Cloud Considerations for Enterprises

In the enterprise world, the term Cloud has so many connotations ranging from Public Cloud, Hybrid Cloud, Multi Cloud on the infrastructure side, to Platform as a Service Cloud to Software as a Service Cloud. Here, we will focus on Infrastructure as a Service and Platform as a Service components of Public Cloud. According to Gartner, in 2022, more than $1.3 trillion in enterprise IT spending is at stake from the shift to cloud, growing to almost $1.8 trillion in 2025.

RATIONALE FOR PUBLIC CLOUD

Before Public Cloud came into existence, enterprises needed to get into a CapEx intensive technology consumption model. They needed to get a space for data center/colocation, procure, rack and stack servers and storage, wire the networks, install operating systems and/or virtualization software and then start coding and rollout of products and/or services. More than Capital needs, this process was time-consuming and change averse. It took 10 to 12 weeks to get the infrastructure going and once you had infrastructure in place, you could not change the type or scale of the infrastructure on short notice. In other words, infrastructure was illiquid. Convergence of other technological advancement like REST along with cloud has lowered the barrier to writing software in-house. This accelerated the developer’s ability to go from idea to implementation in a matter of days rather than months. Besides the speed of building systems, Public Cloud has improved the economics of enterprises by reducing the risk of failed ideas taking lots of Capital Expenses. Following are top six advantages of Public Cloud for Enterprises.

1. CapEx to OpEx: Getting started with a new idea with a lot less capital is made possible via IaaS and PaaS layers of Cloud. Public Cloud lets enterprises build applications in small increments, as ideas get traction, more infrastructure can be added on as-need basis. Unlike old stacks, you do not have to sit on idle infrastructure and wait for application users to come.
2. Agility: As described earlier, cutting down time for procurement for IaaS from 8 to 10 weeks into minutes, saves great amount of time and money for the organization. It increases time to market thus increases ROI.
3. Elasticity: Pay for what you use model makes the application infrastructure expenses proportionate to its usage patterns. Spend on compute and storage increases only if greater number of users are using the application. It helps enterprises double down on ideas which work and shed the ideas (without dire financial consequences) when ideas do not work.
4. Ecosystem Leverage: Public Cloud helps enterprises leverage a wide variety of tools and skillsets via Cloud Marketplaces and Cloud Services Provider (CSP) facilitated Partner Ecosystems.
5. Trained Workforce: Public Cloud promotes standard practices through time tested frameworks and platforms. CSPs invest a lot in skills development of workforce to promote their stacks and all that investment by CSPs works to the advantage of the technology consumers/Enterprises.
6. New Business Models and New Markets: We have seen repeatedly that Public Cloud helps in the introduction of new business models without incurring heavy CapEx. Public Cloud’s distribution model via CSP’s CDN and/or CSP’s organic reach helps an Enterprise reach new markets in emerging economies, especially where a big part of their products and services are digitally delivered.

We have seen repeatedly that Public Cloud helps in the introduction of new business models without incurring heavy CapEx.

CLOUD USAGE PATTERNS?

For enterprises, internal software development has always been a key piece of their applications strategy, be it for extending functionality of their vendor-provided software applications/platforms or writing newer in-house applications from scratch. All enterprise application workloads can be placed into three categories, Systems of Record, Systems of Engagement and Systems of Innovation. Distinct parts of the public cloud play different roles in these three categories. From my experience of working with hundreds of enterprise Public Cloud customers, some patterns have emerged. First pattern is that in the short term, most Systems of Record are moving to the Public Cloud into two categories: Managed Cloud Services (sometimes it’s provided by the vendors which provide system of record software solution) and Public Cloud Infrastructure as a Service (mostly moved as-is migrations). Peeking into the future tells us that eventually, Systems of Record will be procured as Software as a Service. So, I believe that for Systems of Record IaaS part of Public Cloud is an interim stage. Second pattern is that most Systems of Engagement are moving into consuming a sizable chunk of Platform as a Service tier of Public Cloud along with consuming IaaS in Public Cloud. Remember that Systems of Engagement applications tier has a mix of vendor provided applications and home-grown applications. Third pattern is that a large majority of Systems of Innovation are mainly leveraging core parts of Infrastructure as a Service (also referred to as Cloud Primitives), mainly due to the nature of the tooling and platforms needed by these new breeds of software systems.

?

CLOUD CONSIDERATIONS FOR ENTERPRISES

Throughout all my engagements in Cloud, it has been observed that without having clear goals, moving to Public Cloud can prove to be expensive and a frustrating experience. Below is a list of major reasons/factors for evaluating and moving to Public Cloud Service Providers.?

• Mergers & Acquisitions.
• Datacenter and Applications Consolidations.
• Developer Productivity.
• Enablement of New Business Models.
• Entering New Markets.
• Improving Price to Performance Ratio of Infrastructure AKA Lowering Cost.
• Leveraging Legacy Investments via Enablement of Hybrid Cloud Architectures.
• Leveraging Existing Skills of Practitioners while Moving to Public Cloud.
• Migration Costs
• Using Specific Compute Instances for Specific Workloads. (e.g. PIC, Performance Intensive Computing). To increase the portability of workloads, a special consideration is needed for this factor. If the choice of hardware selection is left to the Public Cloud Provider (CSP), they may pick the hardware which suits their goals of ensuring higher margins (or reducing management costs), whereas enterprises may want to repatriate (bring back to their own data centers) their workloads at a later stage. Intel can help enterprises pick an appropriate architecture through their Cloud optimization services which will ensure better price to performance ratio for short term as well as long term.

CLOUD, FUTURE TRENDS IN ENTERPRISES

Current observation from the market is that the use of Multi-Cloud architectures is on the rise and most enterprises are embracing a multi-cloud strategy. Performance Intensive Computing, along with the specialized computing needs, Artificial Intelligence (AI) and Machine Learning (ML) which should be considered when picking a Public Cloud Provider. Far and few enterprises are getting their feet wet with Web3, Crypto, and Decentralized Web technology stacks. Public Cloud Providers are bound to play an important role in Edge Computing via Metro Datacenters and/or via Cloud in a Box offerings for Private Datacenters. Due to rising spend on Public Cloud Services, a new discipline of FinOps is being introduced by forward-looking enterprises, the goal is to contain the costs of Public Cloud.

Another area to watch in Public Cloud is the mapping of the right workloads to the right processing power. This lowers the cost to Cloud Providers and their customers. A recent IDC study, sponsored by Intel? shows that organizations will realize benefits worth an average of $245,300 per developer ($8.77 million per organization) by using?Intel? Optimized Software. These benefits come from performance and functionality enhancements, resulting in infrastructure cost savings, development productivity gains, and more efficient and effective use of open-source software to generate improved business results. This total savings amount includes revenue and productivity gains worth an annual average of $116,500 per developer ($4.17 million per organization), IT efficiencies and productivity gains at an average of $65,700 per developer per year ($2.35 million per organization), IT infrastructure cost savings that IDC values at an annual average of $59,100 per developer ($2.11 million per organization) and realizing productivity and revenue gains worth an annual average of $4,100 per developer ($145,500 per organization). This is huge and Intel? has made a considerable amount of investment in this area, as you notice, investments are paying off.

CONCLUSION

Despite all the hype and capabilities of Public Cloud, we have merely moved less than 20 percent of total enterprise workloads to the Public Cloud. So, there is a long way to go.

As described above, there are enormous benefits of Public Cloud for Enterprises. Having said that, there are a few hurdles and cautions as well. Public Cloud is best fit for building Systems of Innovation and Systems of Engagement BUT latency between Systems of Record and other new types of systems is a key consideration. Besides data latency concerns, existing skills gravity is still a huge issue. People are afraid to move to new skills as they are used to performing their day-to-day tasks in a certain way. Finding new talent who are experts in Public Cloud technologies is also a challenge.?

Technology is like Medicine, every pill has a side-effect.

I usually say that Technology is like Medicine, every pill has a side-effect, Public Cloud is no exception to this rule. To reduce the side-effects of the Public Cloud induced change, we must keep the existing skillsets of the workforce in mind as well. You should maximize the leverage from your existing systems. Also, enterprises must master what I call “micro-consumption” mechanisms when procuring new technology, which means bringing the technology change into the organization/enterprise in micro-chunks, and bringing it often. Sticking to familiar, yet Public Cloud-based technology architectures is the best approach to usher in the next wave of computing. Change is inevitable, but we must make it digestible by making it people-friendly as well as economically feasible!

Happy Clouding!

[ THIS CONTENT IS SPONSORED BY Intel?]