The Important

There has never been more “information”, analysis, noise, and stress to execute. Executing the wrong thing fast is not only a bad outcome, but also a waste of resources; resources that will never be reclaimed. Superior performance comes from focusing on The Important, so that strategic synergies are achieved across a value chain, leading to more impactful use of resources and superior performance.

Watch List

Some examples of The Important (last updated August 24th, 2020). Will synthesize down to a smaller list later.

Internet Devices & Users

Strong Growth

  • Devices with voice-based interfaces
  • E-commerce / on-demand
  • Cloud service revenue
  • Digital payments
  • Wearables


  • Digital media consumption continuing to grow, driven by mobile devices
  • Interactive / social gaming
  • Internet devices dominate Advertising spend, TV still significant; mobile growth strong
  • APAC dominates Internet users (especially China & India)
  • USA dominates top 5 capitalized Internet properties

Slowing growth

  • The maturation of the smartphone market in developed countries frames Service Provider/Telecom economics
  • High global access to Internet makes growth for entire Internet ecosystem more difficult (points of low access availability notwithstanding)


  • AR/VR/MR
  • Broad access to AI/ML-based expertise/expert advice
  • Machine-mind interfaces
  • Robots/cobots (collaborative robots), drones, automated vehicles
  • IoT

Internet Applications and Services

Consumers are becoming increasingly concerned with privacy and information rights:

  • Large Internet sites being hacked and personal information stolen
  • Consumer information is being mined, used, and monetized
  • State-sponsored cyber attacks
  • Social media overload concerns rising (the impact of spending so much time online)

Blockchain emerges as one of the solutions:

  • Well suited to immutable transaction journals.
  • Not a mutable database, so not well suited for all consumer concerns
  • Mutable use cases witll drive alternatives to blockchain

A flood of data

  • AI/ML emerges as a new tool for competitive advantage across many industries
  • Period of learning, exploration, tools development, etc.
  • Data is the new oil

Increasing software modularity

  • Virtual machines, containers, and serverless
  • Modularity improves statistical multiplexing across compute resources
  • Modularity improves code change isolation and maintenance
  • Modulartiy complicates management, requiring new tools

Customer Experience

  • Cloud disrupts customer experience
  • New expectation: frequent, frictionless, feature/maintenance velocity
  • New expectation: access data/service anywhere
  • New expectation: new capabilities based on new information

Operations Excellence

  • Cloud disrupts IT operations scale, efficiency, and agility
  • IT resources are elastic, highly efficient/automated, and fast to add/delete (in general)
  • Economic considerations aside, all companies / entities must assess whether they get benefit from operating there own IT environment, especially for new applications / services

Subscription Revenue Models

  • Cloud SERVICES are a good match for subscription revenue models, driving a business model inflection in software
  • Cloud also offers IT resources on a usage basis, driving low get-started costs, with potential high long run costs, for applications/services that are always on, unless terms are renegotiated with the provider.
  • Low upfront costs was one of the catalysts for shadow IT exploding


Packet transport

  • IP/MPLS to shift to SR MPLS & SRv6 over the next 1-5 years. Driven by a perceived reduction in complexity, server-to-server/handset networking, edge compute, and programmable traffic engineering.
  • Linux becomes network operating system industry defacto standard.
  • Internally containerized NOS software architecture becomes industry defacto standard.
  • 400ZR/ZR+ drives packet/optical integration over short WAN distances (0-80/120 KMs), particularly in DCI
  • Spine / leaf architectures appear in more parts of the network
  • Spine / leaf specific approaches to routing emerge

Edge Services

  • Shit to BGP-based PE for all VPNs, Ethernet and IP, as a common approach.
  • Possible separation between service routing and transport routing (unclear probability)

Mobile Networks

  • 5G, Flex-E/, Open RAN, Millimeter


  • Centralized analysis and recommendations augment distributed control planes
  • Automation drives configuration improvements
    • Centralized automation
    • Distributed self-configuration
  • Traffic engineering and fault management become predictive rather prescriptive or reactive through the application of AI/ML
  • Cloud-based network management has the ability to propogate across multiple networks, insights drawn from the Internet and multiple Enterprise networks

Compute / Silicon

  • The end of the CPU for everything era
  • Productivity gains driven by silicon that is purpose built, to solve specialized problems, with significant price/performance improvements over CPU
  • Potential of GPU being replaced by specialized ML training silicon
  • Expanding use cases for merchant network processing units
  • DSP technology continues to be e differentiator in the optical domain


  • Flash adoption
  • Multi-cloud data mobility
  • Non-volatile memory express (NVMe) / NVMe over fabrics
  • Software-defined storage
  • Ethernet storage fabric
  • Storage class memory


  • Cloud-based services
  • Potential integration of cloud-based security and networking services
  • Data breaches, state-sponsored cyberwar, mobile device risks, IoT risks, AI weaponization
  • Phishing, ransomware,..

On the Horizon (3-10 years out)

  • Quantum computing / networking augments classical approaches
  • Neuromorphic computing
  • IPv6 becomes default for new networks, despite some concerns
  • Serverless computing matures
  • Autonomous networking
  • Gallium nitride (GaN) and silicon carbide (SiC)