After decades in enterprise networking, Cisco entered the compute market in 2009 with its Unified Computing System (UCS). Ten years later, demand for fast application deployment and for private and multi-cloud infrastructure was changing how businesses build and run their core IT. Kaustubh Das, then Vice President of Product Management and Strategy for Cisco’s Computing Systems Product Group, shared his view of where the data center is heading. This article summarises the main ideas and what they mean for IT leaders.
The biggest trends affecting the data center
- Higher user expectations. In a digital world, IT must speed up how it delivers services so that applications, DevOps teams, lines of business and end users get what they need at the pace they expect. For most companies, applications are now the main way to interact with customers, partners and suppliers.
- Many more technology choices. Data center leaders once only had to pick the right compute, storage and network. Today there are many design patterns – converged infrastructure, hyperconverged infrastructure (HCI), public cloud, hybrid cloud or traditional standalone systems – and similar choices at every layer: hypervisor, file system, management stack, workflow automation, security, ticketing and IT operations management.
- Exponential pace of change. Analytics, fast storage and cloud, then hyperconvergence, and now AI and machine learning. Development has moved to containers, microservices and Kubernetes, with new languages and PaaS frameworks. A new technology often arrives before the previous one has been evaluated.
- Workloads moving to the edge. More workloads run in remote offices and branches, creating new opportunities but requiring a management model that covers distributed computing everywhere. Many management tools do not scale to this, so organisations end up with several tools and more operational complexity.
- Multi-cloud as the norm. The discussion is no longer which applications move to the cloud, but how one application can span several clouds – for example a web tier in the cloud and a database on-premises – and how cloud fits into the application lifecycle, such as developing in the cloud and running in production on-premises. As experience replaces hype, IT is also seeing the real cost of cloud: very efficient for some use cases, more expensive than owned infrastructure for others.
How changing demands are changing the role of IT
IT’s role used to be simple: source the best, cheapest infrastructure and deliver SLAs for infrastructure and applications. That has become far more complex.
- From buying software to running development platforms: as most companies become application companies, IT now creates and maintains development environments, which has driven the rise of DevOps. In a data-driven world, IT’s key customers also include data scientists and business units.
- Distributed data and applications: applications are becoming connected networks of functions and data spread across on-premises and public cloud. With IoT, devices and connections are multiplying, and it is often more efficient to process data where it is created than to move it.
- Holistic, silo-free infrastructure: IT is expected to avoid silos and reduce integration, management and connectivity problems from edge to core, and to deliver convenience – not only lower cost and higher uptime.
- Policy-driven, elastic infrastructure: the new mandate is to provide near-infinitely elastic infrastructure and to set policies so that machines manage machines. Public cloud becomes one of many infrastructure elements, while cloud-like elasticity, scale and programmability are brought to everything IT owns.
- Security everywhere: as attacks grow more sophisticated, security cannot sit only at the perimeter; it must be built into hardware, software, management and the supply chain.
How the industry is responding
Large infrastructure and cloud vendors innovate to deliver the experience customers want, but often by keeping customers inside their own ecosystem, which creates new silos or reinforces old ones. Many start-ups build open-source tools, PaaS and SaaS that solve specific problems; these are easy to access but lead to fragmentation and ongoing complexity. Meanwhile component technologies – CPUs, GPUs and accelerators, flash and persistent memory – often evolve faster than the industries and software ecosystems that use them, so their benefits can be lost in the complexity they add.
Cisco’s approach to compute and storage
- Revolutionary infrastructure, evolutionary operations: customers are at different stages of adoption, so new infrastructure must coexist with what they already run and extend the value of past investments.
- Making next-generation technology enterprise-ready: architectures born in the public cloud and open source need to be hardened for enterprise scale, mixed environments and incompatible technologies. Cisco captures tested blueprints as Cisco Validated Designs (CVDs) for databases, ERP, big data analytics, software-defined storage and newer areas such as AI, ML, IoT and edge computing.
- Multi-cloud by design: customer environments will span several clouds, one of which will be an automated on-premises data center. IT needs to connect clouds, split applications across them, develop in one and deploy in another, while keeping networking, security and operations consistent.
- Policy and automation across the portfolio: UCS separated a server’s identity from its hardware; Cisco Intersight extends this with policy-based infrastructure management delivered as a cloud service, giving a consistent experience whether servers sit in the core, at the edge or anywhere else. Analytics and machine learning point to a future where machines manage machines, issues are fixed proactively once found, and people focus on creating new value.
Questions IT leaders should ask about data center and cloud strategy
- What do we want to control, and what are we happy to buy as a service? Are we optimising cost for every infrastructure element, or prioritising something else, such as paying for automation to free up engineering capacity?
- Which IT vendors do we trust as long-term partners? Assess their long-term financial sustainability and whether they optimise for themselves or for their customers.
- Which lock-in are we willing to accept, and which should we resist? Where is it worth paying for control and integration, and where do we want choice and openness?
- Has our IT organisation really changed in the last 10–15 years? If technology, priorities and operating models have changed but the organisation has not, ask why.
IT leaders work in a complex, fast-moving and often unpredictable world. They can still meet new challenges by preparing to respond quickly, keeping their options open and choosing their IT transformation partners wisely.
Need a check of your network, servers or backup?
IT Systems Vietnam reviews LAN/Wi-Fi, servers, cloud, VPN, DNS and backup to reduce downtime and the risk of data loss.




