As we’ve reported several times recently, edge computing is tipped to be big, driven largely by the demands of IoT. And it seems every vendor is jumping on the edge computing bandwagon.
In June this year Hewlett Packard Enterprise (HPE) announcing plans to invest $US4 billion in intelligent edge technologies and services over the next four years. And last month Nutanix announced its entry into the edge computing market with the launch of Nutanix Xi IoT edge computing technology.
The latest vendor touting its edge computing credentials is Vertiv. It has just issued its Top Five 2019 Data Centre Trends, proclaiming that “Edge Will Drive Change.”
Verti heads its press release announcing these predictions with the following statement:
“The edge of the network continues to be the epicentre of innovation in the data centre space as the calendar turns to 2019, with activity focusing on increased intelligence designed to simplify operations, enable remote management and service, and bridge a widening skills gap.”
Vertiv CEO Rob Johnson, said: “Today’s edge plays a critical role in data center and network operation and in the delivery of important consumer services. This is a dramatic and fundamental change to the way we think about computing and data management. It should come as no surprise that activity in the data center space in 2019 will be focused squarely on innovation at the edge.”
Vertiv’s enthusiasm for edge computing raises the question: what exactly is edge computing? Vertiv specialises in data centre infrastructure: power and cooling systems. So it is hardly likely to get excited about a booming market in boxes that plug into a power point.
For example HPE has just announced the first fruits of $US4b edge computing investment, the HPE Edgeline EL300, describing it as “a fan-less, low-energy system equipped with Intel Core i5 processors, up to 32GB of memory and 3TB of storage.”
The Edge Computing Consortium, in its 2016 white papersays edge computing is performed on “an open platform at the network edge near things or data sources, integrating network, computing, storage, and application core capabilities and providing edge intelligent services.”
Defining the edge
Some insight on how Vertiv views the edge can be found in a Vertiv white paper Defining Four Edge Archetypes and Their Technology Requirements.
In it Vertiv argues there are few, if any, resources that provide a comprehensive view of the edge ecosystem. “A close analysis of the market reveals a wide variety of current and emerging use cases and, while they share some similarities based on the broad definition of edge computing, they are also distinct in some significant ways.”
After analysing various edge computing use cases Vertiv claims to have identified four main archetypes for edge applications:
– Data Intensive
– Human-Latency Sensitive
– Machine-to-Machine-Latency Sensitive
– Life Critical
Its white paper presents “a description of each archetype with examples of the most impactful use cases, along with an overview of their connectivity requirements to local, metro and regional hubs, which represent the edge transmission layer and core and are sometimes differentiated as edge, fog and cloud computing.”
Vertiv argues that the infrastructure required to support these current and established use cases consists of devices that produce or consume data and four layers of storage and compute.
This is where it gets interesting, and where Vertiv’s view is perhaps at odds with other visions of edge computing. Vertiv’s four layers are
– a processing endpoint
– a local data hub
– an urban data hub
– a regional data hub.
Vertiv’s view of the processing endpoint seems close to what many view as an edge computing device.
A processing endpoint that “may be as simple as the PC or tablet a consumer is streaming video to, or could be the microprocessors embedded in automobiles, robots or wearable devices.”
However it does add “These components are application-dependent and are typically designed in by the equipment manufacturer or retrofitted to existing devices,” which is at odds with the idea of an edge computing device as being fairly generic.
The next bit of infrastructure in the chain is certainly not within the ambit of what most people would consider and edge computing device.
It fulfils the widely accepted idea of edge computing in that it “provides storage and processing in close proximity to the source.”
The 300kW edge ‘device’
However Vertiv says this local hub will most likely be “a rack- or row-based system providing 30-300kW of capacity in an integrated enclosure that can be installed in any environment. That’s a far cry from HP’s fanless low energy edge computing device.
Does it matter? Well yes it does, especially when people start making predictions about the edge computing market.
Take this report from GrandView Research with the mouthful of a title Edge Computing Market Size, Share & Trends Analysis Report By Technology (Mobile Edge Computing, Fog Computing), By Vertical, By Organization Size, By Region, And Segment Forecasts, 2018 – 2025.
Its 92 pages will set you back close to $UD6000 and it promises to answer six key questions about edge computing. “What exactly do we mean by edge computing?” is not one of them.