Twilio Microvisor Supports IoT

With the caveat that it typically is easier for an application provider to add new features and create new platforms, compared to a connectivity provider, Twilio’s new Microvisor platform shows how an enabler of voice, text, chat, video and email services in any application is migrating towards support for internet of things. 

The move illustrates how a communications enabler adapts its current business to developing new markets, much as connectivity providers hope to add support for internet of things and edge computing. 

Microvisor is said by Twilio to be “an IoT connectivity and device management platform that offers embedded developers a one stop shop for building connected devices, keeping them secure, and managing them through their lifetime.”

As was the case with the original Twilio platform that basically would voice-enable any application, the Microvisor platform abstracts away many of the common infrastructure challenges burdening IoT developers, the company says.

That includes functions such as security, debugging and updates and language and embedded operating system support, including hardware such as the Cortex-M microprocessors and select STMicroelectronics chips.

Industrial IoT Hinges on Partnerships, Vendors Say

5G Private Networks Often Will Connect to Edge Computing Facilities

The possible or predicted use of private 5G networks will likely often involve edge computing, as the purpose of the private 5G will be to support networks of sensors, even if many other deployments are designed simply to provide indoor mobile phone connectivity. 

Factory or seaport automation use cases provide examples of the former. Sports venues provide examples of the latter. 

That is not to deny many early use cases where 5G is used to provide a fixed wireless access service competing with fixed network services. That is an obvious use case with clear demand drivers and a business case. For most consumer phone customers, 5G will simply mean “faster mobile internet access.”

You can see that in this forecast of edge footprint by Tolaga Research, which shows significant use cases for mobile internet access and fixed wireless. 

source: Tolaga Research 

But most of the innovative use cases will revolve around the internet of things and ultra-low-latency use cases where edge computing is a necessary or helpful companion technology. Many IoT use cases will involve monitoring of physical processes, where the raw data collected will benefit from local processing. 

Many of the ultra-low-latency use cases will likewise benefit from local processing, matching the ultra-low-latency local or inside-the-building communications links. 

source: STL Partners 

Researchers at Real Wireless, for example, say up to 75 percent of enterprise small cell deployments will be installed by third parties, not mobile operators, for example. In many cases, the devices communicating over those small cell networks will be sensors, connected to some edge server that provides alarms, for example. 

source: Semiconductor Engineering 

Underwater Data Center Server Reliability 8 Times Greater than in Traditional Data Center

Server failure rates are about eight times lower in an underwater data center that had no human touch for two years. Enclosed in a capsule about the size of a shipping container, the data center was submerged in seawater and contained 12 racks and 864 servers.

The Microsoft research team hypothesizes that the atmosphere of nitrogen, which is less corrosive than oxygen, and the absence of people to bump and jostle components, are the primary reasons for the difference in reliability.

In addition to possibly being able to apply the lessons to land-based data centers, the test also provides an idea of ways to support edge computing facilities. 

Verizon Deploys Edge Computing in 3 Additional Cities

Verizon has deployed AWS Wavelength edge computing in three new locations: Atlanta, New York and Washington, DC. Verizon and AWS launched the mobile edge computing (MEC) platform in August 2020 in Boston and the Bay Area.

Verizon plans to add five more cities by the end of 2020. 

It might be worth noting that the hoped-for upside from multi-access edge computing that marries 5G and other access networks to local computing facilities might be more limited, at least initially, than many expect. Edge computing still is “computing as a service.” 

STL Partners, for example, says the range of forecasts for MEC revenue ranges from a low of less than $2 billion to a high of perhaps $4 billion globally in 2025. That is a small figure, on a global basis. 

source: STL Partners 

The supplier of the computing as a service in this case is Amazon Web Services, not Verizon. Verizon will supply 5G access and real estate (space, cooling, power, security), to be sure. There is some revenue upside. 

But as with the internet of things, most of the revenue will be earned in other parts of the ecosystem. In this forecast by Yandex, about four percent to five percent of total IoT revenue will go to suppliers of connectivity. 

source: Yandex 

That is helpful, to be sure. But it is not going to make a huge difference for most service providers, mobile or fixed.

Edge Computing Drives Nvidia Interest in ARM

Nvidia's interest in buying ARM has to do with the expected growth of edge computing. Traditionally, Nvidia's graphic processors have been used in centralized cloud data centers. But edge computing will shift many processing chores to devices or premises-based servers or close to the network edge, where low power draw and small footprint are important.

Graphics processing unit improvement seems to be occurring faster than Moore’s Law would predict. Nvidia’s GPUs today are 25 times faster than five years ago. If they were advancing according to Moore’s Law,  they would have increased their speed only by a factor of 10, says Nvidia CEO Jensen Huang.

Between November 2012 and May 2020, Nvidia’s chip performance increased 317 times for a major class of AI computing, says Bill Dally, chief scientist and senior vice president of research at Nvidia. On average, the performance of these chips has more than doubled each year.

Enterprises Continue Shift to Multi-Cloud, Edge Computing Volume Still Does Not Register

Fully 93 percent of enterprises have a multi-cloud, a survey of 750 global cloud decision-makers and users conducted on behalf of Flexera finds. That appears to up from 2018 levels. As some note, edge computing is a coming trend, but does not yet approach these sorts of adoption rates, even if forecasts are robust.

source: Zion Market Research

source: Flexera

source: Rightscale

Some 87 percent have a hybrid cloud strategy and 33 percent of all participating organizations use multi-cloud management tools. Respondents use an average of 2.2 public and 2.2 private clouds. 

source: Flexera

About 20 percent of enterprises spend more than $12 million per year on public clouds, the study finds. More than 50 percent of enterprise workloads and data are expected to be in a public cloud within 12 months. 

source: Flexera

The top three public cloud providers remain AWS, Azure and Google, the study finds. Azure is narrowing the gap with AWS in both the percentage of enterprises using it and the number of virtual machines (VMs) enterprises are running on those services.

source: Flexera

Some 40 percent of enterprise AWS users spend at least $1.2 million annually, compared to 36 percent for Azure. Google experienced the fastest growth in adoption since last year’s survey, Flexera says.

Cloud Computing Does Not Always Offer the Best Total Cost of Ownership

As useful as cloud computing is for testing new applications and use cases, entering new geographies or serving new customer segments and supporting relatively lower volume applications, cloud repatriation--shifting workloads back to private clouds, on-premises data centers or hybrid approaches--makes sense as workload volumes grow.

As often is the case, per-user charging mechanisms work very well at low volumes, often saving capital investment or opex costs. In high volume, the economics often reverse. 

In other words, total cost of ownership--when an app supports high volume--often is lower when supported by on-premises data centers, rather than using a cloud service provider. 

At low volumes of use, renting often makes more sense. That seems to apply for enterprises looking at edge computing, according to a Mobile Experts analysis. Over three years, “it costs about 35 percent to 55 percent more to access a hyperscaler's cloud for a heavy industrial workload at the edge,” Mobile Experts says.

“Unified communications as a service” provides a simple example of the total cost of ownership. One analysis, for example, In a 100-user situation, the total cost of ownership over a five-year period favors a premises deployment over either private cloud or public cloud, according to TTx.

The economics favor on-premises support. 

• On-premise UC: $220-$240 per user per year

• Public cloud UCaaS: $360-$480 per user per year

• Private cloud UCaaS: $250-$350 per user per year

That includes upfront capital investment and recurring costs. A rough estimate of the cost per user for the first year for each system would be:

• On-premise UC: $70,000-$90,0000 for hardware/infrastructure, software licenses, endpoints, and installation/deployment support.

• Public cloud solution: In a 100 user environment, in most cases the upfront costs are $0.

• Private cloud solution: Highly variable but declining

On-premise UC represents costs between$120 to $145 per user, per year for ongoing maintenance, cost of upgrades and public network access. 

Public cloud UCaaS has costs between $360 to $480 per user per year which includes the license cost per month and included deployment support. (This assumes an average of $30-$40 per user for licenses which includes renting the endpoints)

Private cloud UCaaS has recurring expenses between $250 to $350 per user per year for financing and 3rd party maintenance. 

Cloud workloads often are the best choice. But not always.

Can MEC Platform Business be Built?

Some telcos are optimistic about prospects for multi-access edge computing, including the opportunity to function as a platform for edge computing. The MobiledgeX initiative is a good example of that belief. MobiledgeX hopes to build an ecosystem of developers able to use common global interfaces, telco data centers and integrated 5G access. 

source: MobiledgeX

The MEC is envisioned as being complementary to the current hyper-scale computing as a service platforms such as AWS Wavelength and Azure Edge Zone. The issue, as always, is how much value actually can be created by a telco-owned edge computing platform or even actual edge computing as a service. 

Already, hyperscalers are moving to create their own edge computing as a service offers. In those instances, telcos become providers of real estate services and connectivity, but not the platform for developers or actual edge computing as a service.

It never is easy for any participant in an ecosystem to emerge in additional roles, and has historically been challenging for telcos. Edge computing might not prove much different.