Talk to any IT administrator about what they look for in a server platform and performance, openness, and security will top the list. Specifically, run my most demanding applications fast, enable my organization to support the range of workloads that power my modern business, secure my environment and protect my data. Ask those same IT administrators who can deliver on these and a combination of vendors will be quoted back, with a third party integrating. However, I believe these characteristics are, in essence, IBM’s z15.
Dispel the outdated notion of mainframe technology. I believe IBM Z is a new class of compute platform that is open and well suited to drive the digitally transformed business. In one system, legacy back-office co-resides and integrates with cloud native. Likewise, structured and unstructured data types are integrated to deliver all of the necessary intelligence to the business in a secure, cloudified way. And the z15 is the latest generation platform.
Where the IBM Z” brings it”
Before getting into the workloads supported, let’s cover what I think are the four features that make IBM Z a smart choice for any enterprise IT organization.
Reliability: There is a reason why IBM is the core backbone to the industries that make such large investments in the management and utility of data. Banking, financial services, and healthcare to name just a few. It’s because of the confidence around the availability of systems (and the data residing on these systems). Seven 9s availability, serviceability without disruption of operations, and consistency of performance all equals SLA adherence that IBM Z can deliver.
Performance, and consistency of performance: organizations need to be able to rely on their systems to be both performant and consistent in that performance profile. In multi-tenant virtualized and cloud environments, this concept is easier to promise than to achieve. Because of the IBM Z architecture and system software, features like VM isolation deliver not just security, but protection against performance disruptions often associated with “noisy neighbors.”
Security: Security is more complex than locking down perimeters or user access management. It’s about protecting the systems and data that make the business run. Data must be protected in all phases of its existence – at rest, at work, and in flight. IBM’s pervasive encryption is a hardware based crypto engine that ensure data is protected both at rest and in flight with zero impact to system performance. And Data Privacy Passports is an IBM technology that allows an organization to protect data even as it is shared cross-organization and cross cloud with partners, customers and others. Imagine being able to revoke access to sensitive customer information that was shared with a third party directly from your console? This is Data Privacy Passports and for any company that is concerned with Sensitive Personal Identifying Information (SPII), this feature alone should make the z15 from IBM worth evaluating (I detail this and more in this review of the z15).
Openness: Mainframes populate the vast majority of enterprises and house the applications and data that are considered mission critical. At the same time, these systems and the applications and data housed on them must be able to integrate with newer application and data architectures – and span the on-prem to cloud compute model employed by virtually every company. Throw in the thousands of virtual servers and countless containers that drive the everyday functions of the enterprise, and IT becomes very messy.
Contrary to the outdated view some hold of mainframes, IBM Z is perhaps the most open compute platform on the market as it can support this very messy environment. Legacy? Check. Linux? Check. Containerized? Yup. Virtualized? Yup.
So, what are those workloads?
Based on what IBM Z delivers in terms of performance and security, it’s natural to assume that any workload would benefit. And this is pretty much true. For organizations that run a mixed environment, there are a few workloads in particular that really shine on the Z system, even if not obvious to an IT administrator. Transaction processing and analytics processing (OLTP, OLAP), real time analytics, and massive container and VM consolidation are ideal workloads for IBM Z. Let’s dig into these a little bit.
On-Line Transaction Processing (OLTP) and On-Line Analytics Processing (OLAP): Real time responsiveness to customer orders and inquiries is critical to customer retention. And in this digital economy, “real time” means, in fact, real time. Because of this, the infrastructure supporting these workloads must be able to vertically scale with needs of the applications that serve customers. This is true for both long-term incremental growth and shorter term “bursty” and elastic requirements that are more event driven.
Further, the ability to scale appropriately must not disrupt the current application performance. The abstraction of compute from I/O in the Z system allows for this “scale without disruption” requirement. And Z CPU performance (up to 5.2GHz) with access to massive memory delivers on the requirement for real world performance.
Finally, the Z system tool chain enables IT organizations to better integrate the multiple data types coming from multiple sources. COBOL and Java are not mutually exclusive in the datacenter and any IT professional or data architect that has worked on a project of scale knows that the ability to seamlessly integrate across legacy and modern is an absolute necessity to successfully drive digital transformation.
Real Time Analytics: “Real time” is the operative element of real time analytics. In order for real time analytics engines to be effective, they must be able to access multiple data sources simultaneously. And in the age of edge and cloud computing, security is a must.
While the z15 hardware specifications are impressive, what I find most interesting about IBM’s approach to enabling real time analytics across the enterprise is the tools developed for IT organizations. The company developed tools to execute real time queries against standard SQL environments and the mass of unstructured data flooding the business, and provides an environment where data scientists can quickly develop and deploy on the tools they are familiar with – Java, Scala and Python with Apache Spark, Anaconda, and the like. Sounds like a win.
Enterprise VM Consolidation: Consolidation of Linux VMs on the z15? Serious? Yeah. Running Linux on IBM z15 enables organizations to literally consolidate hundreds of Linux based VMs on a single server, reducing the space requirements and management headaches enterprise IT faces on a daily basis, while at the same time cutting roughly in half the power required to run the same workloads. What is LinuxONE? Think the z15 running Linux to support your enterprise needs. It’s worth checking out.
IBM makes some incredible claims around z15 supporting virtualized environments – 7x better per core performance (versus “x86”) and 4.8x better workload performance with z/VM memory. What this adds up to is near native performance. But equally compelling to me is the way this virtualized environment is made more performant and resilient through the VM security and isolation that results from IBM’s Pervasive Encryption technology, disk mirroring and overall system design.
Am I forgetting a couple workloads?
You may notice a couple of obvious workloads missing from the above list. Specifically, digital asset management (DAM) and Artificial Intelligence/Machine Learning (AI/ML). I really wanted to highlight a few of the workloads that were maybe not the obvious candidates for IBM’s Z system. However, the support for DAM and AI/ML has never been stronger on the z15.
The integration of compression and pervasive encryption combined with Data Privacy Passports make the z15 the most formidable IBM Z platform for digital asset management and custody as they enable the highest levels of asset protection without compromising on performance.
Likewise, IBM has long been on the forefront of the artificial intelligence (AI) market (raise your hand if you’ve seen a Watson commercial). The company has built a framework and toolchain that enables enterprise IT to quickly deploy Watson machine learning (ML) on IBM Z to better enable real-time operations such as fraud detection for banking and finance.
Enterprises have long viewed the mainframe as critical to the business due to its reliability, resilience, scale and performance supporting line of business applications. As IBM developed the z15 to align the evolving needs of the enterprise transformation, openness and security became key capabilities to expand. And with the z15, the company has done just that. Perhaps this is why IBM saw Z system revenue increase by 60% in its latest quarterly earnings.
The world runs on the mainframe. Whether well-established banks, healthcare providers, government organizations or the partners and customers that digitally interact with these organizations. Even more sophisticated startups and newer companies should consider deploying Z system, as I do believe all will benefit from how IBM has embraced openness and the tightest security in the z15. From a die-hard server-only guy for nearly 30 years, I hope that means something.
Note: This analysis contains contributions from server and compute analyst Matt Kimball.