Modern applications like 5G connectivity, crypto/blockchain, Internet of Things (IoT), social/video media, and artificial intelligence/machine learning (AI/ML) are redefining business processes and user experiences for companies of all sizes. There’s another thing these applications do, however: they create tremendous strain on existing networks that may have been designed and deployed for the requirement of last generation’s apps.
It is undeniable that the networks of today must be able to process increasingly complex requests and update information in real time. This shines a glaring spotlight on the need for not only higher performance networks but also intelligent operations that can manage both the infrastructure and the apps, workload, data, and workflows that it supports.
This column examines one specific application (AI/ML) and makes the case on why companies today should consider investing in modernizing their networks to harness the value of this app and maximize its full potential.
Why AI/ML and why now?
Artificial intelligence as a technology concept goes as far back as the 1950s, originating as a summer research project at the University of Dartmouth where the term was coined. But today we are rapidly approaching mainstream adoption of artificial intelligence and machine learning, the most well-known form of AI. For example, ChatGPT may be fun and games now, but the potential for revolutionizing many business functions is immense.
In general, there are several business-focused benefits that companies are evaluating for their AI/ML investments. One, for higher levels of intelligence and insights that can help them make better and faster decisions. Two, enriching their customer experience by delivering more engaged, personalized products and interactions. Three, update or even transform their business processes to drive efficiency and increase productivity. Finally, to help find, hire, and empower their workforce with higher levels of efficiency and engagement.
As AI/ML adoption accelerates, companies will need to invest in the tools and expertise to remain competitive. However, many companies today do not yet have the ultra-high-performance computing and networking bandwidth to implement them efficiently. It is then imperative that enterprises take steps to ensure optimal performance within their multicloud networking infrastructure and also make sure that their operational models are in top shape when processing intensive workloads in the data center and cloud such as AI/ML.
The case for 400G and 800G networking
Despite this, many enterprises are either just evaluating or have not yet decided to transition to higher performance networking. These companies may be surprised to learn of the breakneck pace of innovation happening within data center and multicloud networking. Today 100G (gigabit) fabric technology is standard fare. It is not a stretch to think that the industry will soon be reaching a tipping point where network architectures require flexible scalability of 400 and 800G capacities to support new applications, data, and workflows including AI/ML.
While the gains in bandwidth and capacity are compelling a reason enough to migrate to these new network standards, it’s important for these companies to consider other key benefits of this investment:
- Enhanced visibility and control: Modern, high-performance networks are designed to enhance the visibility of application flow levels across their networks, i.e., more application-aware, especially in hybrid cloud use cases. Specific use cases can involve the ability for proactive monitoring and automating troubleshooting to improve the quality of experience metrics of their applications such as large-scale AI/ML clusters.
- Accelerate new experiences: Modernizing network infrastructure can pave the way for new classes of applications that will deliver revolutionary capabilities and enhanced experiences for customers, employees, and partners. These include the specific business benefits of AI/ML mentioned above.
- Redefine data center economics: The latest network infrastructure can deliver tremendous economies of scale such as power optimization, cooling efficiencies, and investment protection. These new classes of networking products are also optimized to work hand-in-hand with intelligent operations platforms, many of which are cloud and SaaS-delivered.
- Support sustainability goals: Environmental, social, and corporate governance are top priorities for companies today and this has put added pressure on IT organizations including network operations teams to support sustainable IT and data center initiatives. Modernizing network and data center infrastructure can be a key step as many of the newest products were designed at the architecture level to operate more efficiently.
Recommended steps for moving forward
Modernizing large enterprise networks is not a simple, unilateral decision. There are several considerations that all companies should be aware of as they evaluate their options for products and suppliers.
One, it is important to evaluate the full breadth and depth of supplier’s entire portfolio, including software compatibility, comprehensive testing validation, industry standards backing, and architectural consistency.
Two, consider other factors such as supply chain reliability, built-in security capabilities, application intelligence, and the ability to deliver cloud agility.
Three, consider how well the supplier can design a network to meet your specific requirements. AI/ML may be available to any company, but how you use it to your advantage is likely unique.
There is no doubt that there is an unrelenting expansion of data center traffic that is fueling the demand for high capacity and highly intelligent data center networking solutions. In fact, the explosion in the capacity demands of data-hungry applications like AI/ML is outpacing the current high-speed transport abilities.
However, many companies have not yet taken the proper steps to modernize their multicloud networking infrastructure and operational model to keep up with these new requirements. The time is now. Companies who lack the network infrastructure to handle apps and workloads like AI/ML risk being less competitive and leaving a lot of unrealized business value on the table.