AWS enabling edge computing, supports mobile & IoT devices, 5G core network and new services
In an AWS re-Invent Leadership session titled “AWS Wherever You Need It,” [1.] Wayne Duso, vice president of engineering and product at AWS, expressed similar goals. “Today, customers want to use AWS services in a growing range of applications, operating wherever they want, whenever they require. And they’re striving to do so to deliver the best possible customer experience they can, regardless of where their customers or users happen to be located. One way AWS helps customers accomplish this is by bringing the AWS value to our regions, to metro areas, to on-premises, and to the furthest and evolving edge.”
Note 1. You can watch the 1 hour “AWS Wherever You Need It” session here (top right).
“We’re helping customers by providing the same experience from cloud to on-prem to the evolving edge, regardless of where your application may need to reside,” Duso explained. “AWS is enabling customers to use the same infrastructure, services and tools to accomplish that. And we do that by providing a continuum of consistent cloud-scale services that allow you to operate seamlessly across this range of environments.”
Duso explained how AWS is enabling edge computing by adding capabilities for mobile and IoT devices. “There are more than 14 billion smart devices in the world today. And it’s often in things we think about, like wristwatches, cameras, cellphones and speakers,” he said.
“But more often, it’s the stuff that you don’t see every day powering industries of all types and for all types of customers.” Duso cited the example of Hilcorp, a leading energy producer, which is using smart devices to monitor the health of its wells, optimize production and proactively predict failures so it can minimize capital expenditures.
With IoT devices becoming common among energy providers, edge computing is on the rise to handle the volume of data these devices generate. “Now, AWS IoT provides a deep and broad set of services and partner solutions to make it really simple to build secure, managed and scalable IoT applications,” Duso added.
Duso pointed to Couchbase as a use case for flexible AWS services: “Couchbase is a non-SQL database company that uses AWS hybrid edge services such as Local Zones, Wavelength, Outposts and the Snow Family to deploy its applications and highly scalable, reliable and performant environments to reduce latency by over 18 percent for its customers.” Each of these AWS managed services enables Couchbase to move data from the edge to the cloud or manage and process it where it’s generated.
“What we built on these AWS compute environments was a highly distributed, managed or self-managed database,” Duso explained. “For the cloud, an internet gateway for accessing that data securely over the web and synchronizing that data down to the edge. And that works across cloud, edge and on the offline, first-compute environments.”
“Our goal is I want to make AWS the best place to run 5G networks. That is the overarching objective. How can I make AWS, whether we are running it in the region, in a Local Zone, on an Outposts, on a Snow device, how do we make it the best place to run a 5G network, and then provide that infrastructure.”
AWS’ 5G network efforts include a cloud architecture that can support an operator’s 5G SA core network and applications, similar to what AWS is doing with greenfield U.S. wireless network operator Dish Network. Sidd Chenumolu, VP of technology development and network services at Dish Network, recently explained that the wireless carrier’s 5G core network was using three of AWS’ four public regions, was deployed in “multiple availability zones and almost all the Local Zones, but most were deployed with Nokia applications across AWS around the country.”
AWS is also working with Verizon to support a part of that carrier’s public MEC system. This includes use of AWS’ Outposts and Wavelengths, the latter of which AWS recently expanded in the United Kingdom with Vodafone.
Hofmeyr continued, “I think you have a spectrum (of different wireless carrier networks), from the total greenfields like what we did with Dish to the large tier-ones. The one thing that’s common across the board is the desire to modernize and become more cloud-like. That is common. Everyone wants that. Each one has a very unique job. There’s not one way that they all are executing in the same way. They’re taking this one workload and then building, so all of them are focusing on different workloads in the network and put it in the cloud.”
In conclusion Hofmeyr said, “I think all over the edge we find these use cases for which purpose-built systems were designed to handle that. And our goal is how do you make that available in the cloud.”
References:
https://reinvent.awsevents.com/leadership-sessions/
https://reinvent.awsevents.com/on-demand/?trk=www.google.com#leadership-sessions
https://www.sdxcentral.com/articles/analysis/aws-wants-to-be-the-best-place-for-5g-edge/2022/12/
https://biztechmagazine.com/article/2022/12/aws-reinvent-2022-harvesting-data-cloud-edge
3 thoughts on “AWS enabling edge computing, supports mobile & IoT devices, 5G core network and new services”
Comments are closed.
Amazon says: “AWS edge services deliver data processing, analysis, and storage close to your endpoints, allowing you to deploy APIs and tools to locations outside AWS data centers. Build high-performance applications that can process and store data close to where it’s generated, enabling ultra-low latency, intelligent, and real-time responsiveness.”
https://aws.amazon.com/edge/
Light Reading: Demand for edge computing is taking longer than expected to develop
According to two giant US technology companies – data center operator Equinix and 5G provider Verizon – demand for edge computing seems to have stalled.
Both companies have argued that they’re well positioned to provide speedy access to computing services via a geographically distributed network of data center locations. But both have also reported slower-than-expected demand for such edge services.
“Equinix is the leading edge provider for use cases today, as it can cover over 80% of the US population with 10 milliseconds of latency,” noted the financial analysts at Wells Fargo in a report to investors, citing their recent conversation with Equinix’s VP of business development Jim Poole.
However, the analysts argued that there’s not much demand for anything faster. “There are no material use cases today that require <10 milliseconds, which haven't necessitated Equinix to go deeper into the 'metro edge' (i.e. secondary or tertiary markets)," the analysts explained. "In other words, the challenge on moving deeper into the edge is not a technology issue, but rather an ecosystem one in which developers have not been able to create economic models around new use cases that require those types of low latencies. But Equinix is clearly watching closely and could potentially make a move deeper into secondary/tertiary markets if the opportunity presented itself."
Equinix is one of the world's biggest data center operators with more than 200 sites around the world.
Resetting expectations
On edge computing, Verizon executives have made similar comments.
"On the MEC [multi-access edge compute], what we are finding is demand is taking a little longer to go," explained Verizon Business CEO Sowmyanarayan Sampath during a recent investor event.
He added that Verizon now expects to generate MEC revenues starting sometime late next year and into 2024.
As noted by FierceTelecom, that timeline represents yet another delay for the operator. Verizon CEO Hans Vestberg said in 2020 that Verizon would generate meaningful revenues from MEC in 2022. Then, earlier this year, Verizon CFO Matt Ellis said the operator's public MEC business would generate "significant" revenues starting in 2023.
"So it's going to take a little longer," Sampath said.
Verizon, for its part, offers more than two dozen public edge computing locations scattered across the US via a deal with Amazon Web Services (AWS).
Of course, the lack of widespread demand for edge computing hasn't stopped vendors from hyping the technology. For example, AWS touted its edge computing offerings during its recent re:Invent trade show, according to SDxCentral.
And Ericsson – one of the world's biggest suppliers of 5G radios – recently wrote that edge computing ought to be a core component of future mixed reality (XR) services. "Without having a sufficiently densified network in terms of RAN [radio access network] and multi-access edge computing (MEC) with an optimized network configuration, wide-area and mobility-supporting XR will be difficult to deliver," Ericsson's Du Ho Kang wrote on the company's website, in outlining the potential networking implications of metaverse-style services.
https://www.lightreading.com/the-edge/demand-for-edge-computing-is-taking-longer-than-expected-to-develop/d/d-id/782116?
BofA is very bullish on edge computing. From a Dec 5, 2022 report to clients:
Key takeaways:
-5G rollout, new AI/ML capabilities, big data & autonomous vehicles/devices to drive strong demand for edge compute.
-Edge compute, in turn, will power a faster and more capable Internet, enabling new applications and driving Cloud spend.
-Potential beneficiaries in Internet group include AMZN, GOOG, META, UBER & LYFT. We also see new competitive risks.
………………………………………………………………………………………………………………..
1. An enabler of a faster and more capable Internet:
In addition to a multi-billion revenue driver for Cloud providers such as Amazon Web Services and Google Cloud, edge compute will be a key enabler of delivery networks and the Internet of things (IoT). Edge compute can lower latency times to very low milliseconds from seconds, and as well as significantly reduce data transfer costs.
Autonomous vehicles, drones and other devices will require low latency and edge compute capacity to optimize performance. We also see a big opportunity for new digital services built on edge compute such as real-time, in-store, mobile promotions, automated checkout, and local travel experience offerings.
2. Edge spend will boost Cloud infrastructure revenues:
Edge compute capabilities require both capex investment and services spend. IDC estimates edge compute spending, across Cloud software and services could grow from $176bn in 2022 to $274bn in 2025. For infrastructure Cloud specific spend, we estimate that spending for edge computing capabilities will grow from $58bn in 2022 to $188bn in 2028, growing to represent roughly 1/3 of total spend. Cloud companies are partnering with telecom companies to provide enhanced services.
3. Cloud, mobility, retail and even AR/VR are LT beneficiaries:
Edge compute will enable new/enhanced Internet services and help grow Cloud revenues. Beneficiaries in our Internet coverage group include Amazon, Google, Meta, Uber and Lyft. Both Amazon Web Services and Google Cloud will see incremental infrastructure services demand. Alphabet’s Waymo division is a leader in the autonomous vehicle (AV) sector and its vehicles will be enabled by edge computing, which will be critical for V2V (vehicle to vehicle) or V2I (vehicle to infrastructure) communication. Uber and Lyft (mobility networks) can benefit today from better services for users and drivers, but the big long-term cost savings could come from AVs. Meta’s AR and VR devices (including headsets and eyewear) should have functionality improvement and new use cases enabled by lower latency and local compute capacity for devices.
Those at risk of greater competition include mobility network operators (Uber/Lyft), eCommerce pure plays from enhanced local retail capabilities (Amazon/eBay/Wayfair), and hyperscale cloud providers given likely new cloud services from telecom providers (AWS/Google Cloud). See Appendix 1 for high-level thoughts on tech sector beneficiaries from edge compute.