The metaverse will challenge every aspect of running a web-scale company. Metaverse technologies such as AR/VR, 8K video, and cloud gaming in the consumer space, as well as Industry 4.0 technologies such as digital twins, blockchain, IoT, and autonomous operations, will require a reimagining of today’s web-scale networks.
Web-scale enterprises and the cloud are at the heart of the future of communications. Therefore, data center fabrics and IP and optical WANs must evolve. That risk will only get higher as 5G and 6G networks connect everything from pacemakers to low-Earth orbit satellites. Automated assurance and security must be built into the DNA of your network. Preparing for these future consumer and enterprise services will require a multifaceted strategy.
Scaling in all dimensions
Until the early 2000s, scaling computer processors relied on a one-dimensional strategy: increasing clock speed. Once the physical limits of clock scaling were reached, multicore processors emerged. We are now reaching new limits, leading to a proliferation of specialized task-oriented processors and other ways to offload parallel processes. Future scaling proposals are envisioned, most of which increase hardware complexity and must be handled in software.
We see that a similar trajectory occurs in the network. Optical transport uses a different strategy than routed wide area networks. Both are distinct from data center fabrics, where increasing bandwidth and speed has long been the primary scaling strategy. However, we are reaching the limit where simply overprovisioning is neither affordable nor successful. Providing a metaverse requires more than extraordinary speed and massive amounts of capacity.
Optical transport, IP WAN, and data center fabric, previously treated as separate domains, are now working together in smarter, more complex ways to scale, guarantee, and secure end-to-end connectivity. need to do it. This requires orchestration across all aspects in a responsive and scalable manner. Many different but interrelated technologies need to work together. Companies can no longer over-engineer network layers and expect cost, efficiency, and scale to meet customer needs.
Data center, IP routing, and optical integration
The growing importance of data center interconnect (DCI) has put WANs in the spotlight. With the rise of edge clouds and the need to achieve ultra-low latency, especially for autonomous and industrial applications, networks need to become more layered and decentralized. Traditionally, workloads moved primarily within data centers, but increasingly they are moving not only within metro area networks but also between data centers.
At the same time, workload requirements are becoming more demanding. AI workloads and their training requirements often involve communication between workloads across multiple data centers. These communications are susceptible to packet loss and delay. Not only must the data center network fabric be designed to accommodate these requirements, but the IP routed WAN must also support these performance metrics.
Routing decisions made at the IP layer must be supported at the optical transport level. Overprovisioning at the optical layer has historically been the norm, as router Ethernet ports have grown from 100 Gigabit Ethernet (GE) to 400GE and even 800GE. This means that the underlying optical capacity cannot simply be assumed. It is also important to know the quality and reliability of the optical link. Does the optical signal travel far enough and support the Ethernet port speed to reach the new edge cloud? Web scalers planning DCI networks consider both the IP routing layer and the optical transport layer is needed.
Similar to the example of computer parallelism, complex network architectures must be handled by increasingly complex software. Such network software dynamically adjusts various elements, turning network elements on and off, depending on the workload or application being supported, to achieve the lowest cost, most power efficient, or best Responsible for transitioning tasks to performance resources. And this all needs to be done automatically and securely.
Unlike a single computer design, networks are built incrementally with layers of legacy network elements and software that cannot be simply overhauled. Despite these limitations, entire networks are increasingly being automated to provide the necessary operational scale, efficiency, and cost-effectiveness. Automation is not available in all network technologies so far, but this level of control is being built in the DCI space more than in public networks with legacy services such as voice that do not present new requirements. there is.
For webscalers, the future of metaverse applications and increased reliance on the cloud means automation of both IP routed and optical networks is driven by economics and opportunity. Especially when it comes to the industrial metaverse, web scalers have strong incentives to invest in their networks. To meet the performance requirements of these new use cases, it is essential to establish a converged multilayer network infrastructure with unified control and open software-defined networking (SDN) interfaces across both the IP and optical layers. .
In addition to delivering new services, this approach to networking also improves operational efficiency and cost effectiveness. In addition to automating many tasks, it simplifies and abstracts many network processes and allows programming of the underlying infrastructure (often referred to as infrastructure-as-code). This not only empowers network engineers but also saves on network equipment costs by optimizing resources through cross-domain coordination and orchestration.
Quickly analyze network telemetry powered by AI/ML analytics to troubleshoot network and service issues. Dynamic awareness of traffic flows across multiple layers enables coordinated operations, such as minimizing the impact of optical topology changes on IP traffic. Diversity analysis allows you to automatically assess truly comprehensive path diversity to ensure highly reliable service delivery. The result is a more efficient, reliable, and scalable data center interconnect that can meet the workload requirements of AI and metaverse applications.
Securing the metaverse
From self-driving transportation and telemedicine to financial services and smart city infrastructure, even a cursory review of the many proposed metaverse use cases highlights the increasingly mission-critical aspects of tomorrow’s services. It will be. This also requires a multifaceted approach to security.
If you look at distributed denial of service (DDoS) alone, it is currently the fastest growing traffic category, even surpassing gaming and video. To provide the scalability and functionality needed to protect highly distributed, mission-critical networks, network security must become more like packet forwarding, a high-performance, scalable feature of routed networks. It is necessary to become. Bolt-on security appliances cannot be expanded. Like assurance, security must be built into the DNA of your network. The router itself must be a highly accurate attack sensor and mitigation element without compromising other services running on the same router.
The opportunities for web scalers are unprecedented. They are building the Metaverse platform of the future. As the cloud becomes more generalized, distributed, and mission-critical, every aspect of the network must be modernized, restructured, and reimagined to meet the needs of the new metaverse and the critical services it provides.