Marvell Technology, Inc. · NASDAQ:MRVL

Marvell Technology — Investor Day (October 2026)

161 slides from Marvell Technology's investor day presentation, Oct 6, 2026

Slide 1: Marvell Investor Day October 6, 2026
[=| MARVELL’ INVESTOR DAY OCTOBER 6, 2026
Slide 2: Ross Seymore
Ross Seymore SVP, Investor Relations
Slide 3: Agenda
Agenda Matt Murphy Chairman and CEO Chris Koopmans President and COO Sandeep Bharathi President, Data Center Group Break Xi Wang SVP and GM, Connectivity Business Russ Esmacher SVP and GM, Data Center Interconnect Business Dave Lazovsky EVP and GM, Data Center Networking Business Will Chu EVP and GM
Slide 4: Basis of presentation
Basis of presentation Fiscal Year vs. Calendar Year ▪ Financial measures in this presentation correspond to Marvell’s fiscal calendar ▪ Marvell’s fiscal year is the 52- or 53-week period ending on the Saturday closest to January 31 ▪ As an example, FY26 refers to the period February 2, 2025 through
Slide 5: Forward-looking statements
Forward-looking statements Except for statements of historical fact, this presentation contains forward-looking statements (within the meaning of the federal securities laws) including, but not limited to, statements related to market trends and to the company’s business and operations, business o
Slide 6: Non-GAAP financial measures
Non-GAAP financial measures ▪ During this presentation, we will refer to certain financial measures on a U.S. Non-GAAP basis. ▪ We believe that the presentation of Non-GAAP financial measures provides important supplemental information to management and investors regarding financial and business tre
Slide 7: Matt Murphy
[=| MARVELL’ Matt Murphy Chairman and CEO
Slide 15: Architecting the Marvell platform
Architecting the Marvell platform Build organically Strategic M&A Integrate + invest Invest in Marvell’s existing technology foundation + Expand the platform through targeted M&A Integrate the platform around one strategic direction One integrated platform, built and scaled inside Marvell ~$40B ofan
Slide 16: AI workloads are reshaping the network
AI workloads are reshaping the network Reasoning models Agentic AI More data Greater movement connectivity Mixture of experts More diverse demands workloads Inference at scale Moving data efficiently is becoming a fundamental constraint on AI system performance © 2026 Marvell. All rights reserved. 1
Slide 17: Hyperscalers rearchitecting their networks
Hyperscalers rearchitecting their networks AWS Google Cloud Meta Microsoft “ The way networks “ In the era of reasoning “ Whether these clusters “ As we enter the era had been built wasn't models and MoE, are training the next of frontier-scale going to scale into where any chip may frontier model o
Slide 18: Connectivity is the next defining unlock
Connectivity is the next defining unlock Compute-first Memory-first Revenue split* Revenue split* Connectivity Memory Compute Company A Company B Company C Company A Company B Company C *Approximate revenue split © 2026 Marvell. All rights reserved. 18
Slide 19: Connectivity is the next defining unlock
Connectivity is the next defining unlock Connectivity-first Revenue split* Compute-first Memory-first Revenue split* Revenue split* Connectivity Connectivity Memory Compute Compute Company A Company B Company C Company A Company B Company C Marvell *Approximate revenue split © 2026 Marvell. All righ
Slide 20: Marvell is a different kind of AI company
Marvell is a different kind of AI company FY26 total revenue Connectivity-first $8.2B Comms Revenue split* and other Interconnect Connectivity 30% 70% compute connectivity Switch Compute Custom and storage Marvell *Approximate revenue split © 2026 Marvell. All rights reserved. 20
Slide 21: Marvell connects AI infrastructure across every distance
Marvell connects AI infrastructure Interconnect Switch and storage across every distance Custom Distributed DCs Data center Rack | Multi-rack Package Up to 1,000km Up to 2km Up to 50m Up to 100mm Scale-across Scale-out Scale-up Scale-in Fiber optics Copper © 2026 Marvell. All rights reserved. 21
Slide 22: Marvell connects AI infrastructure across every distance
Marvell connects AI infrastructure Interconnect Switch and storage across every distance Custom Distributed DCs Data center Rack | Multi-rack Package Up to 1,000km Up to 2km Up to 50m Up to 100mm Scale-across Scale-out Scale-up Scale-in Fiber optics Copper © 2026 Marvell. All rights reserved. 22
Slide 23: Marvell connects AI infrastructure across every distance
Marvell connects AI infrastructure Interconnect Switch and storage across every distance Custom Distributed DCs Data center Rack | Multi-rack Package Up to 1,000km Up to 2km Up to 50m Up to 100mm Scale-across Scale-out Scale-up Scale-in Fiber optics Copper © 2026 Marvell. All rights reserved. 23
Slide 24: One switching foundation for scale-out and up
One switching foundation for scale-out and up Interconnect Switch and storage Custom Scale-out switch Scale-up switch Proven at hyperscale New opportunity ▪ Connecting AI clusters across pods ▪ Connecting accelerators within pods ▪ High-radix ethernet at cloud scale ▪ Supporting every major protocol
Slide 25: Unmatched leadership in storage and memory
Unmatched leadership in storage and memory Interconnect Switch and storage Custom Merchant / standard products Custom solutions Dense SRAM SRAM technology Custom DRAM / Proven IP memory expansion Memory controllers/CXL AI infrastructure + decades Merchant technologies Flash NAND controllers of exper
Slide 26: Dramatically expanding custom opportunity
Dramatically expanding custom opportunity Interconnect Switch and storage Custom NIC Storage Multiple strategic agreements Custom silicon now shipping Memory Security expansion to the top four hyperscalers XPU attach is semi-custom Near-memory Inference Management compute accelerator co-processor ©
Slide 28: Custom revenue is accelerating
Custom revenue is accelerating Updated FY29 custom target FY29 custom target $10B $12B+ >2x >3x FY29 vs. FY28 FY29 vs. FY28 © 2026 Marvell. All rights reserved. 28
Slide 29: Data center CAPEX drives long-term growth
Data center CAPEX drives long-term growth Marvell data center Data center Accelerated compute TAM CAPEX TAM CY30 CY30 CY30 ~$3T ~$1.6T ~$385B ~35% CAGR ~48% CAGR CY25-CY30 CY25-CY30 ~50% CAGR CY25-CY30 Source: Industry estimates, public information © 2026 Marvell. All rights reserved. 29
Slide 30: Data center CAPEX drives long-term growth
Data center CAPEX drives long-term growth Marvell total Marvell data center Comms and other TAM TAM TAM CY30 CY30 CY30 ~$385B $15B $400B ~50% CAGR ~3% CAGR CY25-CY30 CY25-CY30 ~45% CAGR CY25-CY30 Source: Industry estimates, public information © 2026 Marvell. All rights reserved. 30
Slide 31: Growth across the stack
Growth across the stack TAM CAGR Industry growth drivers TAM CY25 – CY30 CY30 Growth of scale-out Interconnect ~65% ~$65B New scale-up and across Switching Growth of scale-out ~40% ~$85B and storage New scale-up Increase of XPU Custom ~55% ~$235B New XPU attach Comms ~3% Modest growth ~$15B and othe
Slide 33: Our growth continues to accelerate
Our growth continues to accelerate Total revenue $20B $12B 67% YoY $20B 47% $8.2B YoY FY28 updated revenue outlook FY26 FY27E FY28E © 2026 Marvell. All rights reserved. 33
Slide 36: Today | Data center target pulled in one year early
Today | Data center target pulled in one year early Investor Day 2026 Data center revenue $20B ~$18B Total revenue Data center $2.2B FY24 FY28E FY29E Updated outlook © 2026 Marvell. All rights reserved. 36
Slide 37: Today | Data center target pulled in one year early
Today | Data center target pulled in one year early Investor Day 2026 Data center revenue ~$18B Data center ~70% DC CAGR $2.2B FY24 FY28E Updated outlook © 2026 Marvell. All rights reserved. 37
Slide 38: Setting our long-term outlook
Setting our long-term outlook $70-90B $2.5B 55-70% $67.5- Total revenue $20B DC CAGR 87.5B Comms and other Data center ~$18B ~70% $8.2B DC CAGR $6.1B FY26 FY28E FY31E © 2026 Marvell. All rights reserved. 38
Slide 39: Marvell long-term revenue target
Marvell long-term revenue target $70-90B FY31 total revenue Total revenue 55-60% $8.2B FY26-FY31 revenue CAGR FY26 FY31E © 2026 Marvell. All rights reserved. 39
Slide 40: What underpins the long-term outlook
What underpins the long-term outlook TAM Revenue Marvell Implied revenue CAGR CAGR growth drivers (midpoint) CY25 – CY30 FY26 – FY31 FY31 Leadership in scale-out / across Interconnect ~65% 60-70% Growing scale-up ~$37.5B Switching Growing scale-out and storage ~40% 40-50% New scale-up opportunity ~$
Slide 41: Why invest in Marvell?
Why invest in Marvell? $400B $20B $12B+ $70-90B EPS CY30 FY28 updated FY29 updated FY31 total … Marvell total revenue custom revenue revenue TAM forecast outlook target outlook Trusted by the world’s AI infrastructure leaders © 2026 Marvell. All rights reserved. 41
Slide 42: Chris Koopmans
Chris Koopmans President and COO
Slide 45: AI buildout continues at massive scale and speed
AI buildout continues at massive scale and speed 2025 Custom AI Day Total DC CAPEX $1T 20% CAGR $593B CY25 CY28 Source: Industry estimates, public information © 2026 Marvell. All rights reserved. 45
Slide 46: AI buildout continues at massive scale and speed
AI buildout continues at massive scale and speed Investor Day 2026 $1.2T Total DC CAPEX 73% YoY $711B CY25 CY26 Update Update Source: Industry estimates, public information © 2026 Marvell. All rights reserved. 46
Slide 47: AI buildout continues at massive scale and speed
AI buildout continues at massive scale and speed $3.1T Total DC CAPEX 35% CAGR $711B CY25 CY30 Update Source: Industry estimates, public information © 2026 Marvell. All rights reserved. 47
Slide 48: Large hyperscale data centers driving majority
Large hyperscale data centers driving majority $3.1T Total DC CAPEX 35% CAGR 40% CAGR >70% $711B All others Top 10 hyperscalers >60% CY25 CY30 Update Source: Industry estimates, public information © 2026 Marvell. All rights reserved. 48
Slide 49: Focus on large customers, deep partnerships
Focus on large customers, deep partnerships Hyperscalers Marvell Own the workload, the data center, The core IP partner and the system architecture that helps realize that vision Data center operations Advanced node at massive scale silicon and packaging System development High speed analog mixed an
Slide 50: Marvell customer journey spans full platform
Marvell customer journey spans full platform Off-the-shelf Custom system, Semi-custom Full custom system standard silicon silicon silicon Merchant Custom Diverse product portfolio at every customer © 2026 Marvell. All rights reserved. 50
Slide 53: Scaling engineering to meet opportunity
Scaling engineering to meet opportunity Non-GAAP R&D $1.5B ~4X Investing ahead of the data center opportunity Data center revenue $2.2B Data center revenue growth* FY24 *4X growth rate based on annualized FY27 Q2 data center revenue © 2026 Marvell. All rights reserved. 53
Slide 54: Building the company for the long term
Building the company for the long term Research funding Curriculum Scholarships Internships 5 countries Analog design 200+ scholarships >500 annually 14 universities Physical design 50+ universities Majority convert Design for test Marvell has become a premier destination for talent © 2026 Marvell.
Slide 55: Embracing AI to accelerate silicon engineering
Embracing AI to accelerate silicon engineering Marvell Faster ramp to impact Design Assistant Faster chip design Proprietary in-house Accelerating Accelerating design tool new engineer development product development cycle from day one to full speed from concept to production Built on decades of Mar
Slide 56: Cloud-first design engineering mindset drives scale
Cloud-first design engineering mindset drives scale 2023 2024-2025 2026 First announcement Optimized E2E flow Broad adoption ~80% EDA infrastructure in the cloud © 2026 Marvell. All rights reserved. 56
Slide 59: Scaling for growth
Scaling for growth Building Deep customer Talent Supply chain for the long term relationships pipeline relationships built on trust World-class team ready to deliver © 2026 Marvell. All rights reserved. 59
Slide 60: Sandeep Bharathi
Sandeep Bharathi President, Data Center Group
Slide 61: 2020 vision | Data infrastructure silicon provider
2020 vision | Data infrastructure silicon provider Pivot to Process Best IP on first shuttle Fast follower with leading foundry 16nm 12nm Architecture Large reticle-size chip Medium-size chips ~200mm2 Design SerDes building blocks 28G SerDes for Wi-Fi, storage for data centers 28G Source: 2020 Marve
Slide 62: 2026 | Delivered and proven at every step
2026 | Delivered and proven at every step 2020 2026 Achieved 5nm 3nm 2nm 14Å Process 5nm (High volume, taped out, IP-ready) First-to-node leadership Skipped 7nm 16nm 14Å 100% A0-to-production 12nm 2020 2026 Architecture Industry’s largest reticle-size chips T100 Optimized compute for AI >850mm2 Swit
Slide 63: AI tokenomics is a connectivity challenge
AI tokenomics is a connectivity challenge Tokenomics drivers Cost per million output token Bandwidth Latency Power Reach/Distance AI no longer Data center only a silicon connectivity problem challenge Performance Performance refers to number of tokens allocated to a given user every second © 2026 Ma
Slide 64: Memory wall stalls compute
Memory wall stalls compute Ideal world Real-world memory wall Compute Memory XPU Memory wall HBM HBM HBM Infinite memory DRAM Uniform latency High bandwidth Memory wall is a bottleneck for performance as AI infrastructure scales © 2026 Marvell. All rights reserved. 64
Slide 65: Crossing the memory wall
Crossing the memory wall Memory as shared resource Memory controllers Memory access CXL technologies Memory expander Photonic interconnects Memory fabric Power-efficient silicon Memory on a chip Rearchitecting systems to bring memory closer over scale-up fabric © 2026 Marvell. All rights reserved. 6
Slide 66: Optics enables multi-rack, scale-up domains
Optics enables multi-rack, scale-up domains Copper The limit Optical Today Tomorrow 1 rack Across racks Up to 144 XPUs 1000+ XPUs Copper limit No limit Scale-up represents >85%* of AI data center traffic Solving the scale-up bottleneck is the biggest gain *Internal company estimates © 2026 Marvell.
Slide 67: Breaking the copper wall with photonics
Breaking the copper wall with photonics 3D integrated silicon photonics light engine Electronics Photonics Substrate Integration on reverse side Pursuing every optical path, across every distance High-speed, Ultra-low power, All-optical Board-level, Short-reach, Ultra-fast, long-reach high density s
Slide 68: Common technology foundation
Common technology foundation Photonics Advanced node Analog Advanced and process mixed signal packaging Connectivity across every reach © 2026 Marvell. All rights reserved. 68
Slide 69: Going beyond CMOS
Going beyond CMOS Old world: Silicon New world: Heterogeneous integration CMOS Silicon Germanium Indium Phosphide CMOS SiGe InP Silicon Photonics Lithium Niobate Plasmonics SiPho TFLN All ICs CMOS + SiPho + SiGe + TFLN in Light Engine Co-integrated in system and in package Fully integrated, process
Slide 70: Analog mixed signal powers data movement engine
Analog mixed signal powers data movement engine Chip I/O Die-to-Die Complex System Move data Connect multiple chiplets Multiple chips and packages in / out of a chip on one package combine into full system NIC Storage Memory Security expansion Near-memory Inferenc e Managem ent compute ac celerator
Slide 71: Packaging innovation for photons and electrons
Packaging innovation for photons and electrons 2D package 2D package 2.5D package 3.5D package CPO Lid Lid Silicon Stacked Silicon Substrate Silicon Substrate chiplets Interposers silicon chiplets 1 die 2 die >8 die >8 die Full integration and co-design to scale in volume © 2026 Marvell. All rights
Slide 72: Industry-leading core technology
Industry-leading core technology Process and packaging IP and architecture System-level solutions NIC Storage Memory Security expansion Near-memory Inferenc e Managem ent compute ac celerator co-proc essor Proven execution, deployed at scale © 2026 Marvell. All rights reserved. 72
Slide 74: Xi Wang
Xi Wang SVP and GM, Connectivity Business
Slide 75: The broadest end-to-end AI connectivity portfolio
The broadest end-to-end AI connectivity portfolio Marvell is a connectivity-first company Distributed DCs Data center Rack | Multi-rack Package Up to 1,000km Up to 2km Up to 50m Up to 100mm Scale-across Scale-out Scale-up Scale-in Coherent modules PAM DSP PAM DSP LED, VCSEL Coherent DSP Coherent-l
Slide 76: Industry-first technology from the ground up
Industry-first technology from the ground up High-speed analog Advanced node DSP Photonics Broadband SiGe/GaAs/InP First to advanced node Broadest portfolio IPs around optics First to A0 production of optical technologies TIA PAM4, PAM6, PAM8 MZM, MRM, EAM, silicon photonics, Driver coherent, cohere
Slide 77: TIAs and Drivers required for every optical interconnect
TIAs and Drivers required for every optical interconnect Performance rules Lowest noise, high gain, high linearity Dozens of specialized products every year What’s inside an optical module? Pairs with wide variety of optics SerDes and Digital Analog Photonics Driver Laser Modulator DSP Expanding int
Slide 78: Marvell high-speed analog foundation
Marvell high-speed analog foundation Product leadership Technology innovation ▪ First to every generation ▪ Multi-foundry process leadership ▪ Both PAM and coherent ▪ Broadband GaAs, SiGe, and InP pluggable platforms ▪ 30-40 tape-outs per year ▪ Enabling next-gen NPO ▪ Optimized for every platform ▪
Slide 79: Multi-generational DSP platform leadership
Multi-generational DSP platform leadership 1.6T 3nm PAM DSP, 8x200G 200G/Lane TIA, Driver 800G 5nm PAM DSP, 8x100G 100G/Lane TIA, Driver 400G 7nm PAM DSP, 4x100G 100G/Lane TIA, Driver 200G 16nm PAM DSP, 4x50G DSP, TIA, Driver 50G/Lane TIA, Driver Pioneering connectivity solutions, first to every gen
Slide 80: Early co-definition with hyperscalers
Early co-definition with hyperscalers Solution defined years ahead Build-in unique features for key hyperscalers Collaboration across the optics ecosystem Production ramp ready from design phase Addressing the scaling challenges of tomorrow with customers © 2026 Marvell. All rights reserved. 80
Slide 81: Build optimized solution for diversifying AI workload
Build optimized solution for diversifying AI workload Ara Ara T Ara X Ara M Petra 1.6T 3nm PAM 1.6T 3nm PAM 1.6T 3nm PAM 1.6T 3nm PAM 1.6T 3nm PAM DSP TRO DSP DSP MACsec DSP Gearbox DSP Optimized for Optimized for Optimized for Optimized for Foundation Power Robustness Security Flexibility Shipping
Slide 82: Scaling beyond 1.6T
Scaling beyond 1.6T 3.2T Scalable for volume deployment solution Link robustness for AI Be more power efficient Maintain pluggable form factor © 2026 Marvell. All rights reserved. 82
Slide 83: Addressing the complexities of 3.2T
Addressing the complexities of 3.2T Marvell® 3.2T optical DSP solution 2nm PAM DSP 2nm Coherent-lite DSP Marvell developing both technologies for complete infrastructure needs © 2026 Marvell. All rights reserved. 83
Slide 84: Marvell industry-first 2nm live demos at ECOC
Marvell industry-first 2nm live demos at ECOC 3.2T PAM4 3.2T Coherent-lite Refreshing the definition Data center and campus reach, of bandwidth balanced power, and cost 8x400G Optical Module 2x1.6T Coherent-lite Module Using PAM 448G SerDes Path to 3.2T 1600GE Coherent-lite DSP and beyond © 2026 Mar
Slide 85: The path to 3.2T and beyond
The path to 3.2T and beyond 6.4T 3.2T 2nm 3.2T Coherent-lite 2nm 3.2T PAM 1.6T 800G 400G DSP, TIA, Driver 200G Marvell continues to lead the speed and bandwidth for every transition © 2026 Marvell. All rights reserved. 85
Slide 86: Next-gen chip-to-chip interconnect for scale-in
Next-gen chip-to-chip interconnect for scale-in Distributed DCs Data center Rack | Multi-rack Package Up to 1,000km Up to 2km Up to 50m Up to 100mm Scale-across Scale-out Scale-up Scale-in Coherent modules PAM DSP PAM DSP LED, VCSEL Coherent DSP Coherent-lite DSP ACC, NPO, CPO, VCSEL Optical I/O
Slide 87: Extending optics native to compute for scale-in
Extending optics native to compute for scale-in Today: Copper I/O Future: Optical I/O Electrical Optical ASIC ASIC I/O SerDes I/O chiplet I/O by CMOS SerDes New light sources, μLED, μVCSEL Reach: ~10mm Reach: 1.5m+ Enabling scale-in architecture with optical chiplet integration © 2026 Marvell. All r
Slide 88: Marvell is a connectivity-first company
Marvell is a connectivity-first company Most advanced Deep customer Architected to scale technology engagements for AI volume Leading every hop, every distance, every generation © 2026 Marvell. All rights reserved. 88
Slide 89: Russ Esmacher
Russ Esmacher SVP and GM, Data Center Interconnect Business
Slide 90: Marvell connects AI infrastructure across every distance
Marvell connects AI infrastructure across every distance Distributed DCs Data center Rack | Multi-rack Package Up to 1,000km Up to 2km Up to 50m Up to 100mm Scale-across Scale-out Scale-up Scale-in Leadership across the full connectivity stack © 2026 Marvell. All rights reserved. 90
Slide 91: DCI market tailwinds
DCI market tailwinds Power Data center Line system densification density Scale-across efficiency Emerging AI cluster to AI cluster Up to 1,000km Transition from Space and power More fiber boxes to pluggables constraints drives capacity 15x bandwidth more data centers vs. traditional DCI Expanding AI
Slide 92: Marvell led the pluggable transition a decade ago
Marvell led the pluggable transition a decade ago Previous Marvell coherent pluggables Transport box DCI box COLORZ® 100 COLORZ® 400 COLORZ® 800 COLORZ® 1600 2016 2020 2024 2026 First at every generation, and the pace is accelerating © 2026 Marvell. All rights reserved. 92
Slide 93: Scale-across connects AI clusters
Scale-across connects AI clusters AI cluster AI cluster Up to 1,000km 15x Bandwidth vs. DCI Accelerating the adoption of 1.6T and 3.2T © 2026 Marvell. All rights reserved. 93
Slide 94: Two paths to market
Two paths to market Coherent modules Coherent DSP Enabling hyperscalers with Enabling broader market with Marvell® COLORZ® modules Marvell DSPs powering partner modules Marvell fulfills accelerating demand with modules and DSPs © 2026 Marvell. All rights reserved. 94
Slide 95: Scaling coherent optics for high volume
Scaling coherent optics for high volume Anatomy of a Marvell coherent optics module ASIC Electronics Photonics DAC Driver Mo dulator Coherent Laser DSP Optical ADC TIA PDs frontend Marvell core competencies developed over 20+ years TIA DRIVER Largest coherent Advanced Highest speed Deep photonics Hi
Slide 96: Multi-generational scale-across platform leadership
Multi-generational scale-across platform leadership 3.2T Coherent-lite 14Å ZR 14Å Plasmonics ZR+ 14Å Plasmonics 1.6T Coherent-lite 2nm ZR 2nm TFLN ZR+ 2nm TFLN 800G 5nm ZR ZR+/MACsec 2nm SiPho SiPho 100G ZR 400G ZR 28nm 7nm SiPho SiPho Multiple DSPs, CMOS node leader, more advanced photonics © 2026
Slide 97: Marvell leading next gen at 2nm
Marvell leading next gen at 2nm COLORZ 1600 COLORZ 800 Powered by Electra DSP Powered by Libra DSP Industry-first 2nm 1.6T ZR/ZR+ Industry-first 2nm 800G ZR/ZR+ extending links to 1,000km+ with built-in MACsec encryption © 2026 Marvell. All rights reserved. 97
Slide 98: Marvell industry-first 2nm live demos at ECOC
Marvell industry-first 2nm live demos at ECOC 1.6T ZR/ZR+ 800ZR/ZR+ with MACsec More bandwidth, Secure AI traffic, lower power, longer reach lower cost, simpler operations 2nm 1.6T Coherent DSP 800G ZR/ZR+ Pluggable Optics for ZR/ZR+ Pluggable Optics with MACSec © 2026 Marvell. All rights reserved.
Slide 99: Leading scale-across with coherent pluggables
Leading scale-across with coherent pluggables Market leader First to market Delivering at scale Supplying to Leveraging core >1M shipped top 5 hyperscalers technologies Leadership through delivering innovation at scale © 2026 Marvell. All rights reserved. 99
Slide 100: Dave Lazovsky
Dave Lazovsky EVP and GM, Data Center Networking Business
Slide 101: Marvell Data Center Networking
Marvell Data Center Networking Cloud Optics Switching Integrated Optics Pluggable transceivers Scale-out and NPO and CPO and coherent DCI scale-up © 2026 Marvell. All rights reserved. 101
Slide 102: Established switching portfolio, accelerating momentum
Established switching portfolio, accelerating momentum Proven multi-generational execution Ethernet Line card and Liquid-cooled Silicon die package chassis ESUN 100T BGA/CPC/NPO/CPO Scale-out and Scale-up Switch bandwidth 12.8T 51T 100T 200T 400T © 2026 Marvell. All rights reserved. 102
Slide 103: Established switching portfolio, accelerating momentum
Established switching portfolio, accelerating momentum Proven multi-generational execution Ethernet ESUN UALink Switch bandwidth 12.8T 51T 100T 200T 400T Customer deployment at scale | Comprehensive switch portfolio extending to 200T, 400T, and UAL © 2026 Marvell. All rights reserved. 103
Slide 104: The industry's only portfolio of all switch protocols
The industry’s only portfolio of all switch protocols Logical layer: Protocol UALink NVLink Ethernet / ESUN Marvell network solutions Protocol agnostic | Tailored to each customer Marvell: One partner across every protocol © 2026 Marvell. All rights reserved. 104
Slide 105: Marvell connects AI infrastructure across every distance
Marvell connects AI infrastructure across every distance Distributed DCs Data center Rack | Multi-rack Package Up to 1,000km Up to 2km Up to 50m Up to 100mm Scale-across Scale-out Scale-up Scale-in Leadership across the full connectivity stack © 2026 Marvell. All rights reserved. 105
Slide 106: AI workloads drive new infrastructure demands
AI workloads drive new infrastructure demands Agentic AI + AI adoption at scale MoE + long context ▪ Dynamic all-to-all traffic Reasoning + inference-time ▪ Higher connectivity bandwidth Foundation compute ▪ More tokens per dollar ▪ More concurrent interactions models at scale ▪ Larger KV cache ▪ Mo
Slide 107: Reasoning at scale creates a memory challenge
Reasoning at scale creates a memory challenge Longer, more KV cache grows Required KV entry Reload or Recompute Impact on system complex context with context is not in fast memory (and evict to make room) efficiency Prompt Required KV entry Higher is not in fast memory latency Reasoning Reload Recom
Slide 108: Memory scaling beyond the server and rack
Memory scaling beyond the server and rack Memory architectures scale with inference demand XPU / server Rack Pod / multi-rack Direct-attach memory (SRAM / HBM) Memory-semantic network Larger scale-up domains Fastest access More capacity Pod-scale access Capacity constrained within Scale-up network e
Slide 109: Scaling beyond the rack requires optics
Scaling beyond the rack requires optics Beyond the XPU package and the server Beyond the rack Inference-time compute / reasoning models The network must extend beyond require access to more memory the rack to access more memory Higher bandwidth Memory can only scale if the network Lower scales with
Slide 110: Scale-up requirements drive optics closer to compute
Scale-up requirements drive optics closer to compute XPU 1 accesses memory in XPU 2 through the scale-up network Scale-up bandwidth per XPU XPU 1 + HBM ≥16 Tbps XPU 2 + HBM Equivalent to roughly 10 × 1.6T links Scale-up network requirements 10x bandwidth density Low latency Low power Reach Supportin
Slide 111: Hyperscale data center optical scale-up networks
Hyperscale data center optical scale-up networks Physical layer: Copper Optics Module Providers Marvell ”Fast Pipe” Light Engines Marvell “Flat Pipe” Light Engines 200G NPO 200G Light Engine 400G Light Engine MRM / OCI MSA Photonic FabricTM SiGe Drivers / TIAs NPO / CPO NPO / CPO NPO / CPO NPO / CPO
Slide 118: Scale-up networks co-engineered end to end
Scale-up networks co-engineered end to end XPU Scale-up switch XPU I/O I/O I/O Physical layer Physical layer I/O chiplet chiplet chiplet chiplet Protocol Protocol Compute side Compute side Design for link performance Architect for rapid bring-up Operate with confidence Signal integrity, low power an
Slide 119: Customer operational priorities shape architecture
Customer operational priorities shape architecture Physical layer: Copper Optics NPO: Near-package optics Lower failure blast radius Different priorities Different architectures Field-replaceable One goal: optics optimize the system for Flexible and each customer configurable © 2026 Marvell. All rig
Slide 120: Customer operational priorities shape architecture
Customer operational priorities shape architecture Physical layer: Copper Optics CPO: Co-packaged optics Lowest power and latency Different priorities Different architectures Greater integration One goal: and reduced system complexity optimize the system for Optimized each customer for maximum integ
Slide 121: Shared memory beyond the XPU
Shared memory beyond the XPU Marvell® Photonic FabricTM Memory Module Photonic FabricTM Memory Appliance High-bandwidth, low-latency access to shared memory across the scale-up domain © 2026 Marvell. All rights reserved. 121
Slide 122: The trusted partner for scale-out, scale-up switching and optics solutions
Scale-out Switching Scale-up Switching Scale-up Optics Growing to >$1B in FY28 Strong customer traction Significant Marvell content Multi-billion $ annual trajectory Additional multi-billion $ annually Revenue opportunity substantially greater than scale-up switching The trusted partner for scale-ou
Slide 123: Will Chu
Will Chu EVP and GM, Custom Cloud Solutions Business
Slide 124: Multiple multi-generational wins at top hyperscalers
Multiple multi-generational wins at top 4 hyperscalers Custom Cloud Solutions XPU Marvell core IP XPU attach © 2026 Marvell. All rights reserved. 124
Slide 125: XPUs growing in complexity to scale AI performance
XPUs growing in complexity to scale AI performance N N+1 N+2 1-2 die | 4-6 HBM 4-8 die | 12 HBM | 2-4 copper I/O 8+ die | 12+ HBM | 4+ optical I/O Marvell critical technologies SerDes Die-to-die Advanced packaging Custom SRAM Custom HBM PIVR 3D DRAM NPO CPO NVLink Fusion Connectivity IP is the catal
Slide 126: Custom opportunity expanding rapidly
Custom opportunity expanding rapidly Traditional data centers AI data centers Standard / merchant products Custom / semi-custom products NIC Storage NIC Mem ory CPU Mem ory Security expansion Security Management Near-memory Inference M an agement compute accelerator co-processor Standard product IP
Slide 127: AI workloads outgrow off-the-shelf NICs
AI workloads outgrow off-the-shelf NICs Traditional data center AI challenges Marvell ▪ Foundational NICs are prevalent ▪ Specific I/O speed, BW and config Marvell® Custom NIC ▪ SmartNICs with integrated CPUs ▪ Optimized data path, accelerators ▪ I/O speed <100G ▪ Custom network protocols ▪ Optimize
Slide 128: Memory expansion and near-memory compute
Memory expansion and near-memory compute Traditional data center AI challenges Marvell ▪ CPU-bound workloads ▪ Inference and KV cache workloads Marvell® Structera Custom ▪ Direct-attach memory sufficient ▪ Memory-bound, latency-sensitive ▪ Marvell early standard CXL effort ▪ CPU and XPU capacity and
Slide 129: AI storage demands semi-custom silicon
AI storage demands semi-custom silicon Traditional data center AI challenges Marvell ▪ Merchant controllers for SSDs ▪ Multi-gen FW / SW control and ▪ Flash-agnostic storage controllers ▪ Merchant accelerators for offload multi-vendor NAND flexibility ▪ AI-optimized storage accelerators ▪ Storage of
Slide 130: A new class of AI infrastructure management silicon
A new class of AI infrastructure management silicon Traditional data center AI challenges Marvell ▪ BMC manages standard servers ▪ Extremely complex AI servers AI infrastructure ▪ ~150 sensors, 100s KB/day ▪ >1500 sensors, 100s MB/day management ▪ Minimal processor and I/O perf ▪ ~1000X more perform
Slide 131: Inference offload opens a new XPU attach market
Inference offload opens a new XPU attach market Next-gen AI challenges Marvell ▪ Inference architectures evolving ▪ Need to offload XPUs for inference SerDes Custom SRAM ▪ Specialized co-processors attached to XPUs Die-to-die Adv. package Well-positioned to win in exciting new XPU attach market © 20
Slide 132: Significant custom opportunity
Significant custom opportunity More More More Higher Higher Marvell customers XPUs attach attach ASPs opportunity categories rate Building a balanced portfolio © 2026 Marvell. All rights reserved. 132
Slide 133: XPU and XPU attach driving growth
XPU and XPU attach driving growth Adding XPU attach $12B+ in FY29 $30B+ in FY31 to growing XPU business Updated forecast at midpoint target, based for greater diversity on design win diversity Customers, team, and platform to deliver on the opportunity © 2026 Marvell. All rights reserved. 133
Slide 134: Dan Durn
[=| MARVELL’ Dan Durn EVP and CFO
Slide 135: Long-term model execution FY21 → FY26
Long-term model execution FY21 → FY26 Revenue accelerated Op income expanded Capital allocation >$8B ~$11B $4B 23% CAGR 32% CAGR Organic investment* M&A Return to shareholders vs. 15–20% target set in FY21 Delivered earnings leverage Disciplined capital allocation We exceeded our targets while addin
Slide 136: What you have heard today
What you have heard today Market opportunity Technology leadership Revenue opportunity DSP and optical ~$400B TAM $70-90B Analog mixed signal ~45% CAGR CY25-CY30 Silicon and packaging 55-60% CAGR FY26-FY31 Purpose built to deliver significant growth © 2026 Marvell. All rights reserved. 136
Slide 137: What's in front of us
What’s in front of us Revenue Earnings Cash flow 55-60% Faster than Strong 5-year CAGR FY26-FY31 revenue and rising Revenue growing faster Operating leverage Funds growth and than market at scale shareholder returns Faster growth, expanding margin, and rising cash generation © 2026 Marvell. All righ
Slide 138: Faster revenue growth, more data center
Faster revenue growth, more data center Total revenue $70–90B Data center Comms and other Data center ~65% Rev CAGR FY26–FY31 55-60% CAGR Comms and other >3% $8.2B Rev CAGR FY26–FY31 FY26 FY31 © 2026 Marvell. All rights reserved. 138
Slide 139: Data center growth is broad-based
Data center growth is broad-based Interconnect Switching and storage Custom ~65% ~45% ~80% FY26-FY31 revenue CAGR* FY26-FY31 revenue CAGR FY26-FY31 revenue CAGR ~$65B TAM in FY31 ~$85B TAM in FY31 ~$235B TAM in FY31 Three large and diversified high-growth engines *FY26-FY31 midpoint of revenue CAGR
Slide 140: Significant operating leverage
Significant operating leverage Gross Margin Operating Margin Influenced by product mix Increases with scale Growth in custom OpEx to rise at ~half the % rate of revenue growth Earnings compound faster than revenue © 2026 Marvell. All rights reserved. 140
Slide 141: Capital allocation priorities
Capital allocation priorities Organic growth Strategic M&A Return capital Talent, IP, and supply Proven track record >50% of free cash flow, to fund the roadmap of acquiring technology on average over time and teams Funding growth and delivering returns while maintaining a strong balance sheet © 202
Slide 142: Long-term Non-GAAP financial model
Long-term Non-GAAP financial model Metric FY31 target model Revenue $70-90B Gross Margin Operating Margin 56-59% 44-46% >$30 Tax Rate 15% Free Cash Flow >36% FY31 implied EPS © 2026 Marvell. All rights reserved. 142
Slide 143: Why invest in Marvell?
Why invest in Marvell? $400B $20B $12B+ $70-90B >$30 CY30 Marvell FY28 updated total FY29 updated custom FY31 total FY31 implied TAM forecast revenue target revenue target revenue target Non-GAAP EPS Trusted by the world’s AI infrastructure leaders © 2026 Marvell. All rights reserved. 143
Slide 144: Marvell Investor Day October 6, 2026
[=| MARVELL’ INVESTOR DAY OCTOBER 6, 2026
Slide 146: Key terms
Key terms ▪ ACC: Active copper cable ▪ MACsec: Media access control security ▪ AEC: Active electrical cable ▪ LED: Micro-light-emitting diode ▪ ADC: Analog-to-digital converter ▪ MZM: Mach-Zehnder modulator ▪ BMC: Baseboard management controller ▪ MRM: Micro-ring Modulator ▪ CPO: Co-packaged optics
Slide 148: Discussion of Non-GAAP financial measures
Discussion of Non-GAAP financial measures Non-GAAP financial measures exclude the effect of stock-based compensation expense, amortizat ion of acquired intangible assets, acquisit ion and divestiture related costs, rest ructuring and other related charges (gains), (including, but not limited to, cha
Slide 149: Reconciliation of GAAP gross profit to Non-GAAP gross profit
Appendices Reconciliation of GAAP gross profit to Non-GAAP gross profit (Millions) (Unaudited) Q2’26 Q1’27 Q2’27 GAAP gross profit $ 1,010.6 $ 1,260.8 $ 1,455.6 Special items – expenses (income): Stock-based compensation 13.4 14.2 15.9 Amortization of acquired intangible assets 167.4 150.8 142.7 Res
Slide 150: Reconciliation of GAAP gross margin to Non-GAAP gross margin
Appendices Reconciliation of GAAP gross margin to Non-GAAP gross margin (Unaudited) Q2’26 Q1’27 Q2’27 GAAP gross margin 50.4% 52.1% 53.1% Stock-based compensation 0.7% 0.6% 0.6% Amortization of acquired intangible assets 8.3% 6.3% 5.2% Restructuring related charges (gains) (a) ⎯% (0.1) % ⎯% Other co
Slide 151: Reconciliation of GAAP operating expenses to Non-GAAP operating expenses
Appendices Reconciliation of GAAP operating expenses to Non-GAAP operating expenses (Millions) (Unaudited) Q2’26 Q1’27 Q2’27 GAAP operating expenses $ 720.5 $ 921.4 $ 995.9 Special items – (expenses) income: Stock-based compensation (140.2) (193.4) (310.3) Amortization of acquired intangible assets
Slide 152: Reconciliation of GAAP operating income to Non-GAAP operating income
Appendices Reconciliation of GAAP operating income to Non-GAAP operating income (Millions) (Unaudited) Q2’26 Q1’27 Q2’27 GAAP operating income $ 290.1 $ 339.4 $ 459.7 Special items – (expenses) income: Stock-based compensation 153.6 207.6 326.2 Amortization of acquired intangible assets 243.7 225.2
Slide 153: Reconciliation of GAAP operating margin to Non-GAAP operating margin
Appendices Reconciliation of GAAP operating margin to Non-GAAP operating margin (Unaudited) Q2’26 Q1’27 Q2’27 GAAP operating margin 14.5 % 14.0 % 16.8 % Stock-based compensation 7.7 % 8.6 % 11.9 % Amortization of acquired intangible assets 12.1 % 9.3 % 7.8 % Restructuring related charges (gains) (a)
Slide 154: Reconciliation of GAAP interest and other loss to Non-GAAP
Appendices Reconciliation of GAAP interest and other loss to Non-GAAP interest and other loss (Millions) (Unaudited) Q2’26 Q1’27 Q2’27 GAAP interest and other loss, net $ (56.4) $ (256.1) $ (81.4) Special items – expenses (income): Change in fair value of contingent consideration ⎯ 250.7 52.0 liabil
Slide 155: Reconciliation of GAAP net income to Non-GAAP net income (1 of 2)
Appendices Reconciliation of GAAP net income to Non-GAAP net income (1 of 2) (Millions) (Unaudited) Q2’26 Q1’27 Q2’27 GAAP net income $ 194.8 $ 34.5 $ 308.0 Special items – expenses (income): Stock-based compensation 153.6 207.6 326.2 Amortization of acquired intangible assets 243.7 225.2 214.9 Rest
Slide 156: Reconciliation of GAAP net income to Non-GAAP net income (2 of 2)
Appendices Reconciliation of GAAP net income (loss) to Non-GAAP net income (2 of 2) (a) Restructuring and other related items include gain on sale of property, changes in contractual obligations, employee severance costs, facility exit related charges, and other. (b) Other costs in operating expense
Slide 157: Reconciliation of GAAP diluted EPS to Non-GAAP diluted EPS
Appendices Reconciliation of GAAP diluted net income per share to Non-GAAP diluted net income per share (Millions, except per share amounts) (Unaudited) Q2’26 Q1’27 Q2’27 GAAP weighted-average shares outstanding ⎯ basic 862.6 882.0 897.4 GAAP weighted-average shares outstanding ⎯ diluted 870.4 893.3
Slide 158: Reconciliation of GAAP outlook to Non-GAAP outlook (1 of 3)
Appendices Reconciliation of GAAP outlook to Non-GAAP outlook (1 of 3) (Millions, except per share amounts) (Unaudited) Q3’27 GAAP net revenue $3,150 +/- 5% Special items: ⎯ Non-GAAP net revenue $3,150 +/- 5% GAAP gross margin 52.9% - 53.9% Special items: Stock-based compensation ~0.8% Amortization
Slide 159: Reconciliation of GAAP outlook to Non-GAAP outlook (2 of 3)
Appendices Reconciliation of GAAP outlook to Non-GAAP outlook (2 of 3) (Millions, except per share amounts) (Unaudited) Q3’27 Total GAAP operating expenses ~ $1,015 Special items: Stock-based compensation 285 Amortization of acquired intangible assets 72 Restructuring related charges 1 Other 2 Total
Slide 160: Reconciliation of GAAP outlook to Non-GAAP outlook (3 of 3)
Appendices Reconciliation of GAAP outlook to Non-GAAP outlook (3 of 3) (Millions, except per share amounts) (Unaudited) Q3’27 GAAP diluted net income per share $0.53 +/- $0.05 Special items: Stock-based compensation 0.34 Amortization of acquired intangible assets 0.21 Other income tax effects and ad
Slide 161: Marvell Investor Day October 6, 2026
[=| MARVELL’ INVESTOR DAY OCTOBER 6, 2026

Marvell Technology — Investor Day (October 2026) on 6ix

Marvell Technology — Investor Day (October 2026)

Slide 1: Marvell Investor Day October 6, 2026

Slide 2: Ross Seymore

Slide 3: Agenda

Slide 4: Basis of presentation

Slide 5: Forward-looking statements

Slide 6: Non-GAAP financial measures

Slide 7: Matt Murphy

Slide 8: © 2026 Marvell. All rights reserved.

Slide 9: © 2026 Marvell. All rights reserved.

Slide 10: A strong point of view on where industry is going

Slide 11: A strong point of view on where industry is going

Slide 12: A strong point of view on where industry is going

Slide 13: A strong point of view on where industry is going

Slide 14: A strong point of view on where industry is going

Slide 15: Architecting the Marvell platform

Slide 16: AI workloads are reshaping the network

Slide 17: Hyperscalers rearchitecting their networks

Slide 18: Connectivity is the next defining unlock

Slide 19: Connectivity is the next defining unlock

Slide 20: Marvell is a different kind of AI company

Slide 21: Marvell connects AI infrastructure across every distance

Slide 22: Marvell connects AI infrastructure across every distance

Slide 23: Marvell connects AI infrastructure across every distance

Slide 24: One switching foundation for scale-out and up

Slide 25: Unmatched leadership in storage and memory

Slide 26: Dramatically expanding custom opportunity

Slide 27: Custom revenue is accelerating

Slide 28: Custom revenue is accelerating

Slide 29: Data center CAPEX drives long-term growth

Slide 30: Data center CAPEX drives long-term growth

Slide 31: Growth across the stack

Slide 32: Our growth continues to accelerate

Slide 33: Our growth continues to accelerate

Slide 34: 2024 | A bold data center target

Slide 35: 2025 | Raising the data center target

Slide 36: Today | Data center target pulled in one year early

Slide 37: Today | Data center target pulled in one year early

Slide 38: Setting our long-term outlook

Slide 39: Marvell long-term revenue target

Slide 40: What underpins the long-term outlook

Slide 41: Why invest in Marvell?

Slide 42: Chris Koopmans

Slide 43: Scaling the company to meet the opportunity

Slide 44: Scaling the company to meet the opportunity

Slide 45: AI buildout continues at massive scale and speed

Slide 46: AI buildout continues at massive scale and speed

Slide 47: AI buildout continues at massive scale and speed

Slide 48: Large hyperscale data centers driving majority

Slide 49: Focus on large customers, deep partnerships

Slide 50: Marvell customer journey spans full platform

Slide 51: Cementing strategic partnerships

Slide 52: Scaling the company to meet the opportunity

Slide 53: Scaling engineering to meet opportunity

Slide 54: Building the company for the long term

Slide 55: Embracing AI to accelerate silicon engineering

Slide 56: Cloud-first design engineering mindset drives scale

Slide 57: Scaling the company to meet the opportunity: Supply chain

Slide 58: Supply chain excellence at scale

Slide 59: Scaling for growth

Slide 60: Sandeep Bharathi

Slide 61: 2020 vision | Data infrastructure silicon provider

Slide 62: 2026 | Delivered and proven at every step

Slide 63: AI tokenomics is a connectivity challenge

Slide 64: Memory wall stalls compute

Slide 65: Crossing the memory wall

Slide 66: Optics enables multi-rack, scale-up domains

Slide 67: Breaking the copper wall with photonics

Slide 68: Common technology foundation

Slide 69: Going beyond CMOS

Slide 70: Analog mixed signal powers data movement engine

Slide 71: Packaging innovation for photons and electrons

Slide 72: Industry-leading core technology

Slide 73: Tokenomics is physics

Slide 74: Xi Wang

Slide 75: The broadest end-to-end AI connectivity portfolio

Slide 76: Industry-first technology from the ground up

Slide 77: TIAs and Drivers required for every optical interconnect

Slide 78: Marvell high-speed analog foundation

Slide 79: Multi-generational DSP platform leadership