2026 Showcase

2026 Predictions Showcase

We kicked off 2026 with a series of predictions from leaders in infrastructure technologies. You'll find their takes on AI, networking, computing, and mobile technologies and what to expect this year. Curious how they did last time? Check out the 2025 predictions to see how accurate they were.

Also, grab our latest research brief right from this page.
Analysis by AvidThink

What the 2026 predictions point to

Across these infrastructure leaders, 2026 is the year AI stops being about bigger models and starts being about the physics and economics of running them at scale. The recurring throughline: inference, power, and the network fabric become the real battlegrounds, with Ethernet and a maturing optical roadmap underpinning it all.

  1. The year of inference

    AI shifts from training to decode-heavy inference, demanding far more memory capacity and forcing new economics. Credo, YOUnified AI, and Zayo all frame 2026 as the year inference costs reshape pricing, deployment, and ROI expectations.

  2. Power and cooling become the gating constraint

    Compute is limited by electricity and heat, not silicon. Adtran, Black & Veatch, and YOUnified AI point to collocated generation, SMRs, GPU-driven water management, and local politics making geography a first-order design factor.

  3. Ethernet wins the AI fabric

    Broadcom, DriveNets, Ethernet Alliance, and IEEE converge on Ethernet as the standard from scale-up to scale-out, with the 802.3dj roadmap pushing 200 Gbps signaling toward 1.6T and 400 Gbps signaling toward 3.2T and beyond.

  4. Optics, memory, and timing move center stage

    Marvell, OIF, Astera Labs, and SiTime highlight the enabling layers: co-packaged and increasingly fully optical connectivity, 400/448 Gbps and coherent lite, tiered memory beyond HBM, and resilient timing to keep GPUs and robots running.

  5. Autonomy, sovereignty, and a reality check

    Telecom shifts from automation to autonomy (Blue Planet), geopolitics drives sovereign and private cloud (Ericsson, Adtran), quantum and post-quantum security rise, and Graphiant expects an AI-bubble correction that refocuses the market on real business value.

More about this report →

Highlights from industry thought leaders

Video interviews

Adtran

AI, Security & New Revenue Opportunities

  • AI continues to dominate the industry in 2026, but power consumption is the binding constraint
  • Digital sovereignty and security advance, including post-quantum cryptography
  • Service providers evolve beyond infrastructure into digital platforms
Abstract
Stephan Rettenberger, SVP of Marketing and Corporate Communications at Adtran discusses AI's continued industry dominance constrained by power consumption; digital sovereignty and security advances including post-quantum cryptography; and service providers evolving beyond infrastructure into digital platforms.
Astera Labs

Specialized Accelerators, Hybrid Fabrics & Tiered Memory

  • Specialized accelerators take center stage as workloads demand adaptability and flexibility
  • Hybrid AI fabrics combine multiple protocols rather than a single interconnect standard
  • Tiered memory architectures address large-scale inference workloads
Abstract
Thad Omura, Chief Business Officer at Astera Labs, expects adaptability and flexibility through specialized accelerators, hybrid AI fabric solutions combining multiple protocols, and tiered memory architectures for large-scale inference workloads to be front and center in 2026.
Black & Veatch

Data Center Power, Water & AI Infrastructure Trends

  • Power suppliers and data center developers deepen partnerships, with collocated generation and SMRs becoming more prevalent
  • GPU cooling demands in warmer regions drive significant water management innovation
  • Neocloud providers offer distributed AI-as-a-service in the 20-80 MW range
  • Integrated campus simulations and AI-enabled operations become standard practice
Abstract
Jenn Cahill, Assistant VP of Campus Infrastructure Integration at Black & Veatch, projects that 2026 will bring increased partnerships between power suppliers and data center developers, with collocated generation and sustainable options like SMRs becoming more prevalent, alongside significant water management innovation driven by GPU cooling demands in warmer regions. She highlights the rise of neocloud providers offering distributed AI-as-a-service in the 20-80 megawatt range, and emphasizes that data-driven decision-making through integrated campus simulations and AI-enabled operations will become standard practice across the industry.
Blue Planet (a Ciena Company)

Telecom's Shift from Automation to Autonomy

  • Telecom shifts from automation to autonomy, with accurate OSS data a prerequisite for agentic AI
  • Autonomous network use cases move from pilots into production environments
  • CSPs accelerate legacy OSS decommissioning, retiring decades-old mainframes that block AI
Abstract
John McVey, Global VP of Customer Success at Blue Planet (a Ciena Company), presents three key predictions for how the telecom industry shifts from automation to autonomy in 2026, including the critical need for accurate OSS data as a prerequisite for agentic AI, the advancement of autonomous network use cases from pilots to production environments, and the acceleration of legacy OSS decommissioning as CSPs retire decades-old mainframe systems that block AI capabilities. McVey emphasizes that winning networks in 2026 will be smarter, self-governing systems built to operate at machine speed as Blue Planet helps CSPs modernize their OSS environments to unlock measurable autonomy.
Broadcom

Ethernet Dominance, Multi-Site Clusters & Co-Packaged Optics

  • Ethernet becomes the standard for scale-up networking
  • AI clusters expand across multiple data centers with Ethernet as the interconnecting fabric
  • Co-packaged optics go mainstream on the back of 70% power efficiency gains
Abstract
Hasan Siraj, Head of Software Products/Ecosystem at Broadcom, presents three networking predictions for 2026 centered on AI infrastructure, including Ethernet becoming the standard for scale-up networking, AI clusters expanding across multiple data centers with Ethernet as the interconnecting fabric, and co-packaged optics going mainstream due to 70% power efficiency gains. Broadcom is investing heavily in these technologies while collaborating with customers and partners to enable these infrastructure transitions.
Clearfield

5 Fiber Industry Trends Shaping 2026

  • BEAD funding and public-private partnerships accelerate global fiber deployment
  • Simplified design and training grow in importance to counter workforce shortages
  • Higher-capacity hardware and multi-gigabit technologies drive speed and density gains
Abstract
Anis Khemakhem, Chief Marketing Officer at Clearfield, identifies five major trends shaping the fiber industry, including accelerated global fiber deployment driven by BEAD funding and public-private partnerships, the growing importance of simplified design and training to address workforce shortages, and speed and density innovations through higher-capacity hardware and multi-gigabit technologies.
Credo

Inference, Neoclouds & Reliability Challenges

  • 2026 is "the year of inference" as AI shifts from training to decode-heavy applications
  • Decode-heavy inference requires significantly more memory capacity
  • AI spending moves from traditional hyperscalers to neoclouds needing vendor support for rapid deployment
Abstract
Don Barnetson, SVP & Head of Product at Credo, talks about "the year of inference" as AI shifts from training to decode-heavy applications requiring significantly more memory capacity, while AI spending moves from traditional hyperscalers to neoclouds that need vendor support for rapid deployment.
DriveNets

GPUs diversity and Ethernet in Scale-Up

  • The GPU market grows more heterogeneous, moving beyond a single dominant vendor
  • Networking becomes a critical AI infrastructure design consideration
  • Ethernet emerges as the leading technology across all connections, from scale-up to scale-out
Abstract
Dudy Cohen, VP of Product Marketing at DriveNets, presents three predictions for 2026 AI infrastructure: a more heterogeneous GPU market, networking becoming a critical design consideration, and Ethernet emerging as the leading technology across all AI infrastructure connections (from scale-up to scale-out).
Ericsson

Cellular, Sovereign Cloud & Supply Chain Issues

  • Cellular becomes a viable alternative to wired connections on technological and financial grounds
  • Geopolitical factors drive adoption of sovereign and private cloud for security and autonomy
  • DDR4 memory shortages impact network product lifecycles across the industry
Abstract
Archana Khetan, SVP, Product Management and Technical Marketing at Ericsson, presents three key predictions for 2026: cellular technology becomes a viable alternative to wired connections as customers actively consider it due to technological advancements and financial viability; geopolitical factors drive adoption of sovereign and private cloud solutions for security and autonomy; and DDR4 memory shortages impact network product lifecycles across the industry.
Ethernet Alliance

Ethernet's Path to AI Dominance

  • Ethernet will succeed in AI applications despite uncertainty over the specific path forward
  • The industry understands its objectives and will determine the approaches needed to meet customer requirements
Abstract
Peter Jones, Chair of the Ethernet Alliance and Cisco Systems employee, presents predictions for Ethernet in AI applications, expressing confidence that Ethernet will succeed despite uncertainty about the specific path forward. He emphasizes that while the exact methods remain unclear, the industry understands its objectives and will determine the necessary approaches to meet customer needs.
Fiber Broadband Association

Fiber Growth, Quantum & AI Infrastructure

  • 2026 sees record fiber deployment driven by private investment and BEAD program funding
  • Quantum networking becomes critical as "Q Day" threats accelerate the need for enhanced encryption
Abstract
Gary Bolton, President and CEO of the Fiber Broadband Association, predicts that 2026 will see record fiber deployment driven by private investment and BEAD program funding, while quantum networking becomes critical as "Q Day" threats accelerate the need for enhanced encryption security.
Graphiant

AI Bubble Correction in 2026: Network Infrastructure's Critical Role

  • A significant AI-bubble correction hits as model scaling and data center expansion plateau
  • Focus shifts from raw scaling toward actual business value and real-world applications
Abstract
Ali Shaikh, Chief Executive Officer of Graphiant, expects a significant correction in the AI bubble as rapid model scaling and data center expansion reach a plateau, shifting focus toward actual business value and real-world applications. He shares how this will impact networking in 2026.
IEEE

The Future of Ethernet: 200G to 3.2T and Beyond

  • The 802.3dj standard nears completion, introducing 200 Gbps signaling for speeds up to 1.6T Ethernet, with products already shipping
  • The industry moves to define 400 Gbps signaling as the foundation for 3.2T Ethernet and beyond
Abstract
John D'Ambrosia, Chair of the IEEE P802.3dj Task Force at IEEE, discusses the future of Ethernet, focusing on the completion of the 802.3dj standard introducing 200 Gbps signaling for speeds up to 1.6 terabit Ethernet, with products already shipping. He highlights the industry's move toward defining 400 Gps signaling as the foundation for next-generation speeds including 3.2Tbps Ethernet and beyond.
Intel

AI CPU Deployments, Edge Computing & 6G

  • Widespread AI deployments use CPU-native capabilities for radio algorithms and infrastructure optimization
  • Enterprises adopt on-premises edge computing with multi-tenancy support
  • The industry intensifies focus on defining 6G standards and monetization strategies
Abstract
Cristina Rodriguez, VP and GM of Network & Edge Group at Intel, forecasts widespread AI deployments using CPU native capabilities for radio algorithms and infrastructure optimization, enterprise adoption of on-premises edge computing with multi-tenancy support, and intensive industry focus on defining 6G standards and monetization strategies.
Marvell

Optical Networks, AI Chip Design & Memory Architecture

  • Hybrid electrical-optical connectivity evolves toward fully optical by 2027
  • AI capabilities become more deeply integrated into chip design technologies
  • Memory capacity expands beyond traditional HBM via attached or disaggregated architectures
Abstract
Noam Mizrahi, EVP and Corporate CTO at Marvell, presents three key predictions for the semiconductor and AI infrastructure landscape in the coming year. He anticipates hybrid electrical-optical connectivity evolving to fully optical by 2027, greater integration of AI capabilities into chip design technologies, and expanded memory capacity beyond traditional HBM through attached or disaggregated architectures.
OIF

OIF on 448G, Coherent Lite, CMIS & Power Efficiency

  • The industry pushes toward 400/448 Gbps implementation while tackling reliability, power efficiency, and reach
  • Coherent lite emerges as a developing market segment
  • CMIS becomes critical firmware infrastructure for AI applications
Abstract
Nathan Tracy, President of OIF, discusses 2026 predictions focusing on the industry's push toward 400/448 Gbps implementation while addressing reliability, power efficiency, and reach challenges, with particular emphasis on coherent lite market development and CMIS as critical firmware infrastructure for AI applications.
SiTime

AI Infrastructure, Bandwidth, Timing & Robotics

  • AI infrastructure drives demand for resilient timing to keep expensive GPUs running continuously
  • Networking bandwidth transitions from 800G to 1.6 Tbps
  • AI-enabled robotics require precision timing under extreme conditions where humans cannot operate
Abstract
Piyush Sevalia, Executive Vice President, Marketing at SiTime, presents three key predictions for 2026 focused on AI infrastructure's demand for resilient timing solutions, the transition from 800G to 1.6 terabits per second networking bandwidth, and AI-enabled robotics requiring precision timing under extreme conditions. He explains that keeping expensive GPUs running continuously, accelerating data delivery, and enabling robots to operate in challenging environments where humans cannot are driving significant design activity for SiTime's timing technology.
Spirent

5G Advanced, Telco AI, Satellite-to-Device & 6G

  • 5G Advanced gains momentum via RedCap IoT and location services
  • Telcos monetize AI with offerings like GPU-as-a-service
  • Non-terrestrial direct-to-device capabilities expand beyond emergency use cases
  • 6G standardization accelerates around new spectrum bands, integrated sensing, and cybersecurity
Abstract
Stephen Douglas, Head of Market Strategy at Spirent, talks about 5G Advanced momentum with RedCap IoT and location services, telco AI offerings like GPU-as-a-service, and expanded non-terrestrial direct-to-device capabilities beyond emergency use cases. He also discusses accelerated 6G standardization focused on new spectrum bands, integrated sensing, and cybersecurity.
VIAVI

AI Data Centers, Quantum Security & Sensing Convergence

  • Convergence of networks, security, AI, photonics, and sensing reshapes how infrastructure is designed, deployed, and tested
  • High-speed data center networking and AI in RAN change testing requirements
  • AI-enabled cybersecurity threats emerge as a growing concern
Abstract
Sameh Yamany, CTO & Chief AI Officer at VIAVI, discusses convergence of networks, security, AI, photonics, and sensing that will fundamentally change how the industry designs, deploys, and tests infrastructure. He forecasts how high speed data center networking, AI-enabled cybersecurity threats, and use of AI in RAN will have implications on testing needs this coming year.
YOUnified AI

When AI Gets Normal (Not Smarter)

  • AI becomes "normal" as infrastructure economics, security, and compute politics eclipse model-intelligence gains
  • Inference costs force new SaaS pricing models
  • AI security shifts from policy to system-design requirements
  • Data center power demands make geography and local politics critical factors
Abstract
Eugina Jordan, CEO of YOUnified AI, presents three key predictions for 2026, arguing that AI will become "normal" as fundamental infrastructure challenges around inference economics, security requirements, and compute politics take center stage rather than model intelligence improvements. She advises that inference costs will force new SaaS pricing models, AI security will shift from policy to system design requirements, and data center power demands will make geography and local politics critical factors, concluding that founders should "build boring with right economics in mind because boring always scales."
Zayo

3 Predictions for Enterprise AI and Infrastructure in 2026

  • Enterprise AI adoption enters an expansion phase with more ROI evaluations and operational integration
  • The Neocloud vertical sees market separation between winners and strugglers amid capital-intensity pressures
Abstract
James Tomko, SVP of Sales at Zayo, predicts that enterprise AI adoption will enter an expansion phase in 2026 with increased ROI evaluations and operational integration, while the Neocloud vertical will see market separation between successful and struggling companies amid capital intensity challenges.
Reader feedback on Data Center Networking in 2026 — Scaling to New Heights: "Appreciate the broad coverage and unique insights." "Up, Out, Across, and Away we go." Download our latest report and send us your feedback!
Download the Latest Report

By downloading this report, you agree to our Terms of Use and Privacy Policy.