Information Technology Reference
In-Depth Information
oldest systems are simply removed from service. This strategy maxi-
mizes utilization against cost but increases complexity over cyclic
replacement strategies and can keep some systems in use beyond
their warranty period or expected life. When used in research or
engineering enterprises which demand greater processing power for
specific users or functions, this can be an effective strategy. When
used in politically based hierarchies in which the top dogs get the
best toys, this strategy creates “have and have-not” feelings between
users and develops greater complexity without adding value to the
enterprise. This strategy shares the highest level of user disruption,
because all systems shift one layer down with each cycle.
•
Full pulse
—Enterprises with a defined replacement cycle, typically
linked to warranty or funding cycles, may elect to replace all systems
each time the cycle concludes. This is the least complex enterprise
possible, but almost never occurs beyond very small-scale network
enterprises or those within specialized isolated environments such as
extreme-environment research facilities. This strategy provides all
users with equal technology and affords integrators a simple environ-
ment with a well-defined upgrade term. If the cycle exceeds the limits
of Moore's and Kryder's laws, life-cycle costs increase toward the end
of each replacement cycle. Shortened cycles avoid this and increase
agility, but they increase costs accordingly. This strategy suffers in
budgetary planning, because some years will have no replacement
costs while the end-of-cycle year will require full funding for the
entire enterprise. Similarly, support personnel requirements are high
in replacement years and reduced during the remainder of the cycle.
•
Steady cycle
—As with the full pulse strategy, the steady cycle
replaces systems on a cycle linked to fiscal or warranty constraints.
Unlike full pulse, the steady cycle replaces a portion of the systems
each term, allowing for easier and less variable budgeting across mul-
tiple years for mid-to-large-scale enterprise networks. User environ-
ment disruption is minimized, complexity is minimized within each
term's purchase group, and all systems are replaced on a regular basis
to provide a continually renewing technology base able to imple-
ment planned integration of emerging solutions.
Because of the impact of Moore's law, major software develop-
ment cycles, and standard warranty periods, the most common
cycle for user workstation and laptop equipment employs purchase
