Deployment
The wrong choice here could lock an enterprise into an architecture that is
dependent on specific versions of its subsystem, and really limiting the role
and value of the mainframe in term of integration. MIPS control and upgrade
avoidance are the order of the day and really need to give deployment options
a high priority to keep costs in line. The alternative could be as bad as increased
costs, or as dramatic as total failure. The proverbial lose-lose.
It is worth noting that Linux on System z holds tremendous potential in this
areas and Linux on System z deserves investigation as the place to house this
workload to ensure a high-performance, low-cost execution option.
Over the past decade, three specialty engines - Integrated Facility for Linux
(IFL), System z Application Assist Processor (zAAP), and System z Integrated
Information Processor (zIIP) - have evolved, each one optimized as a compelling
alternative to attract specific new workloads.
They offload the processing of applications from the mainframe's General Purpose
Processor (GPP) and run the target workload more effectively. Specialty engines
also change the economics by providing a life-cycle cost of the solution that
is substantially lower than the cost of traditional mainframe software. In short,
shifting transactional workloads from the GPP to specialty engines is a smart
move - one that dramatically slashes the costs of mainframe integration within
modern computing initiatives.
Specialty engines save money by allowing the enterprise to leverage mainframe
computing power for new workloads and specific types of processing without substantially
impacting the software costs associated with running and managing the mainframe
environment.
By hosting specialized workloads, specialty engine processing is excluded from
the overall MIPS or MSU ratings that determine mainframe software costs. Also,
any work running on specialty engines frees up cycles on traditional processors;
which stretches out software maintenance cycles, and, of course, reduces long-term
costs.
These reductions in CPU consumption and associated costs, as well as the strong
performance gains offered by specialty engine exploitation, tell a compelling
story - in effect, run intensive workloads on a specialty engine and not only
save money, but achieve higher performance while enhancing the relevance of
a core computing platform.
-1-