum’s external high-power laser business (see the light sources section) should not be seen as affected either way. We see micro-VCSEL as incremental for Lumentum, which moved early in favor of the technology. We believe it is the unnamed second emitter on Credo’s SRMO slide. The field is more crowded than InP and its lead is smaller, so whatever share it wins is upside on top of the core business. The takeaway is a copper-replacement story that grows the optical market and leaves existing optics untouched. Theme 4: Beyond silicon, fiber, lasers and test are now constraints too We kept hearing the same worry: single-mode fiber will be short, whichever approach wins. Lightmatter expects the shortage in late 2027 to 2028 and says the CPO ramp is gated by test equipment in Taiwan; Astera Labs also pointed to component supply as the constraint. Silicon has not dropped off the list. DSPs, TIAs and drivers for 400G per lane are not yet available from suppliers (see the high-density pluggables section). Lumentum’s CPO capacity is booked into early 2029, as its CEO has said. 2. NVIDIA: a CPO version of NVL576 but keeps NPO on the table Keynote: a CPO version of NVL576 Most of NVIDIA’s day-one keynote went to MGX 1.2, 800VDC and the Vera CPU; optics got the last few minutes but carried the most news: Vera Rubin Ultra NVL576 links 576 GPUs over NVLink through a next-gen NVLink 7 CPO switch tray, targets 10T+ parameter models and arrives “next year.” This was probably the first time NVIDIA announced optical scale-up at such a large venue; at GTC it only came up in a small workshop. Image 2: Vera Rubin Ultra NVL576, with 576 GPUs linked through NVLink 7 CPO switch trays. Source: FUNDA; NVIDIA slide. One thing the keynote made clear is that NVIDIA’s next scale-up domain will be optical. NVL576 spans eight racks and 576 GPUs, far beyond the reach of in-rack copper, and NVIDIA is positioning it for 10T+ parameter models, so cross-rack NVLink has to run over fiber. Where the optical engine sits is still open. The keynote showed a CPO version, with optics soldered onto the NVLink 7 switch tray and the slide labeled “Vera Rubin Ultra NVL576 CPO.” In the NPO session, however, NVIDIA’s own slides said it may build NPO with ODMs and will support NPO through NVLink Fusion, and Kuchta mentioned a soldered-NPO proof of concept. Our earlier channel checks pointed to NPO for NVL576’s cross-rack links. We do not read the keynote as a final decision, only as a sign that CPO remains firmly in contention. We expect the choice to depend on yield and test capacity for co-packaged engines, how failures are handled in the field, and laser supply. Either way, demand for cross-rack optics, fiber and lasers is coming; what remains unsettled is how that value splits between the switch tray and the compute side, and between CPO and NPO engine suppliers. Supply-side conversations leaned toward NPO. The scale-up systems people described for 2027 are NPO. They put CPO’s volume in scale-out instead, in NVIDIA’s Spectrum-X CPO switches, where our checks indicate it is still a small share of scale-out switching. They saw CPO inside scale-up as 2029 at the earliest. That is later than the “next year” on NVIDIA’s slide and later than Coherent’s 27H2 target for scale-up CPO production (see the Coherent company note). We take the gap as one more reason to treat engine placement as undecided. On scale-out, NVIDIA presented CPO-based Spectrum-X Photonics multiplane Ethernet, scaling to 512K GPUs on two tiers, with 2.6 ms hardware fault detection, 40% fewer switches, 5x power efficiency and 10x MTBI; Meta, Microsoft and Oracle are listed adopters. NPO session: “Socketed does not mean field replaceable” The NPO session on the 13th had ten speakers, and NVIDIA’s Daniel Kuchta, Principal Hardware System Architect, went last, opening with the line that a few things about NPO “need to be cleaned up.” It was the most direct critique of NPO we heard all week, in three points: • First , based on NVIDIA’s experience, soldered beats sockets. Soldered and socketed describe how the optical engines are attached and either approach can be used with CPO or NPO. NPO’s claimed replaceability advantage comes with its socketed design. Sockets help in prototyping and manufacturing defect screening, but add cost and area, hurt signal integrity and complicate thermals; soldering is cheapest with the best signal integrity. • Daniel Kuchta said the Spectrum-X CPO switch is in production with all 32 optical engines soldered to the package. NVIDIA’s earlier Quantum InfiniBand CPO switch was socketed, hence “been there, done that.” Image 3: NVIDIA’s socketed-vs-soldered CPO comparison: “CPO socketed does NOT mean field replaceable... same for NPO”. Source: FUNDA; NVIDIA slide. • Second , serviceability is “dubious at best.” Even socketed NPO will likely be clean-room assembled with woven pigtails, so it gets shipped back rather than fixed in the field, “a very expensive repair job.” • Third , multi-sourcing is doubtful too. NPO for density or power he understands; for multiple suppliers, “you’re kidding yourself,” since every prototype differs and there are four MSAs, all brand new. NVIDIA did not shut the door, though: the slide below says it may work with ODMs on NPO and will support NPO through NVLink Fusion, and that CPO, pluggables and NPO will likely coexist in scale-out; Kuchta also has a soldered-NPO proof of concept. Image 4: NVIDIA on NPO: it may work with ODMs on NPO and will support NPO via NVLink Fusion, but questions the multi-supplier and serviceability claims. Source: FUNDA; NVIDIA slide. Pluggables stay too. According to Daniel Kuchta, Octal Small Form Factor Pluggable (OSFP) will still be the workhorse. NVIDIA’s announcement of its collaboration with Cisco on Ultra-Dense Small Form Factor Pluggables (USFP) is discussed in more detail in the high-density pluggables section below, together with eXtra-dense Pluggable Optics (XPO) and their collective impact on CPO and NPO adoption. 3. NPO: How everyone else responded Subscribe now Read more
View original →
MRVLMarvellBullishMarvell
OCSOculis HoldingBullishOculis Holding
ASMLNeutral
CRWVCoreWeave, Inc.NeutralCoreWeave, Inc.
GOOGLAlphabetNeutralAlphabet
METANeutral
NVDANVIDIANeutralNVIDIA
OPOptimismNeutralOptimism
TSMTSMCNeutralTSMCWeekly|AI4S, Gemini 4, TSM & ASML, MRVL and OCS, Muse, CRWV
Google is back in the frontier race, and we think a new training cycle is beginning. As we argue in this week’s column, Gemini 4 is evidence that scaling still works; the next test is how much post-training improves coding and long tasks in Gemini 4.1. With OpenAI, Anthropic, Google, Meta and xAI all investing to defend or improve their positions, we remain bullish on the networking and optical interconnects needed to keep larger clusters productive. AI for science gives us another reason to be optimistic about where that compute can go. OpenAI’s latest mathematics release still needs independent scrutiny, but progress in problems with verifiable answers could help AI take on more research and engineering work. Drug discovery, materials and energy research offer much broader potential benefits. Physical experiments and clinical trials will still take time; our optimism is about expanding the ideas scientists can test and the connections they can make across disciplines. The supply chain is preparing for that investment. We see TSMC’s growth increasingly governed by how quickly capacity comes online, and have raised our ASML EUV shipment forecasts again. Marvell’s Investor Day targets are moving closer to our forecasts, backed by longer-term supplier planning. OCS adds another source of optical demand as content per accelerator rises, although deployment beyond Google remains an important assumption in our longer-term model. On the application side, we are watching retention and cost together. Muse’s improving retention gives Meta a reason to expand acquisition once the economics support it, while computer use opens a much larger pool of work to automation. Our expert interviews also underline the obstacles: high token consumption can limit enterprise adoption, and physical capacity still depends on manufacturing yields, equipment and power. Better models and more infrastructure need to translate into reliable, affordable work. This Week’s Reports Going the Way of Go: stronger models make the next training cycle worth watching. Our new column connects AI’s progress in mathematics and programming with Google’s return to the frontier, and explains why we remain optimistic about AI for science and optical interconnect demand. TSM & ASML: capacity expansion sets the pace. We expect TSMC’s advanced-node expansion to support further growth and now forecast ASML total EUV shipments of 100/125 units in 2027/28, as both logic and memory customers need more equipment. MRVL & OCS: more optical content as AI clusters grow. Marvell’s new targets move closer to our forecasts, while our OCS work points to more than $20b of annual demand by 2030, conditional on broader adoption that includes Nvidia from late 2028. META: Muse’s retention is improving before the next growth push. We expect adoption to advance in steps as Meta improves retention and balances user acquisition against the substantial compute cost of serving a personal agent. Computer use: a much larger market for AI providers. Our long-term framework points to roughly $1tn of additional annual revenue and automatable work equivalent to about 180mn full-time roles, an estimate of task hours rather than a forecast of layoffs. Premium Report Snapshot Subscribe now Read more
View original →
MRVLMarvellNeutralMarvellResearch|MRVL: Investor Day Targets Move Closer to Our Forecasts
Optical interconnects and custom silicon driving growth Optical interconnects and custom silicon driving growth
Marvell outlined an ambitious outlook at its October 6 investor day, targeting CY30 (FY31) revenue of $80b at the midpoint and non-GAAP EPS of more than $30.
View original →