Hybrid Multi-Cloud – The New Reality

We hear a lot of buzzwords in the industry right now, almost as if the buzzword engine is running on jet fuel.  In your role as a senior leader in the IT function, how do you cut through the noise and find the signal from that static that can help you make those pivotal decisions that can take your organization to the next level?  As I engage with clients around the globe one key theme is emerging.  Digital Transformation.  The clients I engage with are struggling to drive a digital transformation, most times have driven be their line of business executives, against a backdrop of technical debt and then COVID comes along to further pile on.

One topic that comes up as part of this wider Digital Transformation conversation is the journey to the cloud, not as a destination, but more as a mantra for how to deliver IT in the most performant, scalable, and increasingly importantly cost-effective way.  Every client I speak to is grappling with how they embrace cloud-native containerized workloads as a key technology to empower their transformation.  This is less about Kubernetes as a tech stack, more about the flexibility that can be delivered by a microservices architecture.  A key theme in this cloud journey can be put simply as “what do I put where?”  while containers and cloud orchestration tools powered by Kubernetes deliver this capability, they do not provide the insight into the fundamental questions of which cloud and when to go off-premises and when to stay on-premises.

A key question to ask to better frame the discussion is - When driving workload based on an SLA, which platform delivers the best cost? 

Recently IBM’s IT Economics and research walked me though some detailed analysis that the team did to try and come up with an IBM point of view to this question.  The results were enlightening to me so let me share some of the feedback.  The team focused on Online Transaction Processing (OLTP) workloads to better understand the core challenge faced by clients when it comes to their mission-critical workloads.  As you can see from the chart below IBM internal tests show that the same OLTP workloads on OpenShift require 17 times fewer cores on LinuxONE and deliver a 48% lower TCO over three years than compared x86 servers.

Let us unpack this for a moment.  The conventional wisdom is that commodity x86 servers are cheaper than high-end mission-critical servers such as LinuxONE, however, the results of this detailed analysis speak to the contrary.  The detailed analysis compared 7 x86 servers (totaling 138 cores) against 8 processors on a LinuxONE and found that the overall TCO over a 3-year period to be 48% cheaper on a LinuxONE platform.  How can this be, I hear you ask?

While conventional wisdom is a powerful force, as leaders we need to question groupthink and evaluate all the options on their merits.  LinuxONE represents a different choice.  The IBM designed and manufactured chip architecture enables Linux workloads to typically perform at 2x what can be found on commodity x86 architectures.  This is driven by the world's fastest commercially available chip clock speed of 5.2 GHz and a huge cache nest of almost 1tb that surround the chip.  When coupled with a dedicated I/O subsystem (something not found in the x86 architecture) this enabled LinuxONE to handle huge volumes of OLTP workload and not break a sweat.

LinuxONE the backbone of your private cloud

Raw server performance is not enough though.  To operate as a top tier participant in a hybrid multi-cloud strategy a platform needs to be able to deliver as an IaaS or PaaS layer in a wider cloud strategy. LinuxONE can deliver in this regard.  With Red Hat OpenShift providing the consistent PaaS layer for Kubernetes orchestration and management the story for LinuxONE is complete.  

Clients are talking to me every day about their desire to simplify management and consolidate their infrastructure.  This leads to engagements where my team focuses on workload placement and the total cost of ownership for respective platforms.  In these discussions, I encourage my teams to get their clients to look beyond the total acquisition of the hardware and think more broadly on what each platform takes to operate on a holistic level over a 3 or 5 year period.  Only when my team work with clients on a holistic level do all the costs of operation manifest themselves.  Too narrow a perspective on upfront cap-ex investments leads to eventual frustration on the cost of operation.

Summary

From my global perspective and my daily engagements with infrastructure executives in the world’s largest organizations, I am fortunate to get a unique perspective on the journey to cloud these leaders are making.  The best of this select bunch are willing to look beyond conventional wisdom and industry dogma to gain competitive advantage.  An increasing number are deploying LinuxONE and Red Hat Open Shift as the bedrock for their hybrid multi-cloud.  To find out more about why click here.