IBM’s New IBM z16 – Meeting and Anticipating Critical Enterprise Requirements, Again

The past two decades have been a time of revolutionary technological change in virtually every region and sector, but it is hard to think of any area where things have shifted more radically than in enterprise data centers. Consider that in 2002, large businesses could choose between over half a dozen discrete server architectures and CPUs. While x86-based systems were on the rise, most were aimed at lower-end applications and workloads.

Today, only three of those platforms are still being actively developed: IBM zSystems mainframes, IBM Power Systems and x86.

The massive popularity of x86 servers inspires regular questions about how and why IBM solutions continue to thrive and dominate their target markets, particularly IBM zSystems. There are good technological answers to that point, including IBM’s continual evolution of mainframe features and capabilities, and its sometimes-counterintuitive investments in new areas, like Linux and open source. But focusing on technology alone overlooks a vital factor—the degree to which IBM understands its customers’ current business needs, anticipates new challenges and develops zSystems mainframe solutions to address those points and concerns.

The new next generation IBM z16 solution announced on April 5^th is a great example of how this dynamic works.

Responding to mainframe customers’ challenges

Before delving into the new IBM z16, how have previous IBM zSystems systems addressed enterprises’ existing and emerging needs? Let’s consider three specific areas:

• Advanced analytics – Though the IBM mainframe has long been the market’s preeminent transaction solution (underscoring its wide use in banking, financial services, airline reservations and insurance), the company has invested substantially to enhance those processes with advanced analytics. One example is IBM’s 2010 acquisition of Netezza, resulting in the IBM Db2 Analytics Accelerator (IDAA) that integrated IBM Netezza 1000 analytics appliances with IBM zEnterprise EC12 mainframes which customers could use to combine traditionally separated OLTP and reporting systems. Further developments, like the IBM z14’s support of up to 32TB of system memory and up to 16TB per LPAR, substantially speeded in-memory databases and other analytics workloads.
• Security improvements – Given the sensitivity of much or most of the data relating to financial transactions, IBM has always accentuated leading edge security in its IBM zSystems mainframe and IBM LinuxONE systems. For example, last generation IBM z15 systems expanded the Pervasive Encryption features first offered in the IBM z14 (which supported extensive encryption of data in-flight and at-rest) with advancements like IBM Data Privacy Passports. That is a data-centric audit and protection (DCAP) solution that protects and enforces appropriate use of data after it leaves the system of record, reducing the risk of security breaches and helping clients meet data compliance obligations.
• Resiliency enhancements – A significant reason for IBM zSystems’s popularity in banking and other financial services where time is literally money is the platform’s essentially bulletproof resiliency, supporting up to “seven 9s” (99.99999%) or no more than 3.15 seconds of unplanned downtime per year. Resiliency depends on interwoven functions, including system recovery, business continuity and secure storage of key data and files, processes that IBM is continually improving. For example, IBM z15 solutions were the first to support Instant Recovery which utilized technologies such as Speed Boost, zIIP Boost and IBM GDPS (Geographically Dispersed Parallel Sysplex) to speed system recovery and reduce the financial and operational impacts of outages.

These and other prior zSystems mainframe features and functions set the stage and the bar for IBM’s new solutions.

What IBM’s z16 delivers

The new IBM z16 mainframe delivers the sort of next gen functional boosts over past gen systems that we have come to expect, including improved capacity (25% per drawer and 40% in max configurations), 35% faster system restart and 11% per core performance improvement. Plus, the company has cast a wide net in terms of new features and functions, including those enhancing DevOps processes and hybrid cloud deployments.

Frankly, there is too much to unwrap in a single commentary, so let’s focus on the three areas I discussed concerning past IBM zSystems systems:

• Advanced analytics – The new IBM Telum processor IBM announced last August will make the IBM z16 the first system in the industry to support on-chip AI acceleration. What does that mean in practical terms? That IBM customers will be able to significantly increase the rapidity of analytics-dependent processes like critical decision making and watching for/catching problems like fraudulent transactions in real-time and at scale. What does “at scale” mean here? IBM claims that the IBM z16 can process up to 300 billion inference operations per day with 1 millisecond response time. In other words, in 1/100^th of the time it takes you to blink your eye, an IBM z16 can analyze and approve/refuse credit card purchases, money transfers and other financial transactions. Moreover, this function can be used with existing training models and also supports building and training new models using open source tools.
• Security improvements – The z16 is the first commercial IBM system to use advancements the company has made in quantum computing. These include Quantum-safe Secure Boot which protects system firmware against quantum attacks and supports up to 19 billion transactions per day with Quantum-safe encryption. This latter feature should substantially boost customers’ ability to secure transactions against existing dangers while also offering protection against emerging “harvest now/decrypt later” schemes designed to steal valuable data so it can eventually be decrypted by other quantum systems.
• Resiliency enhancements – With the IBM z16, IBM is significantly expanding the on-system features included in the IBM z15 to embrace multiple sites and data centers. For example, production capacity can be shifted from a IBM z16 at a primary data center to a IBM z16 at an alternative location in seconds, improving responses to both planned and unplanned outages while still meeting SLAs. Workloads moved to other locations can also successfully remain there for extended periods of time which should help highly regulated businesses address compliance requirements. Mainframe customers can also use IBM GDPS to move production systems in as little as two hours, impacting and improving operational procedures.

Final analysis

As in past IBM zSystems development processes, the company spent substantial time (1,100 hours during IBM z16 development) with mainframe customers and partners assessing their requirements, what they believed needed to be improved and what emerging issues and threats concerned them the most. Including those discussions in the planning process is one of the reasons that IBM zSystems has long been and will likely remain one of the industry’s most customer- and user-focused data center platforms.

It is also an important point to remember when considering the IBM mainframe’s remarkable longevity, and why IBM zSystems continue to be strong contenders in crucial global markets and remain vital solutions for many of the world’s largest and most influential enterprises. Overall, IBM’s new IBM z16 qualifies as a significant advance from the prior generation mainframes and provides new features and functions that will be especially valuable to IBM’s enterprise customers and partners.