Can You Enjoy Metaverse Without Edge Computing?

Edge computing is certain to play a bigger role in our computing fabric as augmented reality, virtual reality and future Metaverse environments become possible. “Even at ultra-low latency, it makes little sense to stream (versus locally process) AR data given the speed at which a camera moves and new input data is received (i.e. literally the speed of light and from only a few feet away),” says Matthew Ball, EpyllionCo managing partner.

The conventional wisdom today is that multi-player games, to say nothing of more-immersive applications, do not work when total latency is greater than 150 milliseconds and user experience is impaired when latency is as low as 50 milliseconds, says Ball. 

source: Matthew Ball 

Will the Metaverse require 1,000 times more computing power? Intel thinks so. And that implies we might be decades away from a ubiquitous and widely-accepted Metaverse that people actually use routinely. 

“Consider what is required to put two individuals in a social setting in an entirely virtual environment: convincing and detailed avatars with realistic clothing, hair and skin tones; all rendered in real time and based on sensor data capturing real world 3D objects, gestures, audio and much more; data transfer at super high bandwidths and extremely low latencies; and a persistent model of the environment, which may contain both real and simulated elements,” says Raja Koduri, Intel SVP and GM of Intel’s Accelerated Computing Systems and Graphics Group. “Now, imagine solving this problem at scale--for hundreds of millions of users simultaneously--and you will quickly realize that our computing, storage and networking infrastructure today is simply not enough to enable this vision.”

“We need several orders of magnitude more powerful computing capability, accessible at much lower latencies across a multitude of device form factors,” says Koduri. 

“Truly persistent and immersive computing, at scale and accessible by billions of humans in real time, will require even more: a 1,000-times increase in computational efficiency from today’s state of the art,” he notes. 

Whales Really Do Matter, for Firms of Any Size

Whales matter, in business, and for just about every business. Look at expected profit contributions across a firm’s customer base. The first 20 percent of customers will supply the most cumulative profit. The tail of the last 80 percent of customers might contribute almost nothing in that regard. 

source: Baker Tilly 

How can a software startup better estimate where its revenue might come from? Nnamdi Iregbulem, Partner at Lightspeed Venture Partners, actually has thought quite a lot about that subject. Basically, it appears the Pareto Principle holds for software startups, across the full range of firm sizes. 

“I think people talk about concentration as if there are a couple of companies that have revenue concentration issues, and then the rest are fine,” says Iregbulem. “ It just turned out that literally every company has pretty high customer concentration, not in the sense that there was one customer that was 10 percent of revenue, but in the sense that there was a subset of customers that were a pretty meaningful share, something like 20 percent being 70 percent of revenue.”

source: Nnamdi Regbulem 

In other words, in every industry segment, it is not just “whales” who see Pareto distributions, but firms of every size in every segment. The implications are clear enough: even a smaller firm targeting a smallish niche is going to have its own “whales” (a few customers than anchor total revenues). The same goes for profits. 

source: Baker Tilly 

“It’s a very common mistake I find among investors where they'll meet a company, the company will have X number of customers and the standard ACV (average contract value) will be fairly small because most of their users are either free users or in some kind of lowest-tier version of the product,” says Iregbulem. “But they do have a couple of meaningful customers that are spending real revenue or paying the highest tier of a product or what have you.”

It is not unusual for as much as 60 percent of customers to provide no more than “breakeven” performance, in terms of profitability. 

source: Baker Tilly 

In other words, software monetization is a power law. A power law distribution is a curve that looks like this: most of the results are generated by a small fraction of instances, products or customers. “Rather than the exception, high concentration is the norm in certain verticals (for example cloud infrastructure) or pricing models (consumption/pay-as-you-go) where a "customer" can be as small as tens of dollars per month,” he notes. 

source: Reaction Wheel 

In the data center and connectivity businesses, that rule tends to hold as well. A handful of customers anchor demand for global bandwidth and data center capacity. The rule also seems for  devices connected to Wi-Fi. 

“Combined, the above insights form a mathematical justification for "land and expand"-style go-to-market strategies,” he adds. “Here, land and expand is effectively an indexing strategy: land at as many organizations with as little investment as possible.”

“Every once in a while you'll land a Google, a Facebook, or an Amazon (both figuratively and literally) which will drive a disproportionate share of revenue,” he says.”Even if those customers start off small, any given customer could potentially become quite large.”

Whales matter, in business, for revenues and profits.

Hybrid Access Networks Might Get More Hybrid

The “hybrid” in “hybrid fiber coax” has been a key strategic approach used by the cable industry more generally. Fiber and coax; owned spectrum and mobile virtual network operator; MVNO and Wi-Fi offload; linear video and on-demand streaming; analog and digital; content ownership and distribution.

Keep in mind the "hybrid" business strategy often suggests itself at times of key technological change. Perhaps the classic example is the transition from sailing ships to steam-powered ships, when for 100 years many sailing ships were outfitted with steam boilers, but used both methods of propulsion.

The key point is that any hybrid still represents a transition strategy "from something to something else."

Now hybrid looks to happen for next-generation access networks. Cox Communications, joining others such as Comcast, says it will deploy fiber to the home and DOCSIS 4.0 networks using HFC. 

Though Cox has given few details, the network aims to supply 10 Gbps. Cox does not say whether that is symmetrical or not. DOCSIS 4.0 will feature upstream speeds to 6 Gbps, as the standard calls for. FTTH is almost certainly going to be symmetrical.  

 In a first, Virgin Media O2 plans to operate its gigabit hybrid fiber coax network side by side with the coming fiber to home network. That has never been done before at scale, if ever, by a cable TV company. 

The best analogy is the way mobile network operators run several generations of mobile networks simultaneously, using discrete radios, spectrum and often towers or radio sites. When telcos deploy optical fiber, they decommission the copper access network. 

That decision should have operating cost implications, positive and negative. On the positive side of the ledger, VMO2 will not have to migrate customers from the hybrid fiber coax network to the FTTH network in every case. 

That will save truck roll costs, customer premises equipment costs and some amount of customer service work to explain and convince people to switch over. 

On the negative side of the ledger, supporting two separate networks, with different underlying technologies, will not provide clear operating cost advantages. Energy costs for two networks will be necessary. Support staff will have to be trained to support both networks. 

It is possible migration will not be necessary in a majority of cases, for some time. So we will have to wait and see how the operating cost advantages and disadvantages appear over time. 

Take rates will be important. Though VMO2’s national FTTH network will  not operate in parallel everywhere, it will do so in areas where the HFC network already operates, which might mean 15.6 million or so homes. Seven milliion homes will have FTTH only.  

Cox and Comcast will operate two networks as well; just not two networks in the same area, as a rule. FTTH will undoubtedly be targeted to business-rich areas or higher-demographic suburban areas first, where the demand for multi-gigabit service is expected to develop first.

Virgin Media O2 Will Run HFC Network Side by Side with the New FTTH Network, in a First

In a first, Virgin Media O2 plans to operate its gigabit hybrid fiber coax network side by side with the coming fiber to home network. That has never been done before at scale, if ever, by a cable TV company. 

“We don't intend to shut down the HFC network,” Fries says. “ In fact, it wouldn't be surprising to me if the HFC network were operating for quite some time in a sort of a dual mode because we don't want to force migration to fiber.”

That sort of flies in the face of the argument that running a single network costs less than running two at the same time. That suggests operating cost savings will not be a significant part of the upgrade of the whole network to FTTH. 

“In our case, we'll migrate people to fiber who want 1, 2, 3, 10-gig,” says Fries. “But if you're happy with your 500-meg or your 250-meg, we may or may not incur the cost of migrating a customer from HFC to FTTH. 

“We continue to see cost per premise decline to below GBP 600 (USD 816) per premise as we use more of BT's passive infrastructure, and that only improves the significant return on capital we're already seeing from this investment,” says Liberty Global CEO Mike Fries. “We are accelerating our Lightning build program in 2022 from around 330,000 homes last year to over 500,000 homes this year.”

“Second, we completed the 50,000 home fiber-to-the-premise trial that supports our previously announced intentions to upgrade our entire 15 million homes to fiber by 2028,” says Fries. That build confirms Virgin Media O2 assumptions about build cost. 

“VMO2 will commit to be an exclusive anchor tenet of the network” while also opening up the network on a wholesale basis to other internet service providers, he says. 

In practice, since the whole VMO2 HFC network runs at gigabit downstream speeds, the only customers that would choose to migrate are those that want higher upstream speeds or “faster than gig” downstream speeds.

Mobility is the New Battlefield in War Beween U.S. Telcos and Cable TV Operators; Home Broadband the Current Issue

The U.S. cable industry has fared relatively better in the competition with telcos for at least some logical reasons, aside from arguable industry culture advantages. In the competition for fixed network voice and mobile services, cable has been the attacker, starting with zero market share and winning accounts from the incumbents. 

In home broadband, cable has benefitted from the relative slowness of telco adoption of fiber-to-home platforms. Simply, cable has leveraged its lower-cost hybrid fiber coax platform to radically boost performance at much-lower cost than telcos have been able to do. 

In video services, where cable was the incumbent, there was a slow attrition of some market share to telcos. The big share held by non-cable providers was in the hands of the satellite providers, and AT&T at one point owned all of DirecTV, immediately making AT&T one of the largest providers of linear video subscriptions in the U.S. market. 

That stake in DirecTV has been spun out to private equity group TPG, but AT&T retains a 70-percent interest in the venture. 

Still, terrestrial fixed network market share held by telcos was relatively small. 

source: Cable Compare 

In recent decades, competition between telcos and cable has shifted. Video share has remained relatively constant, with the overall market gradually shrinking. The fixed network voice market is declining, for all leading providers. 

Home broadband has emerged as the revenue growth driver. And while cable has held close to 70 percent of the installed base in that market, many observers--perhaps most--now expect telcos to take more share as FTTH becomes more common. 

Nationally, telcos have about 30 percent share of the home broadband installed base. The issue is how much additional share telcos can gain as they ramp up FTTH platforms and as 5G fixed wireless becomes a factor for Verizon and T-Mobile. 

All that likely means that mobility will become the new growth battlefield between cable and telco.

Does More Privacy Mean Fewer Ads?

One way of the other, users are going to have more privacy and control over their data where it comes to its use for advertising. But that almost assuredly does not mean there will be less advertising. There could plausibly be more advertising, simply because less-targeted ads mean "more waste."

To make up for the less-effective targeting, advertisers will have to spend more.

It is probably the case that most people, most of the time, do not prefer to see or hear advertising. Which is one way of saying they do not like advertising. 

Of course, many argue what people hate is bad advertising. Some surveys suggest 70 percent of consumers globally tolerate ads, including about a quarter who say ads can be enjoyable. 

source: Marketing Week 

 

But that is likely a minority opinion, sometimes advanced by marketing professionals, not consumers who are exposed to the ads. These days, people tend to view ads as intrusive, too numerous and an infringement on privacy. 

source: Stepan Stroe 

It’s more than that: ads are viewed as intrusive, in much the same way that  many people also do not appreciate unexpected phone calls, which likewise are viewed as intrusive. 

And yet advertising is a big business of perhaps $300 billion in annual revenue in the United States alone. But grumbling about targeted advertising seems to be growing in some quarters. The European Parliament is looking at legislation outlawing targeted ads using religious beliefs, sexual orientation and racial or ethnic origin.  

Facebook already does so. And Google is ending behavioral targeting. That means cross-site tracking will not happen, outside properties Google itself owns. 

Perhaps more important, many in the internet ecosystem--especially Apple--are moving to give consumers more control over ad tracking data. 

Apple will allow its users to opt in or out whenever an app wants to track a user’s activity across other companies’ apps and websites. 

source: Medium

And the trend has been underway for some time. Restricting use of third-party cookies provides one example. 

One way or the other, targeted advertising is going to change. Which also means advertising is going to change. But most of us likely do not believe there will be less advertising. It might be more private. 

Ads might be less effective. But who believes there will be fewer ads? Probably nobody. Not so long as advertising enables access to “free content” that people want. And not so long as firms still need to sell their products. 

source: Statista

Video Drives Most Bandwidth Consumption, Mobile Phones the Most Connected Devices

Some really-important trends might not be entirely obvious in this description of a report about how customers used Wi-Fi on the Comcast network. 

You might think the biggest use would be video streaming, and that is likely true in terms of bandwidth consumption, though the report is organized around “numbers of connected devices” not “bandwidth consumption.”

You might be somewhat surprised that personal computers do not even show up as a category worth quantifying. It is not so much home broadband as a platform for work, homework or economic growth is unimportant. PC use for work or learning is important. 

But such use cases do not dominate bandwidth consumption, which is led by gaming, TV and other entertainment purposes. 

source: Comcast  

Bandwidth consumption is another matter, as global bandwidth consumption is dominated by video, including entertainment video and video clips on social media and for advertising. 

source: Comscore

source: Comscore

source: NCTA  

If Digital Transformation Does Not Create Outcomes, Why Bother?

IDC has predicted that 65 percent of global gross domestic product will be “digitalized by 2022.”  IDC also predicts that by 2023, 75 percent of organizations will have comprehensive digital transformation (DX) implementation roadmaps.

We might disagree about precisely what “digitalization” means without disagreeing that it ultimately must result in outcomes. It is one thing to change a business process; it is quite another to change a business outcome. So it is with any applied effort involving digital transformation. The objective is always to change business outcomes. 

source: AWS 

So here is the point: DX is important when it allows entities and organizations to achieve more. 

source: McKinsey 

By 2025, 75 percent of business leaders will leverage digital platforms and ecosystem capabilities to adapt their value chains to new markets, industries, and ecosystems, argues IDC. 

Note the key phrase: “New markets, industries, and ecosystems.” New; not existing or incumbent or legacy. 

That is not to say almost every business process will be affected; obviously that seems the case. But digitized processes do not help much if they do not also support the move into new markets, industries and ecosystems. 

“In some cases, the use of digital tools and technol­ogies can upend entire business models or create entirely new businesses,” McKinsey consultants say. “Look no further than the way the internet has changed the way consumers research and purchase airline tickets and hotel rooms, disintermediating many traditional travel agents—one of the original cases of industry reinvention,” McKinsey has said.

Or look at the way  video-streaming services has disrupted the economics of traditional broadcast and cable TV channels. 

Consider the way cloud computing has changed “how we do computing” but also has disrupted and changed the computing hardware and software businesses. Or look at the way enterprises, businesses and consumers now use computer applications and services: remote and distributed applications are the rule. 

source: Deloitte 

Our apps are stored “someplace else” and accessed using the internet when we need to invoke them. 

We used to say that “digitization” allowed any media type to be transported over a single connectivity medium. It also has proven true that the internet allows almost any application or process to be supported by remote computing and storage facilities. 

We have separated “where computing is used” from “where computing happens.” But there is more to it than that. Some might argue that DX is just what we used to call “information technology.” 

It arguably is much more. According to IDC, DX is the way to achieve business advantage. We often have proxies for advantage. Sometimes we say DX is valuable because it increases resiliency, or provides agility. DX-enabled firms can survive crises better, and innovate faster. All of that really matters if revenue is at stake; if survival is at stake. 

To be sure, customer experience, product development, fulfillment, marketing, human resources and all other business processes are potentially affected by DX. But it is the ability to protect the legacy financial returns and potentially accelerate revenue growth in new areas that is the foremost outcome of digital transformation. 

source: McKinsey 

IOT and Blockchain, Plus AI, Drive Automation, Lower Costs and Cost of Sales

Distributed or automated ordering and fulfillment, plue bandwidth or services on demand, have been desired capabiltiies in the connectivity business for many decades. Blockchain adds automated settlements to the mix. 

Internet of things adds new requirements for low-cost, secure, automated ordering, fulfilment and settlement. Blockchain should help. 

"Digging into the Future"

Highlights of season one. 

Claude Shannon and TeraHertz Spectrum

Claude Shannon’s information theory is the reason we are moving to millimeter and teraHertz frequencies in the mobility business, as well as higher optical frequencies for fixed network communications.

In fixed network optical networks, the higher the carrier’s frequency, the greater the channel bandwidth channel bandwidth and the higher the information-carrying capacity of the system.
The rule-of-thumb for estimating the transmission bandwidth is that its maximum value is approximately 10% of the carrier frequency. 

Shannon defined the maximum data capacity over a communications channel in the presence of noise.

source: CableLabs 

Basically, capacity requirements are a linear function of the bandwidth available. Shannon’s model shows that as the available bandwidth in the frequency range increases, so does the capacity of the channel. All else being equal, a doubling in frequency can double the channel capacity.

Basically, as frequency increases, wavelength decreases. As wavelength decreases, there are more opportunities to code symbols per second.

Basically, if a mobile signal in the 800 MHz to 2 GHz range has a wavelength between one meter and 10 centimeters, a radio signal in the 6-GHz to 30-GHz range has a wavelength in the 5 cm to 1 cm range. 

source: Researchgate 

That shapes potential bandwidth because there is a direct relationship between frequency and capacity. All other things being equal, signals in the above-6-GHZ range can carry 10 times the information of signals in the less-than-2-GHZ range. 

Since end user bandwidth consumption in the wireless space grows over time, as does consumption over fixed networks, there is a continual need to add spectrum (capacity) over time. 

Aside from the fact that lower-frequency bands already are occupied by other users, 10-times increases in capacity (again, all other things being equal) can be gotten only by adding new frequencies higher in the spectrum. 

This is relatively easy to visualize. This chart shows the traditional mobile communications spectrum in light blue, compared to other bands used for point-to-point (non-mobile communications). The orange bar supports Wi-Fi and some other uses. 

source: Researchgate 

The point is that above 38 GHz, most of the spectrum is commercially unused. So as bandwidth demand increases, we will make more extensive use of the sub-40-GHz spectrum, and then begin adding frequencies higher than that. 

So Claude Shannon’s information theory explains why we will move into millimeter and teraHertz radio frequency ranges to support future mobile network generations. Since there is no way to reassign most of the lower-frequency assets to mobility, unused bands must be sought. 

And most of those resources are in the millimeter and teraHertz regions. 

Or, if you like, uncertainty plus information volume is the driver of the move upwards in frequency.  It is the combination of the minimum number of bits required  to represent any bit of information and the amount of information to be coded that fuels the move to millimeter and teraHertz frequencies, as challenging as that might be, in terms of building commercial wireless networks.

Mobility and Internet Access Buying is Similar to Consumer for Micro-Sized Organizations

There is a good reason why telco small business marketing often includes those organizations in a “mass markets” segment. “Micro” businesses with no more than nine employees often buy consumer mobile service plans, for example.

“Small” businesses with 10 to 49 employees more commonly buy business plans of some sort. “Medium” businesses with 50 to 249 employees buy business plans about 85 percent of the time, according to an Analysys Mason estimate. 

source: Analysys Mason 

Roughly the same might be noted for other products such as internet access. 

Perhaps the single biggest difference is that small- and medium-sized organizations are more likely to require dedicated internet access with assurances about bandwidth. Micro-sized organizations are much more likely to buy “best effort” consumer grade services.