IBM is growing its mainframe business with a new machine targeted at cloud, analytics and mobile computing use-cases.
The new zBC12 mainframe is a step up from IBM’s previous entry-level mainframe, the z114.
Kelly Ryan, director and business line executive, IBM System z, explained to ServerWatch that the zBC12 machine is available in two models: the H06, which is a single central processing-drawer model, and a two-drawer model, the H13, which offers the additional flexibility for I/O and coupling expansion and/or increased specialty engine capability.
The zBC12 is powered by up to 18 microprocessors running at 4.2 GHz, and it delivers up to 36 percent improvement in performance per core compared to its predecessor, the zEnterprise 114 (z114).
From a processor perspective, the zBC12 leverages the same microprocessor technology as the zEnterprise EC12 (zEC12) but is tuned and optimized for the processing requirements of the entry-level mainframe.
“The core design combines all the best qualities of the latest mainframe design philosophy with a generous infusion of POWER7-inspired computational muscle, a reflection of the teamwork between IBM’s best processor designers,” Ryan said. “The new microprocessor chip has a higher-frequency superscalar design, improved cache structure, second generation out of order execution sequence and is optimized to provide software performance improvements for Java, DB2 and more.”
IBM’s zEnterprise servers use a type of memory configuration known as RAIM (redundant array of independent memory).
“RAIM is similar to what is known in the disk storage industry as RAID,” Ryan said. “RAIM technology provides protection for the dynamic random access memory (DRAM), dual inline memory modules (DIMMs), and at the memory channel level, delivers the most resilient memory subsystem to date.” According to IBM, the zBC12 supports up to 512 GB of real (usable) RAIM-protected memory.
The zBC12 will offer a specific solution to customers running what IBM calls the IBM Enterprise Linux Server (ELS). Ryan noted that the IBM ELS is a proven platform for Linux workloads and consolidation, and its life-cycle management costs can be considerably less expensive than other Linux platforms.
According to Ryan, ELS has the capability to run a large number of parallel workloads – up to hundreds – in a highly efficient and economical way. It can share system resources at extreme levels of utilization and provision flexible IT services.
“It provides load-balancing and efficient systems management, helping to achieve superior levels of service and improved operational efficiency,” Kelly said.
In contrast with IBM’s POWER Server systems portfolio, Ryan doesn’t really see a comparison.
“Though from a thousand-foot level it looks like there may be a comparison, it’s really like comparing apples and oranges,” Ryan said. “The zBC12 $75k entry point is for a 50 MIPS entry point configuration in support of z/OS or VSE operating environments. IBM Power Systems does not run these workloads, so really, there is little comparison. “