Traffic Between Data Centers Drives Nearly Half of Global WAN Traffic

Bandwidth consumed by all consumer and business users is just a slight bit bigger than the traffic flowing between data centers these days. And data center traffic headed for other data centers clearly drives international WAN traffic already. 

Edge Gets the Hype, But Cloud Data Centers Get the Money

Even as preparation for edge computing gets headlines, investment activity remains focused on hyperscale and regional data center investment by firms that formerly would have invested in other forms of real estate.

“Demand for data centers by global investors remained robust in the second half of 2021, with record-breaking merger and acquisition activity,” according to CBRE. “Nearly 95 percent of respondents to CBRE’s 2022 Global Data Center Investor Sentiment Survey, many of whom are the world’s largest institutional real estate investors, plan to increase their capital deployment in the data center sector,” says CBRE. 

In North America alone, transactions included:

• American Tower’s acquisition of CoreSite for $10.1 billion.

• Blackstone’s acquisition of QTS for $10 billion.

• Cyxtera’s SPAC merger with Starboard Value Acquisition Corp, valued at $3.4 billion.

• Mapletree’s acquisition of the Silas Realty Trust Portfolio of 29 data centers for $1.3 billion.

• Prudential & Digital Realty’s joint venture sale of 10 powered shell data centers for $581 million.

• DigitalBridge’s Vantage SDC acquisition of CA22, a 24 MW hyperscale data center, for $539 million.

Already in 2022, North American transactions include:

• KKR & Global Infrastructure Partners acquisition of CyrusOne for $15 billion.

• DataBank’s acquisition of four CyrusOne data centers in Houston for $670 million.

In the Europe Middle East Africa market, 2021 transactions include:

• IPI Partners acquired the dominant Nordic operator Digiplex.

• Iron Mountain acquired a turnkey facility in Frankfurt from Keppel Data Centers for €76 million.

• Antin Infrastructure Partners acquired leading U.K. managed-services provider Pulsant.

• Azrieli purchased colocation operator Green Mountain in Norway for 7.6 billion NOK.

• Blackstone acquired a triple-net facility leased to Equinix in London Docklands for £196.5 million.

• Keppel DC REIT acquired two triple-net facilities, one in the Netherlands and the other in the U.K. for a combined total of more than €100 million.

• Digital9 Infrastructure acquired Verne Global, a dominant Icelandic colocation provider for £231 million.

• AtlasEdge, the recently formed edge operator backed by Liberty Global, Digital Bridge and Digital Realty, acquired a portfolio of 12 assets from Colt Data Centers.

In 2022, transactions in EMEA include:

• KAO Data, backed by Legal & General Capital, Goldacre and Infratil, have acquired two data centers in West London via a sale and partial leaseback.

• Digital Realty agreed to acquire a majority stake in Teraco, Africa's leading carrier-neutral colocation provider, from a consortium of investors, including Berkshire Partners and Permira, as well as Medallion Data Centres in Nigeria.

In the Asia-Pacific region, 2021 transactions include:

• Vantage Data Centers and lead investor DigitalBridge acquired Hong Kong-based PCCW and Agile Data Centers.

• GLP acquired a 50% stake in Songjiang Internet Data Centre in Shanghai.

• Digital Edge acquired five data centers in Japan for $230 million.

• Equinix entered the India market through the acquisition of GPX India.

Already in 2022, APAC transactions include: 

• Equinix $525 million joint venture with GIC in South Korea.

• Keppel Capital to close $1.1 billion data center fund.

• Mitsui announced intention to invest $2.7 billion to develop data centers in Japan.

• GLP announced plans to develop 900 MW of data centers across Tokyo and Osaka and totaling $12 billion.

In Latin America, 2021 transactions include:

• Piemonte Holding acquired Globo’s Data Center in Rio de Janeiro.

• Squared Capital acquired KIO Networks, a leading data center operator in Mexico.

• Goldman Sachs Asset Management invested in Piemonte Holdings' Brazilian edge data center platform Elea Digital.

• EllaLink and Equinix delivered the first undersea cable between Europe and Latin America.

In Latin America, 2022 transactions include:

• Equinix opened a hyperscale facility in São Paulo and announced plans for two more in Mexico City and another in São Paulo.

• Ascenty launched two data centers in Rio de Janeiro and Hortolândia.

• DigitalBridge-owned Scala Data Centers began construction on two hyperscale data centers in Brazil, with one fully leased to a cloud provider.

• Tencent Cloud launched its first Brazilian data center.

• Ascenty plans to build its fourth data center in São Paulo.

• Microsoft announced plans to establish a Chilean data center region.

Edge Computing and Blockchain-Based Disintermediation

Edge computing is in some ways a complement to cloud computing, in some ways an alternative. To some extent, edge computing is a form of disintermediation, the removal of steps or stages in any supply chain. 

Disintermediation has reshaped computing and communications in the past. Could blockchain launch a new wave of change? Perhaps, many believe. 

Disintermediation is at the heart of the promise and period of a wide range of potentially-huge new enablers of business and economic life, including blockchain, cryptocurrencies, distributed finance and distributed autonomous organizations. 

Disintermediation is the process of removing  “middle men” or distributors, steps or layers or retailers from a supply chain. 

For that reason, disintermediation has been a key business issue for all retailers and distributors since the internet emerged, reducing the roles or value of distributors of all types and increasing the ability of suppliers to go straight to their customers for fulfillment of sales. 

Disintermediation also has affected the connectivity business. Consider only the role of applications in creating messaging apps that directly displace carrier text messaging and voice services. Email had a similar impact prior to the emergence of messaging. 

Now 5G private networks are being evaluated and deployed to replace some parts of the public mobile network. 

source: WallStreetMojo 

Blockchain is a distributed ledger (database) system where details of transactions are stored across a distributed network of servers with no centralized repository of those transactions. The chain of records are indelible: once created they cannot be changed, which is how the integrity  of the information on transactions is maintained. 

Blockchains store data in blocks that are then linked together in series, which is where the name comes from. For that reason a sort of time stamp is created, as the chains form sequentially. 

As new data comes in, it is recorded in part of a fresh block. Once the block is filled with data, it is chained onto the previous block.\

Decentralized finance (DeFi) is an emerging financial technology based on blockchain that allows parties to conduct transactions directly, without the use of an intermediary bank, broker or institution.

A distributed autonomous organization is an online-only group of people who have entered into a contract with one another to reach a coordinated goal. The contract is enforced by the use of blockchain, so that if any member who violates the terms of the contract, their access to owned tokens is revoked automatically by the blockchain itself. 

The promise for users and peril for incumbents is clear enough. As always, disintermediation flattens hierarchies; removes middlemen from a value chain; reduces costs and enables more-direct transactions between buyers and sellers; creators and users. 

What blockchain, DeFi or DAOs could affect, in the connectivity and computing business, is settlements, transaction clearing, ordering and fulfillment, the way sales are handled, bad debt protection and fraud prevention. 

And that should mean the potential for lower operating costs and higher potential margins, partly because sales effort is automated, partly because cost of sales could drop, partly because bad debt is avoided, partly because cash float is improved. 

Other possibilities seem less likely, in large part because vested government interests are involved. DAOs have been used to raise large pools of money to buy assets, for example. Would any regulator ever allow a perceived national connectivity asset to be purchased by a DAO?

Could “service providers” emerge that use DeFi to allow customers to purchase access to computing or connectivity capabilities without directly engaging a facilities-based provider on a longer-term basis? 

Organizations already buy compute cycles “as a service,” typically with a quota of cycles purchased upfront. Could DeFi allow purchasing on a full on-demand basis, with no “bucket of usage?”

Disintermediation is a powerful process that reshapes industries and creates winners and losers. It virtually always leads to lower prices and efficiencies in supply chains. 

We have already seen a first wave of disintermediation in the connectivity and computing businesses. Will we see additional waves as blockchain-based business processes arrive? Probably. But we will see the first impact in settlements, then in sales. 

The other possible changes will involve broader changes that involve additional stakeholders, so would take longer, if they happen at all.

"Narrow AI" is What Most People Will Encounter in Life

Whether we encounter artificial intelligence as a user of apps, services or products, we increasingly will be using AI in the future. Very rarely, if ever, will we actually be buying “AI” as a product, though. It will simply be part of some other things we do buy. 

Which is to say we will normally encounter AI as “weak AI,”  some form of machine learning that is optimized for a particular purpose, such as speech recognition, search results, recommendations,  facial recognition or some customer service app. 

This form of AI will be wildly valuable and ubiquitous, early on. 

source: Toolbox 

What we will not often actually use or see is “strong AI” (“ deep AI”),  when machines apply their intelligence to solve complex problems. Still less are we at a stage when “artificial superintelligence” is possible, allowing machines to surpass human intelligence in terms of completing tasks. 

Weak AI will prove very useful as a way of detecting patterns in human or machine behavior that allow decisions to be made. When people use voice assistants such as Siri, they already are engaging in a process supported by weak AI. 

Almost no consumer or business user will ever buy “AI” in a direct sense, as a product. It will be a feature of a product. 

How Big is the Edge 5G or 4G Private Networks Market?

It should come as no surprise that some forecasts of private 4G and private 5G networks are robust: there are firms that sell infrastructure, services or software supporting such networks. Kaleigo Intelligence forecasts about $7 billion in non-public private networks using either 4G or 5G infrastructure by 2026. 

source: Kaleido Intelligence  

That market will be especially important for a segment of the networking infrastructure industry able to support the building and operating of such networks in a variety of settings where assured bandwidth and latency performance could improve on what is possible using Wi-Fi, for example. 

source: Kore Wireless 

That noted, private 5G networks might wind up being a niche that is quite important for specialists, but not so significant for some possible suppliers. A good chunk of that $7 billion will be spent on infrastructure to build the networks. Another good amount will be spent on system integration, consulting and other costs such as spectrum rights. 

In some cases revenue will flow to mobile service providers who build such networks on behalf of customers. But other system integrators might well emerge as more-important providers, especially since application support will matter and many large system integrators have existing practices that support customers in vertical industries. 

Ability to engineer and build the actual private network is important, but secondary to the task of ensuring the key applications work. And that tends to be a domain better suited for larger integrators with historic relationships with enterprise customers at the application level.

 

No Autonomous Vehciles Without AI, Edge Computing, Vision and Mobile Communications?

Autonomous driving is virtually impossible to imagine with use of edge computing, artificial intelligence, vision systems and mobile communications. Just as certainly, the movement from assisted driving to full autonomy involves increasing application of those core technologies. 

Some categorize autonomous driving in five levels. The first two levels assist drivers in some way. 

At level three the driver must be in the vehicle but is not required to monitor the vehicle at all times. This is known as conditional automation. 

At level four, the vehicle can perform all driving functions under certain conditions. A driver is still required, although the vehicle can travel from point A to point B without intervention. 

At level five, a driver is no longer required. The vehicle performs all driving tasks, even eliminating the need for a steering wheel or gas and brake pedals.

 

source: Vision Systems Design 

Nvidia CEO Jensen Huang on AI, Computing, Graphics, Edge

Jensen Huang, Nvidia CEO on most of the things that matter for artificial intelligence, data science, high performance computing, graphics, edge computing, networking, and autonomous machines.

AMD and Nvidia Battle Over Graphics Processor Performance

There are implications for edge computing, artificial intelligence, extended reality use cases. 

Applied Edge Computing to Save Whales

Whale Safe, a project of the nonprofit Benioff Ocean Initiative, is deploying buoys with acoustical sensors using on-board artificial intelligence--an application of edge computing--in key whale habitat off the coast of southern California. 

The analysis of actual whale locations, correlated with movements of ships in the channel, helps avoid whale strikes by the ships moving in one of the most-trafficked sea lanes globally. 

source: Whalesafe 

Ships are asked to slow down when whales are located in their path, since collisions between ships and whales are often fatal. 

Since deployment of the system in 2020, whale strikes declined significantly in the Santa Barbara Channel, which is one of the most active feeding grounds for humpback and endangered blue whales on the West Coast.

Will IoT Workloads Shape Edge Computing?

Internet of things workloads are likely to be highly disparate, complicating efforts to predict usage requirements, even if the consensus is that IoT workloads will grow as a percentage of total. 

But IoT at scale does raise some issues, such as whether computing workloads will change in some way from current patterns.  

It might be quite reasonable to assume that local processing to support artificial intelligence will be more important than edge storage. But bandwidth and remote computation arguably will be more important for complex, non-real-time analysis.   

source: Analyst Point of View 

Will Edge Computing Prove to be More Sustainable than Cloud Computing?

Edge computing is coming for any number of reasons, including some we do not necessarily and routinely tout, such as support for compute-intensive operations such as artificial reality and other forms of extended reality; real-time process control or any applications that require ultra-low latency. 

Consider power efficiency and footprint. 

“Even with 5G right now, which is a little more power-efficient and has a lot more programmability of the network, from a sustainability point of view, it is impossible to continue transmitting everything to the cloud,” says Stacey Shulman, Intel VP, network and edge computing. 

The actual benefit remains to be verified. Some argue that hyperscale computing facilities and instances are inherently more efficient than small distributed instances. But that only accounts for power consumption per million cycles, and not the full footprint. 

source: Benjamin Davy 

Also, some use cases might require local computing, and the footprint simply must be optimized. The cost of a computing instance or its environmental impact is arguably always secondary to the value of the function: keeping cars from crashing and injuring people, for example, when operating autonomously. 

And different processing tasks consumer different amounts of power.  Still, it has been argued that power consumption does not scale linearly with the volume of computing instances.  

If one considers applications that require the use of artificial intelligence in real time, “here’s a lot of data that can’t go straight to the cloud,” she notes. 

Some analyses already suggest that bandwidth cost savings are a benefit of edge processing. Power footprint might well wind up as among the key benefits of edge computing, Shulman argues. 

Some argue edge computing should be more efficient. But the total impact has to include operating the communications infrastructure, adjust for the types of workloads and utilization of server resources, for example.  

In the past, precisely the opposite argument has been made. AWS has argued that 

remote cloud computing is more efficient, and has a smaller footprint, than “premises” computing at the edge. 

As always with identifying the precise impact of any activity or product, the analysis is complicated and requires many assumptions. 

source: Benjamin Davy 

source: Benjamin Davy

So far, modeling has focused on computing footprint at hyperscale data centers. We do not have enough data, yet, on edge computing instances, except to the extent that we can assume minimum and maximum power consumption for a single physical server or virtualized workloads run on a shared server. 

The full footprint then will hinge on other inputs. Some might also include any substitution effects, where using edge computing reduces the footprint of other activities that are displaced or lessened. 

source: Ericsson 

The point is that we still are not sure how edge computing will wind up, when compared to remote cloud computing,  in terms of carbon and other greenhouse gas footprint.

Google for Games Developer Summit 2022 Keynote

It takes little imagination or experience for us end users to believe that edge computing will be important for gaming, as will virtual reality, given the importance of latency performance. 

Connectivity Providers Will Get 10% of Edge Computing Revenues

No matter where one looks--private 5G networking; internet of things or edge computing--it seems observers believe system integrators, hyperscale cloud computing suppliers, solution specialists and infrastructure providers are positioned to reap about 90 percent of revenue in those businesses. 

Mobile platforms, for example,  are being leveraged for private networking. Non-carrier 5G and networking offerings are gaining momentum, says Dan Bieler, Forrester principal analyst. By about 2025, perhaps 90 percent of private 5G revenue will be earned by hyperscalers, system integrators and solution providers, not telcos, predicts TBR. 

source: TBR 

“All major hyperscalers showcased (at Mobile World Congress) how they are getting more heavily involved in the networking  arena,” says Bieler. 

Google is working on network slicing on Android 12 for providing access to Google Cloud. 

Microsoft’s Azure for operators is a 5G overlay on Azure cloud WAN where Microsoft’s internet backbone carries the customers’ traffic.

Amazon Web Services is offering  private 5G solutions and cloud WAN and edge application offerings. 

“A new threat for telcos is on the horizon,” says Bieler. Importantly, “the lines are blurring between cloud computing and networking.”

No matter the stumbles telcos routinely have had in attempting to diversify their business models, or wring more value out of their positions in value chains, such movement remains necessary, if exceedingly difficult. 

Estimates for the total 5G opportunity range from $4 trillion to $6 trillion by 2030, says Dan Bieler, Forrester principal analyst. But the connectivity part accounts for only five percent to 10 percent of this opportunity. 

You can see the pattern in forecasts of internet of things revenue in Africa and the Middle East. 

source: GlobalData 

Or look at “smart city”  IoT revenue by platform.

source: ABI Research 

In one sense, that should come as no surprise. Connectivity typically represents about 10 percent of market opportunity for any network-based product. The rest is devices, operating systems, platforms, enablement and applications. 

To gain more than about 10 percent of the revenue upside, connectivity providers have to move outside that role. That never has proven easy, or successful.

Here We Go Again: Hyperscalers Will Displace Telcos

One big strategic change in the era of distributed, competitive, virtualized computing and networking is that the demarcation between industry participants is becoming more porous. That is to say, formerly-distinct roles are becoming less distinct. 

“All major hyperscalers showcased (at Mobile World Congress) how they are getting more heavily involved in the networking  arena,” says Dan Bieler, Forrester principal analyst. 

Google is working on network slicing on Android 12 for providing access to Google Cloud. 

Microsoft’s Azure for operators is a 5G overlay on Azure cloud WAN where Microsoft’s internet backbone carries the customers’ traffic.

Amazon Web Services is offering  private 5G solutions and cloud WAN and edge application offerings. 

“A new threat for telcos is on the horizon,” says Bieler. Importantly, “the lines are blurring between cloud computing and networking.”

To put it another way, firms that once were not direct competitors now are emerging in such roles. This is not a new theme in the connectivity business. Encroachment has been happening for close to 20 years. 

source: Cleverism, Michael Porter 

To be sure, product demand also has shifted. People prefer to use mobile networks and devices for voice and messaging. But voice over IP also has destroyed the gross revenue and profit margin of international calling. App messaging has replaced carrier-sold text messaging. 

source: Michael Porter, Cleverism

 

And now mobile platforms are being leveraged for private networking. Non-carrier 5G and networking offerings are gaining momentum, says Dan Bieler, Forrester principal analyst.