IT leaders: Be strategic about your critical workloads

In hindsight, it’s easy to marvel at how cloud computing single-handedly flipped IT operating models worldwide.

Cloud software enabled lines of business to take the DIY path to procure computing services, often by swiping a credit card. IT leaders found their once centralized functions shifting outside their department as applications and infrastructure proliferated in the shadows.

In truth, it happened gradually, then suddenly. IT leaders didn’t stand a puncher’s chance of resisting. And today, they wouldn’t dare.

Where IT leaders once ruled technology capabilities with iron fists, they now collaborate with business stakeholders to deploy and provision technology services across several environments. IT has effectively become decentralized. And while this looks different in every business, common denominators remain.

Workload placement is critical for modern IT

As IT teams work with business stakeholders to implement agile IT services, they choose and procure IT delivery models and environments that are becoming increasingly distributed to accommodate application and data requirements.

Modern computing estates are highly hybrid and span on-premises infrastructure, public and private clouds, colos, and edge devices, which present performance, security, governance, and financial challenges.

In such decentralized environments, where IT resources are distributed across multiple locations, strategic workload placement helps IT and business decision makers assess the criticality of applications to the business, along with their unique characteristics, and select deployment environments.

This exercise in intentionality is the new operating model and IT must get on board with it—if it isn’t already.

What factors go into strategic workload placement? IT leaders must weigh application and vendor-specific criteria, industry regulations, business policies, as well as capital and operational budgets, as noted by IDC research sponsored by Dell Technologies¹.

For example, IT organizations must consider compute and storage requirements such as latency and throughput, which impact performance. Requirements for scaling compute, storage, and bandwidth to meet the demands of applications are also important.

How easily a workload can be migrated from one environment to another (such as on-premises to a public cloud and vice versa) is another factor as requirements for certain applications can shift over time. High-performance workloads may initially run efficiently in a public cloud, but escalating costs to scale them may result in IT staff on-premises, or to a colo—commonly referred to as repatriation.

Placing AI workloads: It’s complicated

For example, machine learning (ML), data analytics, and high-performance computing (HPC) workloads can strain IT budgets, as they are often require in fit-for-purpose infrastructure optimized for performance. As a result, many organizations initiate such applications in public clouds. However, this becomes expensive as usage and resource requirements scale in production.

Organizations should take a more nuanced approach to mitigate the financial impact, according to IDC. For instance, companies executing HPC workloads might build and run their models, which are performance intensive and may require specialized hardware, on-premises. However, they may run their less performance-intensive simulations in edge or public clouds, benefitting from the ability to spin up and spin down compute instances rapidly.

Data security is another reason to choose workload locations. Before and during the process of model building, most organizations can scrub the data of any sensitive information. However, the environment itself must remain secure.

Whether the applications include HPC, client computing, application development, or other categories, IT leaders recognize that workload placement is not a one-and-done activity.

They must constantly evaluate workloads and adjust compute or storage requirements in accordance with business needs. Regulations change, as do the costs to execute applications. This often leads to workload migration and replatforming, refactoring, or rehosting to a new deployment environment.

Even so, relaxing centralized IT management controls while hewing to overarching governance policies can lead to a more agile IT service and provide the kind of consistent cloud experience business staff, developers, and other key stakeholders are already accustomed to.

Choosing the right workload placement partner

Allocating workloads strategically in decentralized IT environments is no trivial matter. To stay competitive, businesses must be able to deliver both their traditional and cloud-native applications at the right place, at the right time, and at the right cost.

Given the complex nature of managing distributed workloads and skill shortfalls, many companies must work with trusted partners with deep connections to tech ecosystems, and that can help them mitigate the challenges of operating multicloud estates.

Our Dell APEX portfolio of solutions is positioned to address a broad range of customer needs and workload requirements across the total multicloud estate via as-a-service and subscription consumption models.

In addition to simplifying the procurement, payment, and consumption of compute and storage on-premises, the APEX portfolio allows organizations to operate their storage, data protection, and cyber recovery services in public clouds, such as Amazon Web Services and Microsoft Azure. Also with APEX: Developers can run cloud stacks and Kubernetes clusters built in on-premises, gaining security, performance, and control.

Decentralized IT environments have always posed unique obstacles for IT and business cohorts. Dell APEX provides the technologies, services, and flexible consumption that organizations require to power their digital business initiatives and gain a competitive edge—today, tomorrow, and in the future.

Learn more about how Dell APEX can help you allocate workloads strategically across your multicloud estate.

1Essential Considerations for Workload-Centric IT Strategies, IDC, August 2023.
2 Dell internal survey of IT decision makers, June 2023