Marvell Forward Catalyst Calendar — Q2 FY27 through FY29+
This calendar extends 05_financials/q1_fy27_preview.md — the T-30-day preview of the next earnings call — into a multi-quarter forward register. Catalysts are organized into four windows: Near-term (Q2 FY27, May–Jul 2026), Mid-term (Q3-Q4 FY27, Aug 2026–Feb 2027), FY28 (~Mar 2027–Feb 2028), and FY29+ (long-term). For each catalyst, the entry includes: date / window, event description, expected impact (high / medium / low), what would surprise positively, what would surprise negatively, and the source / detection signal.
The calendar reflects Marvell’s stated forward roadmap as of 2026-04-29, the post-Q4 FY26 narrative architecture (FY27 ~$11B / FY28 ~$15B / FY28 EPS “well over $5” — Murphy Q4 FY26 ✓ Motley Fool 2026-03-05), and the post-Polariton acquisition information vacuum (no 8-K filed; first disclosure window Q1 FY27 10-Q in late August 2026). It is the framework analysts will use to determine whether the bull or bear thesis materializes.
Near-term window — Q2 FY27 (May–July 2026)
The near-term window is dominated by the Q1 FY27 earnings call — the first reporting period since three M&A closes (Celestial Feb 2, XConn Feb 10, Polariton announced Apr 22), the $1B senior-notes raise (Apr 6/15), the NVIDIA $2B investment announcement (March 31 2026), and a wave of analyst rating changes (Barclays Apr 9 upgrade with PT $156; Benchmark Apr 21 trim; Seeking Alpha Apr 27 downgrade to Hold).
Q1 FY27 earnings call — target 2026-05-29
| Field | Detail |
|---|
| Date / window | ~2026-05-29 (target; Q1 FY26 reported 2025-05-29 — Marvell’s Thursday-after-quarter-end pattern). Quarter-end ≈ 2026-05-02. |
| Impact | HIGH |
| Positive surprise | Q1 revenue beats midpoint by ≥$30M (≥$2.43B); GM 59.5-60.5%; Murphy reaffirms “approximately $15B” / “well over $5” / “double in fiscal 2028” verbatim; first public Polariton color disclosed including expected close + integration roadmap; Celestial first design-win narrative referenced (unnamed hyperscaler “validation engagement”); buyback pace sustained $300-400M Q1; OpEx step <$50M QoQ. Stock +5-10% next-day; Barclays $156 PT validated; consensus PT migrates from current $129 toward $140-145. |
| Negative surprise | Q1 revenue misses (<$2.35B); Q2 guide below $2.55B implied; GM <58.5%; “approximately $15B” softened to “in the range of $14-$15B” or removal of explicit dollar peg; “double in FY28” custom-silicon language hedged; OpEx step >$100M QoQ. Stock -8-15% next-day; Stifel customer-cliff thesis validated; second wave of downgrades; PT compresses to $105-$115. |
| Source / detection | 05_financials/q1_fy27_preview.md full preview (T-30 days); transcript will publish at Motley Fool / Investing.com same day. ✓ |
COMPUTEX 2026 keynote — Matt Murphy speaking June 2 2026
| Field | Detail |
|---|
| Date / window | 2026-06-02 (Taipei; per 02_technology/conference_calendar.md, Murphy COMPUTEX keynote scheduled). |
| Impact | MEDIUM-HIGH |
| Positive surprise | Murphy uses keynote to formally articulate the end-to-end optical AI stack narrative (Polariton modulation + Inphi DSP + Celestial Photonic Fabric + XConn switching + custom xPU) with hyperscaler customer testimonials, named TSMC partnership extensions, NVIDIA NVLink Fusion design demonstration. First public Polariton + Celestial integrated demo. |
| Negative surprise | Generic AI-infrastructure messaging without customer specificity; no Polariton roadmap detail; no Celestial design-win color; lower than-expected media pickup. |
| Source / detection | Marvell IR press releases (investor.marvell.com); COMPUTEX official program; Matt Murphy speaking-engagement disclosures. ◐ |
Q1 FY27 10-Q filing (~late August 2026; Polariton disclosure window)
| Field | Detail |
|---|
| Date / window | ~2026-08-26 (typical Marvell 10-Q cadence: 5-7 weeks after Q1 close ~2026-05-02; consistent with Q1 FY26 filed 2025-06-26 ◐). |
| Impact | HIGH (this is the first public window for Polariton purchase price + structure) |
| Positive surprise | Subsequent Events footnote discloses Polariton terms <$300M cash with no earnout overhang and minimal Q2 FY27 OpEx step; clean integration timeline; ETH licensing transferred without dispute; revenue contribution H2 2027 / FY28. |
| Negative surprise | Polariton disclosed at >$500M with significant earnout / contingent consideration; integration costs >$30M/quarter (vs. expected $20M); ETH license restructuring complicates IP transfer; goodwill allocation issues raised by auditor. |
| Source / detection | EDGAR filing under CIK 0001835632; 10-Q Note “Subsequent Events” or “Acquisitions”; precedent: XConn terms disclosed only in 10-K Note 16 (03_ecosystem/company_profiles/xconn.md ✓). |
Mid-term window — Q3-Q4 FY27 (August 2026–February 2027)
This window covers two earnings calls (Q2 FY27, Q3 FY27) and the conference cadence that bracket them. The strategic question for the mid-term is whether the FY27 ramp tracks to “approaching $11B” or shows early signs of slippage.
Q2 FY27 earnings call — ~2026-08-28
| Field | Detail |
|---|
| Date / window | ~2026-08-28 (target; Q2 FY26 reported 2025-08-28). |
| Impact | HIGH |
| Positive surprise | Q2 revenue ≥$2.55B; FY27 full-year guide tightened upward (e.g., $11.0B → $11.2-11.5B); custom AI silicon revenue cadence accelerating quarter-over-quarter; Polariton design-in disclosure with specific hyperscaler customer or TSMC qualification milestone; Celestial first revenue contribution flagged for early FY28. |
| Negative surprise | Q2 revenue misses; FY27 guide cut to $10.5-10.8B; custom AI silicon flat sequentially; Polariton integration commentary citing yield issues; Celestial OpEx step exceeding $50M without revenue offset visibility. |
| Source / detection | Q2 FY27 transcript via Motley Fool / Investing.com; press release on Marvell IR. ✓ |
Hot Chips 2026 — August 25-26 2026
| Field | Detail |
|---|
| Date / window | 2026-08-25 to 2026-08-26 (Stanford / Cupertino; annual processor architecture conference). |
| Impact | MEDIUM |
| Positive surprise | Marvell-presented or Marvell-mentioned papers on custom xPU architecture (with hyperscaler co-author byline, indicating customer endorsement); TSMC SoIC + CoWoS-L disclosure for Celestial Photonic Fabric integration; Polariton POH modulator architecture paper. |
| Negative surprise | Alchip-led presentation indicating second-source ASIC win on a hyperscaler program; Marvell absence from custom-silicon panels; competing CPO architectures (AVGO Tomahawk-Bailly, NVIDIA in-house) with stronger demos. |
| Source / detection | Hot Chips program (www.hotchips.org); SemiAnalysis coverage; primary papers on hyperscaler architecture decisions. |
ECOC 2026 — September 21-25 2026
| Field | Detail |
|---|
| Date / window | 2026-09-21 to 2026-09-25 (European Conference on Optical Communications, Brussels likely). |
| Impact | MEDIUM-HIGH |
| Positive surprise | Polariton + Marvell joint paper or demo on first 3.2T POH modulator on TSMC silicon photonics; LWLG-Polariton-Marvell joint demo (would resolve the IP2 supply-chain dependency in risks.md by signaling continued material-supply alignment); 1.6T → 3.2T DSP transition demo. |
| Negative surprise | Polariton presentation reverts to MPW-shuttle prototype (signaling TSMC qualification slippage); no joint Marvell + Polariton paper; competing demos from HyperLight (TFLN), Ayar Labs, NVIDIA at higher visibility. |
| Source / detection | ECOC 2026 program; OSA / IEEE Photonics conference proceedings; cross-link with LWLG ECOC presentations historically per lwlg/kb/_user_inputs/polariton_lwlg_verification.md joint-PR cadence. ✓ |
Q3 FY27 earnings call — ~2026-12-01
| Field | Detail |
|---|
| Date / window | ~2026-12-01 (target; Q3 FY26 reported 2025-12-02). Concurrent with major hyperscaler events (AWS re:Invent ~early Dec). |
| Impact | HIGH |
| Positive surprise | Q3 revenue strongly above guide; FY27 full-year tracking to $11.5B+; custom AI silicon revenue accelerating; first material Celestial revenue disclosure; AWS re:Invent 2026 names Trainium 3 production milestones favorable to Marvell; Polariton + COLORZ 3200 timing reaffirmed. |
| Negative surprise | Q3 revenue in-line or miss; FY27 guide trimmed; AWS re:Invent 2026 hints at second-source ASIC vendor (Stifel customer-cliff thesis validation); Celestial design-win window slips to FY28 H2. |
| Source / detection | Q3 FY27 transcript; AWS re:Invent 2026 keynotes; Marvell IR. ✓ |
Industry Analyst Day 2026 — TBD (window: late Q2 FY27 / Q3 FY27)
| Field | Detail |
|---|
| Date / window | TBD as of 2026-04-29; historic April cadence broken; Q1 FY27 call (May 29) is the most likely announcement window per 05_financials/q1_fy27_preview.md §7. Most likely date range: September-November 2026. |
| Impact | HIGH (the IAD is the natural disclosure forum for the disaggregated FY28 revenue bridge that “approximately $15B” requires) |
| Positive surprise | Disaggregated FY28 revenue bridge: how much is custom-silicon, optical, switching, Celestial Photonic Fabric ARR, XConn UALink switch, Polariton-enabled COLORZ 3200; customer-by-customer (anonymized) custom-silicon revenue contribution; multi-year R&D allocation; capex allocation (TSMC SoIC + CoWoS + advanced-node) for FY28 ramp; long-term GM / OpM target post-Celestial integration; Polariton purchase-price and earnout structure. |
| Negative surprise | IAD postponed beyond CY2026; opaque qualitative framing without numerical disaggregation; FY28 revenue bridge components not detailed. |
| Source / detection | Marvell IR announcement (typical 4-6 weeks lead time); precedent: April 2024 AI-Day-style event introduced the $1.5B AI-revenue framing. Cross-link _user_inputs/earnings_call_language_tracker.md. |
Q4 FY27 / FY27 full-year earnings call — ~2027-03-04
| Field | Detail |
|---|
| Date / window | ~2027-03-04 (target; Q4 FY26 reported 2026-03-05). FY27 full-year results disclosed; first FY28 quarterly guide. |
| Impact | VERY HIGH (FY27 closes the +30% growth year; FY28 guide formalizes the ~$15B framing) |
| Positive surprise | FY27 revenue prints ≥$11B (vs. “approaching $11B” guide); FY28 guide >$15B with explicit dollar peg; FY28 EPS guide “$5.50-$6.00” (above “well over $5”); Celestial first quarterly revenue contribution announced; 10-K Note 16 discloses Polariton terms below $400M with strong integration metrics. |
| Negative surprise | FY27 revenue prints below $11B; FY28 guide “in the range of $14-$15B”; FY28 EPS guide “above $5”; Celestial revenue contribution pushed to mid-FY28; goodwill review on Celestial / Polariton flagged. |
| Source / detection | Q4 FY27 transcript; Marvell IR; subsequent FY27 10-K filing (~2027-03-12). |
FY28 window — March 2027 to February 2028
The FY28 window is where the bull thesis is most directly tested: $15B revenue, “well over $5” EPS, Celestial $500M ARR by Q4 FY28, Polariton-enabled COLORZ 3200 in production, and full integration of all three FY26 acquisitions.
OFC 2027 — March 8-12 2027
| Field | Detail |
|---|
| Date / window | 2027-03-08 to 2027-03-12 (Optical Fiber Communication Conference; San Diego / San Francisco rotating). |
| Impact | HIGH |
| Positive surprise | Marvell + Polariton present first commercial COLORZ 3200 module with TSMC SiPh fabrication, full LWLG-Perkinamine or alternative material qualification disclosed; named hyperscaler customer endorsement; LightCounting reports 1.6T optical module shipments doubled in CY26. |
| Negative surprise | COLORZ 3200 demo absent or reverted to MPW-shuttle prototype; AVGO Bailly CPO competing demo with stronger differentiation; LightCounting commentary on optical TAM moderation. |
| Source / detection | OFC 2027 program; OSA / IEEE Photonics proceedings; LightCounting market reports; Marvell + Polariton joint press releases. |
Quarterly cadence FY28 (Q1-Q4)
Each quarter brackets the FY28 ramp. Detection signals to monitor:
- Q1 FY28 (~May 2027): Custom xPU revenue tracking +25%+ YoY (vs. +20% floor guide); Celestial first material revenue disclosure; Polariton TSMC qualification status.
- Q2 FY28 (~Aug 2027): Polariton-derived COLORZ 3200 design-in announced at hyperscaler customer; Celestial Photonic Fabric design-win update.
- Q3 FY28 (~Dec 2027): Custom xPU on track to “double” framing; FY28 full-year revenue tracking to $15B; FY29 guide framework introduced.
- Q4 FY28 (~Mar 2028): FY28 closes; Celestial $500M ARR target test; FY29 framework first quantitative disclosure.
FY28 ~$15B framing checkpoint (Q4 FY28 results, ~2028-03-02)
| Field | Detail |
|---|
| Impact | VERY HIGH (the load-bearing FY28 framing is verified or falsified) |
| Positive surprise | FY28 revenue ≥$15B; non-GAAP EPS ≥$5.50; Celestial Photonic Fabric ARR ≥$500M Q4 FY28; FY29 guide $18-19B with continued double-digit custom xPU growth. |
| Negative surprise | FY28 revenue $13-14B; non-GAAP EPS $4.50-$5.00; Celestial ARR <$300M Q4 FY28; FY29 guide softer than expected. |
| Source / detection | Q4 FY28 / FY28 full-year earnings; cross-validate with hyperscaler 10-K filings re custom-silicon disclosures. |
FY29+ window — long-term (March 2028 onward)
The FY29+ window addresses two structural questions: (a) whether the optical roadmap progresses to 6.4T (post-3.2T), and (b) whether the AI capex cycle inflects from generation-1 custom-silicon deployment to a generation-2 refresh cycle.
6.4T optical roadmap visibility — FY28 H2 / FY29
| Field | Detail |
|---|
| Date / window | OFC 2028 (March 2028) and ECOC 2028 (September 2028) are the natural disclosure forums for 6.4T DSP / POH demos. |
| Impact | HIGH |
| Positive surprise | Marvell + Polariton present 6.4T POH modulator paper; TSMC SiPh roadmap aligned for 6.4T volume CY29; LWLG / NLM material supply chain validated for 6.4T-grade thermal stability; competing platforms (AVGO Bailly, NVIDIA CPO) trail in bandwidth or yield. |
| Negative surprise | 6.4T transition timeline slips to FY30; competing technologies (TFLN, Indium Phosphide) close the bandwidth gap; Marvell’s POH advantage erodes vs. CPO economics. |
| Source / detection | OFC / ECOC programs; LightCounting roadmap reports; SemiAnalysis market commentary. |
Custom xPU generation-2 cycle inflection — FY29-FY30
| Field | Detail |
|---|
| Date / window | FY29 hyperscaler refresh window (CY28-CY29 timeframe); next-generation Trainium / Maia / MTIA / TPU programs need to be in design by FY28 H2 to ship by FY30. |
| Impact | VERY HIGH |
| Positive surprise | Marvell wins design-in on Trainium 4, Maia 300, MTIA next-generation, TPU 9; multi-generational continuity intact; “fab-light” model scales without capex acceleration. |
| Negative surprise | One or more hyperscalers names alternate ASIC vendor (Alchip, AVGO incumbent on Google) for next-generation; Marvell loses multi-generational continuity narrative. |
| Source / detection | Hyperscaler events: AWS re:Invent (Dec annual), Microsoft Build (May annual), Google Cloud Next (Apr annual), Meta Connect (Sep annual). |
NVIDIA CPO market entry — FY29
| Field | Detail |
|---|
| Date / window | NVIDIA’s in-house silicon photonics / CPO program (Risk X5 in risks.md) targets Blackwell Ultra, Rubin, Vera Rubin platforms in CY27-CY29. Production volumes by FY29. |
| Impact | HIGH (sets the structural ceiling on Marvell’s optical interconnect TAM) |
| Positive surprise | NVIDIA CPO deployment limited to internal NVLink fabrics; Marvell merchant DSP volume continues at hyperscalers; NVLink Fusion ecosystem expands rather than constrains Marvell. |
| Negative surprise | NVIDIA CPO replaces merchant pluggable + DSP at major hyperscalers; Marvell merchant DSP volumes plateau or decline FY29; the optical-interconnect TAM splits between NVIDIA CPO (in-NVIDIA-system) and merchant (everyone else). |
| Source / detection | NVIDIA GTC (March annual); SC supercomputing conference (November annual); NVIDIA earnings + 10-K disclosures on optical IP. |
AI capex cycle inflection — FY29-FY30
| Field | Detail |
|---|
| Date / window | The cycle inflection is the FY30+ hyperscaler capex commitment posture. AWS, Microsoft, Google, Meta combined CY29-CY30 capex guides will indicate whether AI infrastructure spend is accelerating, plateauing, or declining. |
| Impact | VERY HIGH (this is the macro variable underlying the entire bull thesis) |
| Positive surprise | Hyperscaler capex guides accelerate FY29-FY30 (driven by inference workload growth, AGI training scale-out, sovereign AI deployments); custom-silicon ramp continues at multi-billion lifetime revenue per socket. |
| Negative surprise | Hyperscaler capex guides flat or decline; AI revenue-per-dollar-of-capex commentary turns cautious; “peak hyperscaler” narrative materializes. |
| Source / detection | Hyperscaler Q4 CY28 calls (Jan-Feb 2029); industry capex aggregators (Bloomberg, Visible Alpha); Macquarie, JPMorgan, Bernstein semiconductor strategy reports. |
Cross-references