Karthik Krishna, Product Manager, Edge and AI Solutions at VMware, will present “Realize the Edge Transformation: Deliver AI to Improve Business and Customer Experiences.
Karthik Krishna:
Hi, and thank you for joining this session. I’m Karthik Krishna, product manager, service provider, and edge business solutions at VMware. Over the last two years, we have worked with many of our service provider and enterprise customers on better understanding their needs and goals when it comes to exciting transition that’s happening around us towards edge technology. I hope to share with you VMware’s approach to this space and help you understand how we will look to collaborate with you on this journey. I’ll start by looking at what’s going on in industry today, right? If we can describe it in one word it’s disruption and that’s going on throughout verticals, right?
Karthik Krishna:
Let’s look at three industries for example, Retail, if a car shows up at a gas station, the station recognizes the car and can provide the driver some offers like stacks or coffee you brought before, or even an oil change to keep that car running. In Industry & Manufacturing, we are looking at smarter machines becoming autonomous and in Healthcare, the wealth of data accumulated hold huge promise in making better decisions that meet the needs of individual patients. You see across the board, regardless of the vertical, there is a seat of change that’s happening. And this change is happening where data is being produced and consumed.
Karthik Krishna:
Customer expectations are changing and industries are facing the opportunity to provide them through digital transformation. Multiple different technologies are converging. From an enterprise perspective, architectures are changing to take advantage of these new capabilities. And we are seeing this happen at many of our customers who are living and breathing this on a daily basis. Their organizations are highly distributed, both in terms of the digital infrastructure they rely on and the human workforce that propels their business every day, data is produced and consumed in a widely distributed fashion.
Karthik Krishna:
This looks like a bit of a chaotic picture, honestly, and many organizations content with the fact that it is. We at VMware are trying to come up with an approach to put some structure around this chaos as we call it, right? And we start by looking at the bottom layer, the data center, the near edge and the far edge. So over the last several decades data center evolved from being run by individual business units into a highly centralized IT function, conserving cost talent and improving efficiency, living major enterprises with a small number of data centers. Now, as the public cloud emerged, many of the applications that used to run in the data center were migrated to leverage those cloud capability, a process that’s still ongoing, honestly, right? The pattern that we have seen emerge was distribute when you must and centralize when you can. That served us well in the 2000s and gave rise to many centralized cloud based applications.
Karthik Krishna:
But now when the source of the data and the consumption of the data is so distributed, you can not simply rely on the centralized models to meet the growing needs of the enterprises. That is a diminishing benefit to using the cloud when it comes to this level of distribution. Now, the reality then that we face is that these are all examples and use cases, whether it’s autonomous vehicles, whether it is omnichannel retail, whether there are connected oil and gas platforms across the board, there are many examples of the distribution of data production, processing and consumption.
Karthik Krishna:
VMware is out there trying to identify the patterns behind these use cases. There are some common elements that emerge showing in the bottom. These are the foundations, what we call the underlay technologies, all of which are emerging and will be used to compose a programmable infrastructure. We are talking about things like high speed and capacity 5G networks that can also be rolled out privately, secure access services and the ability to share content and data faster than before. But that infrastructure will in turn be used by the use cases you see about the overlay, the distributed applications that need data and processing power to deliver their innovation.
Karthik Krishna:
What we envision is a single consistent software stack that enterprise can use and service providers can deploy on top of their connectivity layer to serve these emerging needs. We turn this stack, the edge. So what is edge? The edge is this digital infrastructure that serves the need of workloads that are running at a number of locations close to or where data is being produced and consumed. Now that definition demands some additional discussion here, right? At VMware, we see a variety of situations that meet this edge definition. But from architectural standpoint, we decompose the edge stack into three tiers. Along with the management plane. The layer at the bottom is home to the edge infrastructure. This is the underlay providing the actual network that we talked about. The radio area network, 5G, the fiber and the software that needs to run in order to provide connectivity.
Karthik Krishna:
The layer on top of that comprises what we call the overlay services. These are services facilitating the ability to make the network programmable and secure based on the infrastructure, and also provide the capabilities that the edge applications depend on. The top layer is home to these edge applications, healthcare connected car or retail apps that we spoke about, right? They all need applications services to facilitate execution in a distributed fashion. Now to bring it all together, there must be a management plan that provides not just a telemetry, but that provides the configuration, the installation, the operations, and the maintenance of this entire system. This is our VMware edge vision, VMware follows an important principle in which we aim to avoid looking our customers to a particular architecture or vision. We want to offer you choice and allow this edge to run on a variety of cloud based technologies and different stacks and frequently act as a conduit to a backend hosted by hyperscalers. We call this the multi-cloud edge.
Karthik Krishna:
We believe that we are in the lead to provide enterprises with the choice for the multi-cloud edge, through the service providers who are critical in the delivery of the edge as we move along. Looking at the deployments and use cases for this edge, this is where things come together. At a high level, there are two main scenarios. The near edge is where applications and services are hosted by the service provider whether in their data center, local office, or even at the radio site and the far edge where they are managed by the enterprise on premises, looking deeper at the near edge at service providers, we are seeing increasing disaggregation taking place at the RAN, their network core transitioning into virtualized platforms, running off the shelf hardware. And finally the regional and local points of presence, which we believe will be the main physical presence for edge applications.
Karthik Krishna:
DCSPs will provide the connectivity, security, hardware, and software stack to host edge applications for customers. The far edge locations are on customer premises. This can be as basic as a sassy device offering SD-WAN and now hosting one or more applications. We are also working on support additional devices. One that is slightly bigger than such an SD-WAN device, but capable of hosting containers, virtual machines, and supporting applications with more robust requirements, this would fit large retail locations or distribution centers. You know, on the other X stream, we are working on very thinned out edge that will run a simplified stack with a minimal number of containers that support agile application development. These are the used cases that we are seeing, and we ensuring it supports the multi-cloud edge principle. Now looking at the actual structure we put together, VMware is looking to address the spectrum of form factors and computing capacities through a consistent software stack.
Karthik Krishna:
Instead of focusing on the backend, the data center, our cloud location, we are looking to address a low latency and small footprint at the same time to connect all these devices and locations together. We do that without sacrificing consistency and applications running on VMs and containers. The underlay services are then going to be used by service providers to build networks with trusted VMware offerings like NSX and vSAN but also new edge focused platforms like ESX real time, hypervisor real time basically, right? Running in real time and both open source Kubernetes and our Tanzu Kubernetes Grid (TKG). The overlay edge services are built on top of these foundations with networking and security services, ensuring edge applications can sustain the SLAs and end user experience required by enterprises for their branch locations and distributed workflows.
Karthik Krishna:
We are actively working with hyperscalers and with our partner ecosystem to expand what can be done at this layer. Finally, the compute services exposed to the edge applications and ISVs are moving to leverage this edge capability to offer innovations that were impossible before in a cloud centric board. This all come together and orchestrated through our management stack, working across the underlay with tools like Telco Cloud Automation or Lead Service Management and Tanzu Mission Control, or seeing the compute layer. Consistent management is critical allowing core business services to run and be monitored on known and trusted platforms. As I’m speaking to you all today on infinite vision conference, I would like to highlight our collaboration with Chooch AI to make the AI on edge a reality. You all already knew, well that camera sensors create new possibilities for multiple industries, including retail, healthcare, and manufacturing. Wireless technologies including 5G adds another dimension as it provides connectivity and access to remote data from these sensors.
Karthik Krishna:
Now VMware in collaboration with partners like NVIDIA provide an end to end platform for AI and to deliver a new architecture for the data center cloud and edge to support the needs of next generation applications like Chooch AI. Now together we unlock the potential of AI and next generation apps for every business in every industry, from medical researchers, looking to accelerate breakthroughs in treatments and immunotherapy to financial services organizations protecting assets with advanced fraud detection to retailers exploding new insights that enhance customer care to minimal. And we Chooch AI in the VMware marketplace now. VMware marketplace is a one stop shop for VMware validated and certified ecosystem solutions that enables customer to discover, try, purchase and deploy directly to VMware endpoint. To summarize the benefits of the VMware approach. We are looking to support service providers that are trying to modernize their infrastructure underlay.
Karthik Krishna:
The edge can help them monetize billions of dollars invested in new systems and software through new enterprise services they could not offer before. Enterprises, in turn are looking to accelerate their edge journey as such new capabilities gives them an agile path to rapidly release and improve new services and accelerate their digital transformation often using existing infrastructure they already have. They can do all that without compromising security and while improving their user experience. This is an exciting time for all of us at VMware, seeing the transition to the edge coming to life and work with partners like Chooch AI to make it easier for our customers. I look forward to connecting with you and learning from you all about new and unexpected ways the edge is transforming your business. Thank you.
