New demand, new markets: What edge computing means for hardware companies

Page created by Aaron Ferguson
 
CONTINUE READING
New demand, new markets: What edge computing means for hardware companies
New demand, new markets:
What edge computing means
for hardware companies
With over 100 edge use cases identified, the fast-growing need to power connected devices
demands a custom response from vendors.

JM Chabas, Chandra Gnanasambandam, Sanchi Gupte, and Mitra Mahdavian

OCTOBER 2018 • HIGH TECH PRACTICE
As connected devices proliferate and their                  In considering the sheer variety and volume of
capabilities expand, so does the need for real-time         edge use cases, it becomes clear that the demand
decision making untethered from cloud computing’s           for the edge computing technologies that enable
latency, and from connectivity in some cases. This          them will in turn create myriad opportunities
movement of computational capacity out of the               in a vast number of industries. In this article, we
cloud—to the edge—is opening up a new sector: edge          map out more than 100 edge computing use cases
computing.                                                  across 11 sectors that we believe could create
                                                            more than $200 billion in hardware value in the
By circumventing the need to access the cloud to            next five to seven years (Exhibit 1).
make decisions, edge computing provides real-time
local data analysis to devices, which can include           What drives edge computing?
everything from remote mining equipment and                 To understand what effect edge computing will
autonomous vehicles to digital billboards, wearable         have, it is important to know what causes its
health appliances, and more.                                evolution. Edge computers have an infinitely
                                                            wide range of uses; however, the conditions in
IoT devices tend to operate under different                 which they operate form the driving factors for
conditions from those of the controlled                     this new field and the technologies that serve it.
environments of offices and factories, driving              These are:
demand for a whole new set of technologies that
can allow computing in those situations. Take the           Varied connectivity and data mobility. Edge
scenario of a military drone deployed on a tactical         technologies can operate in places that might
surveillance mission in a high-intensity combat             limit or require intermittent connectivity to
zone. It is essential for the drone to be able to           the cloud for services like computing, storage,
collect, process, and transmit high-quality data in         backup, and analytics.
real time, despite numerous challenges, including
remote location, limited connectivity, and extreme          Need for real-time decision making. Edge
environmental conditions.                                   use cases often require data to be processed
                                                            instantly, for self-driving cars or automatic
While the drone can use mobile satellite                    picking machines, for example. These devices
connectivity to access the secure military cloud, it’s      and platforms need to be able to do analytics
much faster if it does the computing onboard, using         locally, without first sending data to the cloud, so
lightweight data storage and compute power. This            decisions can be made rapidly.
way of computing—on the edge of the cloud—lets the
drone stay in sync with both the command center             Localized compute power. Edge computers
and troops on the ground, without the latency that          need to be lightweight devices that can make fast,
computing in the cloud involves. Once the job is done,      secure decisions without the support of bigger
the drone returns to base and can connect with the          computing power.
larger system, now transferring its data to the cloud,
                                                            New storage and security needs. As the
where it can be used to feed algorithms and other
                                                            numbers of sensors generating data on remote
advanced analytics activities.
                                                            and mobile devices grow, so does the need for

New demand, new markets: What edge computing means for hardware companies                                          2
TMT 2018
                 New demand, new markets: What edge computing means
                 for hardware companies
                 Exhibit 1 of 6

    Exhibit 1    Edge computing represents a potential value of $175B–$215B in
                 hardware by 2025.
                                      % of total edge                                                                  % of total edge
                 Industry             use cases       2025 hardware value1                           Industry          use cases       2025 hardware value1

                1 Hardware value includes opportunity across the tech stack (ie, the sensor, on-device firmware, storage, and processor)
                 and for a use case across the value chain (ie, including edge computers at different points of architecture).

                efficient storage that can be secured in a variety of                            technologies need to be more specialized. For
                environments.                                                                    example, the data storage and computing power
                                                                                                 needed for precision agriculture will be different
                Intermittent power. Power and infrastructure                                     from that needed to run mobile, durable medical
                variations at the edge are pushing the boundaries                                appliances or safety equipment in a mine.
                of performance and capabilities of edge solutions.
                Especially in industrial applications, edge                                      In our research we identified 107 edge computing
                computers need to be able to operate with a power                                use cases (Exhibit 2). These applications are
                supply that might be sporadic.                                                   not conceptual: we identified 3,000 companies
                                                                                                 deploying these use cases today to understand
                What opportunities will edge computing open?                                     the potential opportunities across sectors and
                Unlike recent technological advances such as cloud                               the technology stack. Our calculation of the value
                computing, where most gains were captured by just                                they could generate focused on hardware, but of
                a few major players in the technology sector, edge                               course the opportunity extends to software too. We
                computing creates opportunities across a breadth                                 considered the hardware stacks (the value of the
                of industries. In addition, while much of today’s                                sensor, on-device firmware, storage, and processor,
                technical infrastructure is sector-agnostic—the                                  for example) and use cases across the edge value
                same cloud that powers an ecommerce engine also                                  chain (including edge computers at different points
                powers the workflow of a bank—edge computing                                     in the architecture).

3               High Tech Practice
New demand, new markets: What edge computing means
            for hardware companies
            Exhibit 2 of 6

Exhibit 2   A comprehensive market analysis revealed 107 concrete use cases
            for edge computing. (1 of 2)

            New demand, new markets: What edge computing means for hardware companies   4
New demand, new markets: What edge computing means
     for hardware companies
     Exhibit 3 of 6

     A comprehensive market analysis revealed 107 concrete use cases
     for edge computing. (2 of 2)

5   High Tech Practice
To chart each particular opportunity, we adopted                          improvements in human productivity and safety
            an industry lens to conduct our analysis, identifying                     from edge computing.
            edge use cases and quantifying the potential
            resulting hardware value. Based on the percentage       Given the central theme of edge computing—that
            of edge use cases in each vertical, the top three       a majority of the computing is done closer to
            verticals are:                                          where the data is being generated, and so real-
                                                                    time decision making can’t rely on the traditional
              ƒƒ Travel, transportation, and logistics              cloud or massive on-premises data centers, the
                                                                    varying conditions that each use case involves
              ƒƒ Global energy and materials                        drive the technology needed for it. Looking at
                                                                    this through the lens of specific use cases gives
              ƒƒ Public sector and utilities
                                                                    a sense of the range of technologies that will be
                                                                    needed. For example, autonomous vehicles rely
             The benefits of past technology revolutions were
                                                                    on visual processing, among other technologies,
             concentrated in sectors with heavy tech users, such
            TMT 2018                                                and these systems have to be able to withstand
             as financial services. For edge computing, sectors
            New demand, new markets: What edge computing            ruggedmeans
                                                                            environments that involve variations in
             that have traditionally been less tech-intensive, such
            for hardware companies                                  weather, vibration, and connectivity. (Exhibit 3).
             as energy and materials, stand to make substantial
            Exhibit 4 of 6

Exhibit 3   Individual use cases have specific operating conditions and customer
            needs that shape technology requirements. (1 of 2)

                                                               Computing              Edge
                                       Overview                needs                  computer          Ecosystem             Environment
                                       Autonomous              Real-time            • Autonomous        • Autonomous          Rugged, mobile
                                       vehicles must make      decision-making:       vehicle             vehicle OEMs        outdoor
                                       instantaneous           To avoid fatal       • Garage-based        and integrators     environment with
                                       driving decisions       consequences,          data center1      • Automotive OEM      broad variations in
                                       based on data           data must be                               suppliers           temperature,
                                       collected via LIDAR,    processed in                             • LIDAR, RADAR,       vibration, and
                                       RADAR, and video        real-time for                              and video           connectivity
                                       cameras                 immediate                                  camera vendors
                         Autonomous    Once the car            decisions to turn,
                         vehicles      returns to its          brake, or accelerate
                                       garage, data may
                                       be offloaded to
                                       edge computer for
            Travel,
                                       further analytics
            transport,
            and
                                       Location-based          Localized              Processor         • Local, state, and   Semi-outdoor
            logistics                  advertising in public   computing power:       embedded in         national transit    environment with
                                       transportation uses     Displays must          display systems     authorities         intermittent
                                       the location of a       function               on vehicle        • Advertising         connectivity that
                                       vehicle to              independently with                         agencies            varies by route
                         Location-     customize display       limited connectivity                                           (eg, urban, rural,
                                       advertisements          Compact form                                                   aboveground, and
                         based
                                       near consumers          factor: Devices                                                underground
                         advertising   The advent of 5G        must be
                                       could make              lightweight, small,
                                       location-based          and low-power
                                       advertising possible
                                       at scale

            New demand, new markets: What edge computing means for hardware companies                                                               6
Individual use cases have specific operating conditions and customer
     needs that shape technology requirements. (2 of 2)

                                                                Computing          Edge
                                     Overview                   needs              computer            Ecosystem           Environment
                                      Highly digitized          Real-time           Hyperconverged2 • Oilfield services    Harsh external
                                      drilling rigs             decision-making: edge appliance       companies that       environment
                                      generate data from        With limited                          integrate the        with drastic
                                      multiple sensors on       connectivity, data                    end-to-end           temperature
                                      equipment; the data       must be analyzed                      drilling solutions   variations and
                                      collected by these        and acted upon in                     for oil and gas      limited
                    Offshore          sensors needs to          real-time to avoid                    rigs                 connectivity.
                    drilling rigs     be processed on           damage to                           • Hyper-con
                                      the rig to avoid          expensive                             verged solution
                                      equipment damage          equipment                             vendors
                                      and interruptions in      Rugged form
                                      operations                factor: Device must
     Global                                                     function in harsh,
     energy                                                     at-sea conditions
     and
     materials                        Sensors on                Real time            Hyperconverged   • Mining             Harsh outdoor
                                      monitoring                decision-making: edge appliance         corporations       and underground
                                      equipment, in the         With limited                          • Hyper-             environments,
                                      environment and on        connectivity, data                      converged          with limited or no
                                      employees,                must be processed                       solution vendors   connectivity
                    Health and        generate data that        and acted upon in
                    safety in         is processed in real      real time to prevent
                    mining            time to improve           dangerous or fatal
                                      workforce                 accidents
                                      productivity,             Rugged form
                                      workplace safety,         factor: Computer
                                      and operational           must withstand
                                      efficiency                harsh mine
                                                                environment
                                      Sensor modules with       Real-time            Hyperconverged   • State utility      Outdoor
                                      integrated compute        decision-making: appliance              companies          environment with
                                      are deployed in           In remote settings                    • Electronic         varying tempera-
                                      aquifers, treatment       with no connectivi-                     equipment          ture ranges, and
                                      plants, and pipes;        ty, information must                    manufacturers      indoor factory
                                      modules or                be processed                          • Hyper-con          environment, such
                                      hyper-converged           locally in real time                    verged solution    as central water
                                      appliances process        Rugged form                             vendors            treatment plants
                    Water             data on-site to           factor: device must
                    quality           monitor water quality     withstand outdoor
                    monitoring        in real-time, and in      environments
                                      areas with strong
                                      connectivity, data will
                                      be intermittently
                                      transferred to the
                                      cloud for centralized
     Public                           analytics
     sector
     and                              Sensors and video         Real time           Hyperconverged    • State              Outdoor
     utilities                        cameras deployed          decision-making: appliance              infrastructure     environment
                                      on roads and at           data must be                            companies          exposed to
                                      traffic lights capture    processed in real                     • Third-party toll   the elements,
                                      data on traffic flow;     time to respond to                      operators          with varying
                                      information is            congestion quickly                    • Hyperconverged     temperature
                                      transmitted to a          Efficient storage:                      solution vendors   and moisture
                    Congestion        hyper-converged           large volumes of                                           ranges
                    lanes             appliance2 hosted         video data must be
                                      at intermittent street    quickly accessed
                                      locations to monitor      for analytics
                                      traffic patterns and      Rugged form
                                      implement                 factor: device must
                                      demand- based             withstand outdoor
                                      pricing                   environments

    1 Mobile micro data center deployed at telecommunications network edge.
    2 A hyperconverged appliance combines compute, storage, and networking resources in a single piece of hardware.

7   High Tech Practice
The edge will soon be everywhere
Very soon edge computers will be all around
us performing distributed computing across a
multitude of devices in homes and factories, on
farms, and throughout public infrastructure.
The forces fueling the demand for these devices
and the technologies enabling them are advancing
rapidly. For tech companies, the development of
edge technology will revolutionize industry with
solutions customized for diverse use cases. This
will create a paradigm shift from the device and
original equipment makers all the way through to
how such products are sold, installed, and serviced.
The changes that result will affect all players in
the tech stack, consumers in a vast array of sectors,
and any companies and leaders looking to have a
role in it.

JM Chabas is an associate partner in McKinsey’s
San Francisco office. Chandra Gnanasambandam
is a senior partner in the Silicon Valley office, where
Sanchi Gupte is a consultant and Mitra Mahdavian
is a partner.

The authors wish to thank Joseph Greenberg,
Brahmjot Singh Kohli, Ritika Bipin Singh, and Daniel
Sun for their contributions to this article.

Copyright © 2018 McKinsey & Company. All rights
reserved.

New demand, new markets: What edge computing means for hardware companies   8
You can also read