Vishwamitra Nandlall, VP Technology Strategy & Ecosystems at Dell
The popularized model of the internet is anchored in its roots as a hierarchical network cascading from global backbones operated by big telecom to the so-called “eyeball” networks connecting users. The value of such a network was to connect highly distributed servers hosted in enterprise and university footprints to consumers. The contemporary internet looks much different: the hierarchy has been flattened by hyperscale data centers which are densely peered through global colocation facilities and exchange points. While historically transit networks connected backbone networks to the eyeball networks, today the hyperscale data centers have internalized backbones and are connected directly to the eyeball networks, bypassing transit networks for efficiency. This modern era has re-homed content from highly distributed data centers to the hyperscalers’ centralized compute resources. These changes have transformed the internet into a client server model which has fundamentally altered internet economics.
The next internet era will upend the client server model through the introduction of edge computing. The breadcrumb trail of edge computing begins with content delivery networks (CDNs), which transformed how we scaled service platforms. Instead of pushing content from a centralized server farm from across the network, off-net edge compute clusters were deployed at the periphery of the internet in the eyeball networks. This pattern pushed service delivery towards the client using just-in-case pre-provisioning, and relegated the internet core to function as an opaque and largely private network that replicates and synchronizes services to edge delivery points.
The modern concept of edge moves beyond just content delivery into general purpose and specialized compute resources. In this context, the likely evolution of edge will be to commoditize transmission costs, increase the abundance of compute by orders of magnitude, and orphan the common pooled resource model of the internet from the client service delivery model. For all intents and purposes, the edge becomes the internet.
We built an Internet based on what we needed and had at the time. But, as we started to see new real-time services - like video - being delivered through the internet infrastructure, we had to change the architecture to reduce propagation delays – leading to the rise of CDNs.
Today, we're seeing significant progress in other media, like AR/VR, cloud gaming, and the rising metaverse. These new classes of content will require responsive compute resources which today are unevenly available.
The demand for new applications is not limited to consumer markets, enterprise services also loom large. For instance, the existing Internet cannot meet the real time constraints that you see in networks for industrial control systems. Instead, enterprises must deploy proprietary stovepipe technologies, the costs of such systems limit accessibility to only large asset-intensive industries. By extending the Internet with latency-resistant resources which are closer and more accessible to operating technology, we can start to contemplate delivering industrial control systems more broadly, making them cost-effective and available to a larger percent of enterprises.
I’ve spoken mostly about “poster child” applications for edge, one area that is not as well conceived is security and privacy as part of the internet commons. Historically, the way we implemented such controls was through middle-boxes that performed tasks such as deep packet inspection, intrusion detection, and firewalling at the enterprise perimeter. The emergence of cloud services has migrated enterprise traffic from east-west flows to north-south flows. This trend, combined with the emergence of software defined wide area networks and edge computing has evolved to form secure access service edges (SASE) popularized by companies such as Zscaler. These services provide a secure edge cloud to protect enterprise traffic from malicious attacks. As we see the penetration of more service edges grow, we can imagine an internet model that has an opaque private core surrounded by open service edges, these service edges could peer with other service edges to form a mesh of tunneled connections - an open grid in fact. This new edge grid could proxy user traffic to implement privacy preserving architectures. For instance, it is possible to imagine providing attestation services between endpoints, or overlaying a ToR-like network on the open grid to protect user identity and traffic patterns. Such capabilities could not be scaled sufficiently in the current internet architecture. These privacy capabilities could fundamentally alter the advertising-economics of the internet, dramatically reinventing service incentives and value capture.
For consumers, security services, such as video surveillance and home security, come to mind first. These are applications that will undoubtedly benefit from the emergence of an edge architecture and an Open Grid. The gaming industry will change as well; people will no longer need expensive consoles to play the best games; they’ll be able to benefit from the latest graphics processors delivered from the Open Grid and use a relatively thin client. This will make the most advanced games more widely available without requiring consumers to purchase the latest high-end console.
Smart City infrastructure will also benefit from the Open Grid, as there is a need for highly reliable, consistent, low-latency— yet cost effective —application delivery mechanism to provide value at the scale of a city.
For industries, I believe we’ll soon start to see the Open Grid deliver services that have previously required massive investments in proprietary architectures, limiting their availability to only the most capital intensive industries, which has stunted industry growth.
Dell’s mission is to be the essential infrastructure company. The megatrend of mixing physical and digital, whether we call it IoT, digital twins or metaverse, will be underwritten by edge infrastructure, and Dell is committed to manifesting this future.
Carlota Perez defined two phases in every technology revolution: the installation phase, when the technology comes into the market and the infrastructure is built, and the deployment phase when the technology is broadly adopted by society. The transition between these two phases is an unstable period where profit pools shift. This will likely be driven by a cloud-first to edge-first economy, and Dell will have an important role to play in navigating this turn.
I'm a fundamental believer that society is shaped by the technologies through which it communicates. I think the edge and the Open Grid has the potential to drive a new golden age of Internet growth, and solve the security puzzle that has thrown mud in the axles for essential services to flourish.
Vish Nandlall is the Vice President of Technology Strategy and Ecosystems at Dell Technologies. Previously, he served as CTO for Ericsson NA, designing the regional market strategy and technology roadmap. In this role, he developed and executed a plan to deliver five emerging business organizations - Cloud, SDN, OSS/BSS, Video and Mobile Enterprise – collectively these businesses realized an annualized $2 billion in new revenue at a time when the NA market revenue was $9 billion total. This success hinged on Vish’s ability to develop a portfolio through R&D prioritization, M&A planning and management of the service and sales support recruiting and training.
Vish began his career at Nortel Networks, where he started in 1995 by leading the design of the world’s first distributed call processing platform. Over the next ten years, he went on to pioneer 10Gbps IP forwarding, act as a key contributor and author of the Verizon A-IMS specification, and led the industry’s first LTE trial system. In 2005, Vish was named Nortel’s CTO for the Carrier Network Group as well as a Nortel Distinguished Engineer for his technical leadership across the firm’s IP, optical, and wireless products.
Vish holds a Bachelor of Science in Engineering, First Division, with a focus in Electrical Engineering from the University of New Brunswick. He has been published several times, holds multiple patents, and has been listed by Wireless World as one of their Global 100 Wireless Experts.
The OGA Blog Series highlights the knowledge and strong opinions of today's top executives with regards to the open grid, and how the Internet will evolve over the next ten years. The statements in each blog are of the author, and not of the Open Grid Alliance as a whole.