IBM z14: Enforcing Data As The New Perimeter, Enabling Scale For The Cloud

I recently attended an invite-only sneak peak of IBM’s new z14 mainframe solution in advance of the official release on Monday, July 17th…

“Designed for Trusted Digital Experiences” IBM Z allows for continuous intelligence across the enterprise via advanced analytics and machine learning.

IBM presented the following data as a backdrop for our conversation, against which z14 illuminates solutions:

  • Nearly 5.5 million records are stolen per day, 230,367 per hour and 3,839 per minute.
  • Of the 9 Billion records breached since 2013, only 4% were encrypted.
  • 26% is the likelihood of an organization having a data breach in the next 24 months.
  • Over 90% of the world’s data cannot be Googled.
  • Data is becoming the world’s new natural resource, transforming industries and professions, 44 Zeta bytes by 2020.
  • 220+ billion lines of Cobol code, accounting for more than 70% of business transactions.

IBM z14 Enforcing “Data As The New Perimeter”

z14 is designed with the intent of being the world’s premier system for enabling data as the new perimeter across the value chain. Presenting pervasive encryption as the new industry standard, IBM believes that “100% Encryption is 100% Mainframe.”

Here are the top highlights from the z14 presentation:

  • Key encryption: An industry exclusive; protected and enabled through integration with a tamper-responding cryptographic HSM.
  • True end-to-end data protection: All in-flight network data and API’s.
  • Compliance: real time, self service audit verification.
  • 4x increase in silicon area allocated to cryptographic operations.
  • 4 – 7x faster encryption of data with enhanced cryptographic performance.
  • 18x faster encryption than competition at 1/20th the cost to implement.
  • 2x performance boost on Crypto Express6S.
  • Secure the cloud by encrypting APIs 2-3x faster than x86 systems.
  • Linux exploits Protected Key encryption for data at-rest.

IBM z14 Enabling Scale For The Cloud

IBM z14 has made big improvements to scaling and IO throughput to meet today’s and future business objectives while supporting connected mainframe requirements.

The presentation of the new z14 mainframe solution included the following highlights:

  • Transparent Cloud Tiering: a new cloud storage tier on IBM Z; 50% savings in CPU while archiving large data sets, providing multiple cloud vendor support and providing clients freedom of choice.
  • zHyperLink: a new direct link designed to lower latency by up to 10x, cutting I/O sensitive workload response time by up to 50% without requiring application changes.
  • 3x increase for small data transfer I/O operation, with next generation of I/O channel hardware.
  • Automatic Binary Optimizer reduces CPU usage of applications built with COBOL without source recompilation by up to 80%.
  • Scale-out to 2 million Docker containers in single IBM z14.
  • Scale a single MongoDB instance to 17 TB in a single system and get 2.4x more through put and 2.4x lower latency on IBM z14.
  • 50 new instructions optimized for Java, 1.5x faster Java on IBM Z than alternative x86 platforms.
  • Pause-less garbage collection built into the processor, reducing pause times by up to 10x for predictable high-performing transaction processing at-scale.
  • Enhanced math libraries support through SIMD providing performance improvements for analytics workloads.
  • Up to 32 TB memory (3x) in support for new workloads and in-memory databases; efficiently process huge amounts of information for real-time business
    insights.
  • Leverage and extend existing investment by automatically identifying API candidates and quickly turn them into services available as an API. Quickly identify which API’s are impacted by changes to an application before it goes into production reducing application down time and re-work.

IBM z14: Sexy Tech

IBM z14 introduces new hardware, firmware, operating system, middleware and tooling in an effort simplify and enable systems management for administrators with minimal IBM Z knowledge.

Processor Units (PUs):

  • 41 (49 for M05) PU cores per CPC drawer
  • 33, 69, 105, 141 or 170 PU cores available for characterization
  • Cores 1832 MIPS
  • Up to 23 SAPs per system, standard – New on z14 is SAPs are SMT
  • 2 spares designated per system
  • 85 LPARs
  • Sub-capacity available for up to 33 CPs – 3 sub-capacity points
  • Enhanced performance for compression and crypto coprocessor

Memory:

  • RAIM Memory design
  • System Min of 256 GB – up to 8 TB / drawer – 192 GB Fixed HSA, standard
  • Up to 32 TB for System and up to 16 TB per LPAR (OS dependent)
  • IBM Virtual Flash Memory (replaces Flash Express)

I/O:

  • New PCIe Gen 3 IBM zHyperLink technology
  • 3 PB of all flash storage (More than 3.8x than previous generation in the same footprint)
  • Last high end server to support InfiniBand Coupling Features
  • 16 GBps PCIe Gen 3 I/O Interconnects
  • 6 Logical Channel Subsystems (LCSSs) with 4 Sub-channel sets per LCSS

IBM z14 10-Core Processor Chip Detail:

  • Up to ten active cores (PUs) per chip
    • 5.2 GHz (versus 5.0 GHz IBM z13)
    • L1 cache and L2 cache on core
    • L3 cache on chip – shared by on-chip cores and communicates with cores, memory, I/O and system controller single chip module
  • New instructions for Single Instruction/Multiple Data (SIMD)
  • Single thread or next generation 2-way simultaneous multithreading (SMT) operation
  • Guarded Storage Facility (GSF)
  • Compression enhancements – Huffman Coding
  • Improved instruction execution bandwidth:
    • Pipeline optimization — Improved instruction delivery, faster branch wakeup, optimized hardware/millicode interaction, reduced execution latency, improved OSC prediction
    • Translator / TLB enhancements
    • New instructions for old and new workloads – i.e. Vector BCD arithimetic (COBOL)
  • I/O buses
    • One InfiniBand I/O bus
    • Two PCIe I/O buses
  • Memory Controller (MCU): Interface to controller on memory DIMMs, supports RAIM design
  • 14nm SOI Technology
    • 17 layers of metal
    • 6.1 Billion Transistors vs 3.99 Billion on z13
    • Chip Area: 26.5 x 27.8 mm

Final Thoughts

IBM’s Z14 is poised to provide businesses with significant relief from client data compliance issues with industry leading scale and new levels of agility. Application response times have been cut in half and service build time has been reduced 90% using secure APIs on IBMz and advanced DevOps. In addition, container pricing for IBM Z provides new levels of flexibility for modern digital workloads.

IoT represents enormous scale and impact for connectivity requirements with existing transactional systems, and IBM’s z14 seems set to make its mark.

Thanks for reading!

Don’t forget your free subscription to Quotidian Disruption: “Disruption News Daily.”