GEB1: Diminishing Manufacturing Sources and Material Shortages (DMSMS)

Management Practices
Henry Livingston
Vice-Chair, Government Electronics & Information Technology Association (GEIA)i G-12 Solid State Devices Committee
E-mail: [email protected]; Internet: http://home.att.net/~cetoolbox/

INTRODUCTION The Government Electronics and Information Technology

Association (GEIA) G-12 Solid State Devices Committee developed
Diminishing Manufacturing Sources and Material Shortages a set of DMSMS management practices that can be used by original
(DMSMS) is an increasingly difficult problem for DoD weapon equipment manufacturers (OEMs) during the design and
systems because the manufacturing lives of many critical items get development of electronic systems to mitigate the effects of
shorter while the life cycles of military weapon systems keep DMSMS. This paper provides an overview of EIA Engineering
increasing. Traditionally, efforts to mitigate the effects of DMSMS Bulletin GEB1, Diminishing Manufacturing Sources and Material
have been reactive; that is, the effects are addressed only when they Shortages (DMSMS) Management Practices.
are seen. This reactive approach to DMSMS solutions leads to
decisions that put a premium on faster solution paths with attractive MICROCIRCUIT LIFE CYCLES
short-term gains in order to avoid system inoperability, while
ignoring the long-term solution paths that would lead to generic Life cycle curves provide a basis for developing forecasts and
families of solutions or larger-scale solutions with the capability of planning for technology insertion and product update opportunities.
avoiding future DMSMS issues. In order to solve DMSMS issues GEB1 uses the definitions in EIA-724, Product Life Cycle Data
with lower overall cost, DMSMS solutions must change from Model, for the purpose of describing the life cycle of
reactive to proactive. The building blocks of effective proactive microelectronics. The figure below presents a summary of the
management of DMSMS are established during the design and characteristics associated with each life cycle stage.1
development of systems. If systems are designed with the
inevitability of DMSMS in mind, early solution paths with large- Not all devices conform to the six life cycle stages described in
scale solutions can be started at an appropriately early time to enable EIA-724.2 Some devices undergo a false start and die out, or may be
intelligent choices without the imminent threat of system associated with a niche market. Some devices may be revitalized
inoperability. Such generic large-scale solutions and a consensus on after the decline stage. Other possibilities can also arise due to
where DMSMS threats are most prevalent can be better forecasted by various economic, social, and environmental occurrences. A false
the use of a standard set of DMSMS management practices used by start typically suggests that a device starts out with a strong period of
the foremost members of industry. growth only to stall because of one or more of the following factors:

Product Life Cycle Model

Stage 1 Stage 2 Stage 3 Stage 4 Stage 5 Stage 6

Introduction Growth Maturity Saturation Decline Phase-Out
Increasing Lifetime buys
Sales Slow increase Stable Leveling out Decreasing
rapidly offered
Price Highest Declining Stable Stable Rising High
Usage Low Increasing Stable Stable Decreasing Decreasing
Part Periodic
Frequent Major Few Few or none None
Modification changes
Decline
Competitors Few High Stable number Declining Declining
begins
Reasonable for
Manufacturer Reasonable
Low Increasing Stable Stable survivors
Profit for survivors
(aftermarket)

i
A Sector of The Electronic Industries Alliance
1 of 11
 Development
• Introduction of a superior competing part Start
Projected Lifetime Extended Life
• Improvement of a competing part
• Identification of a technical problem associated with the part 1969 1973 F-14 2010+

• Failure to reach the critical mass to allow economies of scale to 1955 1959 UH-1 U N IT
E D S T AT
E S A
R M Y

2004+

be realized 1969 1975 F-15 2010+

• Lack of a unique and compelling application for the part

1970 1976 SSN 688 2026+

Historical data from industry studies shows that the average life
1953 1955 AIM-9 2025+
span of total microcircuits across all quality ranges is around 10
years overall.3 Military microcircuits average greater than 12.5 years 1954 1957 KC-135 2017+
while commercial microcircuits (including “mil-temp-only,”
industrial, and commercial temperature devices) average less than 8.5 1946 1955 B-52 2040+

years. Certain microcircuit families have different life spans; for

0 Year 50 100
example, certain linear devices average less than 14.5 years while
some microprocessors and memories average less than 5 years. The
following shows the average introduction rate for new generations of
commercial integrated circuits:4 PROACTIVE DMSMS MITIGATION APPROACHES

Average The following describes a variety of methods that can be applied

Device Category Introduction Rate during system design to minimize the impact of future component
obsolescence issues.
Logic Families 6 years
Memory Families 9 months
System Architecture Approaches

Microprocessors 2 years The best known aspect of technology obsolescence is

nonavailability of parts as vendors move on to newer technologies
Digital Signal Processor (DSP) 3 years
and products. A number of other factors contribute to the overall
Programmable Logic Device (PLD) 1 year problem, including:
Linear Interfaces 8 years
• Inadequate system partitioning, which often means that a specific
Gate Arrays 2 years part problem ripples out to affect a larger hardware ensemble,
making the solution more expensive by increasing amount of
hardware which must be redesigned.
Observations from various industry sources indicate that these • The use of commercial off the shelf (COTS) products, which is
average introduction rates have gotten shorter in more recent years. attractive as a way to use successive compatible generations of a
product line but is complicated by the requirement to make such
WEAPON SYSTEM LIFE CYCLES parts survivable in the military environment.

Obsolescence of subsystems is a risk driver for weapon systems • Selective modifications, which are essential to deal with these
because, as with many low-volume electronic systems, they are problems but can create multiple configurations and thus add to
intended for use over extended time, leaving them vulnerable to the configuration management challenge.
obsolescence of the parts, subsystems, and technologies that • Although digital technology has the fastest rate of change, non-
comprise the system.5 Resources have to be committed and digital areas like RF hardware also present parts obsolescence
combined engineering-business strategies need to be implemented to problems, and finding suitable replacements is more complex
sustain supply of obsolete parts long enough to permit redesign and because of the multiple parameters that must be matched.
requalification.
• Software obsolescence affecting the operating system, the
With the collapse of the Soviet Union, very few new weapon application programming interface (API), and the selected
systems are being built, and existing systems are expected to be the programming language. These are areas where military systems
front line military defense far longer than planned. Many weapon can enjoy savings through use of COTS products, but these too
systems can now expect to see a service life of 40 to 90 years. The B- have finite market lifetimes.
52, for example, will be operational for more than 94 years. This is
long enough to provide the potential for five generations of pilots Modular, open, integrated architecture has emerged as a key
from the same family to fly this aircraft. Technology obsolescence enabler of weapon systems that combine affordable cost with the
studies also note that the time period from the start of design to the ability to deal with rapidly changing technology and evolving system
beginning of production is increasing.6 As the time from design to requirements. Key architectural attributes that minimize parts
production stretches out, many technologies used during the design obsolescence problems include effective implementation of open
are obsolete before production starts. systems principles and independence of specific implementing
hardware.7 The basic acquisition technique is to maintain
technology-independent designs and implement any given hardware
entity with the best products available at the time it is needed.
Similarly, as software tools or other entities become unavailable or

2 of 11
unsupportable, the open system structure allows selective it allows description of the structure of a design (i.e. how it is
replacement with new products. decomposed into sub-designs and how those sub-designs are
interconnected). Second, it allows the specification of the function of
Military electronic systems have traditionally been closed and the designs using familiar programming language forms. Third, as a
largely platform-unique. Today, cost and supportability result, it allows a design to be simulated before being manufactured,
considerations motivate open systems that allow wider use of so that designers can quickly compare alternatives and test for
commercial technology and products, competitive sourcing, better correctness without the delay and expense of hardware prototyping.
software productivity and reuse, reduction in the number of unique Through the use of Computer Aided Engineering (CAE) tools, it is
configuration items, and selective modification and upgrading. Open possible to build detailed and accurate simulation models of
systems are most often defined in terms of support for portability of electronic subsystems. The ability to define hardware in a descriptive
applications software and implementation of widely supported model not only improves development efficiency, it also assists in
standards. Equal or greater importance attaches to technology- migrating existing electronic subsystems to the next generation
independence and modular partitioning so that specific areas can be architecture.8
affordably upgraded, expanded, or replaced.
Many components designed today are for specific applications.
An open system design begins with rigorous functional Each of these high-density components is designed to perform
decomposition of allocated requirements followed by allocation of unique functions specific to a single circuit board or application.
functions to hardware and software components. The partitioning When these devices become obsolete, frequently there are no direct
process continues down to the level of hardware and software replacements for them. This can cause two complicated DMSMS
modules. These modules become the “atomic” building blocks of the problems:
system and the basis for commonality across systems as well as
facilitating selective replacement over the system’s life. Careful 1. Since these devices are typically complex, those without
system partitioning allows the effect of a hardware change to be appropriate documentation are very difficult and costly to
localized to minimize the scope and cost of a system modification. reverse engineer.
Ultimately the key is comprehensive functional specifications that
2. If the functions are known, but not in vendor independent
are truly independent of any particular implementing technologies or
products, but retain the flexibility to consider subtleties like degraded format, a considerable amount of engineering time and cost will
be incurred developing a solution.
operating modes that are closely tied to a specific design.
VHDL allows designs to be more cost-effectively transitioned to
Technology Independence new technologies when the original components become obsolete. A
paper published by the Air Force Materiel Command (AFMC)
Technology independence depends on defining all interfaces such reported significant cost avoidance achieved when replacing
that the individual modules can be redesigned with substitute or discontinued technology.9 For example, where VHDL was used to
upgraded components without impacting their functional interfaces document F-22 ASIC designs, re-partitioning and redesign costs
with other modules. This modular design approach can minimize were approximately half the cost incurred by those who did not
system redesign impact. Economies of scale can be achieved if the document ASIC designs in VHDL.
modules are used across multiple equipment. Technology
independence is especially relevant for modular systems, COTS Programmable Logic Devices (PLDs)
assemblies, systems with a high probability of recurrent
obsolescence, and components for specific applications. Programmable Logic Devices (PLDs) are general-purpose
combinational or sequential digital components whose ultimate
Software interfaces must be implemented in such a fashion as to functions are determined by the designer. They leave the
break hardware-software interdependencies, e.g., the use by manufacturer in an unprogrammed state. The configurations of the
applications programmers of unique features of a particular internal switches are fixed after the particular logic function for the
processor. Similarly, the modular partitioning and functional device has been prepared and checked using CAE tools. PLDs are
interface definitions must be independent of any specific manufactured in most digital technologies: fuses, antifuses, floating-
implementing technology such as a particular semiconductor logic gate, MOSFETs and RAM cells. Floating-gate devices can be erased
family or generation of analog signal processing components. and reprogrammed; RAM-based devices are reconfigured
Software must be implemented in widely supported higher order dynamically. Field Programmable Gate Arrays (FPGAs) combine the
languages with excellent real-time support. integration of an ASIC with the flexibility of user-programmed logic.
FPGAs present the user with basic cells and interconnect resources,
One powerful technique for achieving technology independence which serve as the building blocks for design implementation. Users
is to capture the design of a system in the form of executable specify their design with a schematic or hardware description
simulation objects rather than, or in addition to, the traditional text language. This design is then converted to a vendor-specific format
and figures. Such an executable specification starts as a technology- with the components of the design mapped onto the basic cells of the
independent behavioral model, which reproduces the activity of the FPGA. Once the design has been successfully simulated,
module or other entity under any set of stimuli. It then evolves to a interconnect resources are programmed by the user. FPGAs are
structural model when the entity must be instantiated in a given extremely useful in migrating existing designs to new hardware
technology at a point in time. Modern languages like VHSIC technology and are, therefore, an effective means to mitigate
Hardware Description Language (VHDL) support behavioral, microelectronics obsolescence. Though specific FPGA families are
structural, and mixed hardware modeling. discontinued as frequently as other integrated circuit technologies,
they can be cost-effectively transitioned to new technologies using
VHSIC Hardware Description Language (VHDL) the hardware description from the original FPGA design.

VHDL is a language used to describe digital electronic systems

and is designed to fill a number of needs in the design process. First,
3 of 11
Software Portability There are a number of Source
technical problems in the reuse of Language
Obsolescence of embedded computer systems occurs both in source code on a different
hardware and software. Hardware obsolescence arises as hardware platform. This is largely a High-Level Compiler
components become outdated or are no longer manufactured. software engineering issue,
Software obsolescence arises since software is written for a particular requiring the elimination of all Intermediate
platform (i.e. microprocessor or computer) and cannot easily be non-portable aspects of system Language
moved to a new platform without significant rework and cost. Often, design and implementation.
only a part of the system is actually obsolete. Unfortunately, Typically, source code was Virtual Machine
replacement of part of the system is usually as difficult (and originally designed and Low-Level Compiler
expensive) as replacement of the entire system due to the system implemented with a particular
being developed as a single entity, with much interdependence target platform in mind. Thus, the
between hardware and software. task of porting the code is Executable
complex and costly. Often, the Code
Software source code, such as Ada or C, is not directly executable end result is a system that
by the processor in the target computer. It is turned into an behaves differently from the Target
executable code by a compiler. One problem with conventional original system, creating a host of Processor

compilation is that software is compiled with a specific target in other problems.

mind. This ensures the obsolescence of the software as soon as the
hardware needs to be replaced. Ideally, software should be compiled While it is probable that some software cannot be made entirely
independent of the target. Portable code allows compiled software to portable, typical problems and potential solutions for porting existing
be executed on any platform without change. This enables hardware software source code include:
to be replaced when it is obsolete, but software to be moved to the
new hardware unchanged and, therefore, reducing the cost of Problems vs. Potential Solutions for Porting Software
obsolescence.10 Portable code compilation is broken into two stages:
Problem: Solution:
• High-level compilation: Translation of source code, written in a Hardware specific references:
high-level language, into portable code, expressed in an The software contains Write the software such that all
intermediate language. references to specific hardware hardware specific references
functions, such as memory are made via a look-up table
• Low-level compilation: Translation of portable code into locations or interfaces. (either internal or external to the
executable code. actual software) that can easily
be modified, or are made via the
Low-level compilation translates the portable code, in the operating system, which
intermediate language, into an executable form. This compilation can provides a hardware
occur in four main ways: independent abstraction over all
the devices.
1. Host: The translation is performed from portable to executable Operating system references:
code on the host. The executable code can then be loaded onto The software contains calls to Use an operating system that
the target for direct execution. specific operating system has a standard interface layer,
features, which may not be such as POSIX, so that calls to
2. Target pre-run-time: The portable code is translated into an supported on other operating the operating systems are
executable by the target, prior to executing the code. systems, or later versions of the always the same. While
same operating system. providing a standard interface,
3. Target interpretation by software: The portable code is this is not a complete solution
translated ‘on-the-fly’ into an executable form by the target. because the actual behavior of
Often, this is done on a line-by-line level-i.e. only when a the system may change.
portable code statement is needed is it translated.
Architectural assumption:
4. Target interpretation by hardware: The portable code is Assumptions regarding the Data structures are created so
translated ‘on-the-fly’ by hardware into an executable form. underlying architecture are that data types (such as integer)
made such as the precision are always of a specific size.
Methods 1 and 2 reflect conventional compilation. Methods 3 and and/or representation of data. Then data is always
4, involving interpretation, offer more freedom from obsolescence. For example, an integer is manipulated in a way that
Here the portable code (i.e. intermediate language statements) can be stored as a 32 bit word (i.e. 4 makes no assumptions about
viewed as being directly executed by the target, or at least by a ‘thin’ bytes) in the order of most the way in which data is stored,
significant bit (i.e. the one that i.e. direct memory accesses are
layer of software or hardware directly above the processor. This is represents the largest value) is avoided.
commonly known as the virtual machine approach. The virtual- first.
machine layer effectively incorporates the low-level compiler. The
virtual machine interprets the software for the particular target Timing assumption:
processor. Essentially, the underlying hardware is transparent to the Assumptions regarding the Software should be structured
timing behavior of the system so that timing properties are
portable code, high-level source code and the programmer. To move are built into the software, e.g. a handled by the operating
the portable code to a different platform requires that the new loop that effects a specific delay system, e.g. provision of a delay
platform implement the same virtual machine (i.e. low-level on a particular processor. call, or the use of scheduling for
compiler). concurrent software.

4 of 11
Technology Roadmapping Microelectronics technology roadmaps provide the visibility
necessary to maintain a strategy that aligns product life cycles to take
Technology roadmapping is a specific technique for technology advantage of technology insertion and product update opportunities.
planning which fits within a more general set of planning activities. It Examples of resources available to support technology roadmapping
identifies critical product needs that will drive technology selection activity include:
and development decisions, determines the technology alternatives
that can satisfy those needs, helps select the appropriate technology International Technology Roadmap for Semiconductors (ITRS),
alternatives, and helps generate and implement a plan to develop and published by International SEMATECH, provides excellent top level
deploy those technologies. ANSI/AIAA-R100-1996, Parts information for use in obsolescence management and product update
Management, describes the use of technology roadmaps to minimize planning activities.
risk of obsolescence and to develop a strategy for technology
insertion during the entire life cycle. 1998 Technology Roadmap for Integrated Circuits Used in Critical
Applications (SAND98-1914) published by Sandia Laboratories
The main benefit of technology roadmapping is that it provides provides guidance when developing and applying IC's in critical
information to help make better technology investment decisions. It applications and documents major technical trends impacting critical
does this first by identifying critical technologies or technology gaps ICs and the major technical needs.
that must be filled to meet product performance targets. Second, it
identifies ways to leverage R&D investments through coordinating Integrated Circuit Engineering (ICE) Corporation market research
research activities either within a single company or among alliance analysis and technical publications provide information about
members. Some companies do technology roadmapping internally as semiconductors, semiconductor manufacturing and capital
one aspect of their technology planning. At an industry level, equipment.
technology roadmapping involves multiple companies focusing on
common needs. Further insight into future product offerings and technology is
often needed for more effective planning. Some companies establish
The following table provides an overview of the three phases in nondisclosure agreements with device manufacturers to gain this
the technology roadmapping process. 11 insight. Used in conjunction with component obsolescence forecasts,
technology roadmaps can be an effective tool for design review
activity.
The Technology Roadmapping Process
Phase I: Preliminary Activity Some OEMs hold technology interchange sessions with
semiconductor industry representatives to exchange information
1. Satisfy essential conditions
concerning forecasted equipment application requirements and
2. Provide leadership/sponsorship semiconductor industry technology roadmaps. Such technology
3. Define the scope and boundaries for technology roadmap
exchanges are common in the commercial semiconductor industry
and are becoming increasing popular in the military and aerospace
Phase II: Development of the Technology Roadmap industries. Government organizations occasionally host symposia
1. Identify the "product" that will be the focus of the roadmap such as "Industry Days" that provide OEMs opportunity to hear from
the services and government agencies what the perceived technology
2. Identify the critical system requirements and their targets needs are going to be several years into the future. Extending these
3. Specify the major technology areas Government hosted symposia to include semiconductor industry
representatives might stimulate more interest in supporting the
4. Specify the technology drivers and their targets military and aerospace market, increase willingness to consider
5. Identify technology alternatives and their time lines alternatives for supplying parts to these markets, and encourage
suggestions for improvement to the user community through a true
6. Recommend the technology alternatives that should be
"need versus capability" discussion.
pursued
7. Create the technology roadmap report Technology Insertion
Phase III: Follow-Up Activity
Technology Insertion is a means of dealing with the rapidly
1. Critique and validate the roadmap growing problems posed by DMSMS. Technology Insertion refers to
2. Develop an implementation plan “The introduction of new technology into a given design,
manufacturing, marketing or maintenance process in order to effect
measured change in the performance of that process”. If conducted
effectively, Technology Insertion provides a principal means of
Technology roadmaps are developed from market survey data. In
proactively dealing with obsolescence. 12
addition to its use for keeping abreast of technology advances,
changes and trends, market survey data can also be useful for To properly integrate Technology Insertion activities into the
anticipating product life cycles to assist in upgrade and technology System Engineering process, DMSMS must become an element for
insertion planning. Market survey data can be collected through consideration, analysis and planning beginning with the program
researching trade publications and technical societies, vendor product concept phase. It must play a significant role in the entire Integrated
and technology briefings, and other means. Technology roadmaps Product and Process Development (IPPD) effort. It must be
should be updated regularly to sustain their effectiveness. This is addressed in market surveys and trade studies and in the formation
especially important for memory devices, microprocessors, and other and activities of IPTs. DMSMS must be considered in developing
components with shorter life cycles or limited sources of supply. design guidelines and rules, management and manufacturing plans,
and risk management methodologies. It should be incorporated in
Request for Proposals (RFPs), contract Statements of Work (SOWs),
5 of 11
incentive and award fee elements plus entrance and exit criteria in Moreover, as system parts listed are defined, line items should be
program reviews (particularly design reviews). subject to periodic technology / risk screening to identify existing
and potential out year DMSMS problems.

Parts List Review / Prioritization: Once general DMSMS risk factors

Assess Program Identify Candidate Assess Technology
Maturity Technologies Maturities
Trade Studies have been assigned to system parts lists, a prioritized set of targets
for both reactive and proactive DMSMS analyses may be developed.
All current and near term problems should be slated for immediate
Further Yes investigation, with remaining line items categorized by projected out
Review and Select Periodically
Best Alternative(s) Repeat and Refine
Analysis year availability. It may be desirable to further refine rankings to
Required ?
reflect general engineering judgment, specific item / source risk
No elements as identified during case analysis, or individual item
Technology characteristics deemed appropriate (e.g., criticality, number of
Insertion Solution applications, demand volume).
• Cost avoided
• Schedule maintained
• Performance enhanced
Periodic Market Assessment: Although DMSMS screening has the
potential to assist in statistical problem prediction, the accuracy of
The Technology Insertion process should be tailored, as such forecasting for individual line items cannot be guaranteed.
necessary, to accommodate the particular driver(s) confronting a Direct manufacturer coordination is often the only accurate means to
program. While the process would be of most benefit to new evaluate DMSMS vulnerability for specific items. Selected item
programs, it can also be employed to advantage by programs already manufacturers, other industrial organizations and government
well into the acquisition cycle. At each stage of the process, agencies should be contacted to maximize early identification of
appropriate sources of knowledge (e.g., funding profile, tolerance DMSMS issues. JESD48, Product Discontinuance, establishes the
limits, candidate technology performance improvements) are requirements for device manufacturers concerning timely customer
considered. Of particular interest within the System Engineering notification of planned product discontinuance, which will assist
area, is a focus on development of design techniques that capitalize customers in managing end-of-life supply, or to transition on-going
on advances in modeling and simulation tools to facilitate rapid requirements to substitute products.
prototyping to achieve faster, cheaper part replacement or component
redesign to incorporate new technology prior to actual design freeze. A number of industry and government sponsored resources are
available to support life cycle analysis and DMSMS monitoring …
Planned System Upgrades
Computer Aided DMSMS Management Tools, such as those provided
System upgrades should be planned at defined intervals so by TACTech Inc. (i2 Technologies, Inc.) and Manufacturing
microelectronics obsolescence can be dealt with at the same time. Technology Inc. (MTI), provide means to automate research and
This involves predetermining points during the equipment life at notification, component selection, and changing availability. Parts
which the design of all or parts of the system will be brought up to lists are loaded into a database that provides automatic DMS event
date and obsolete items replaced. This approach is particularly notification and presents analysis concerning the risk of
effective when phases of research, development and use take place in microelectronics obsolescence associated with a specific design.
parallel. A limitation of this approach is that these modifications can Many sources of parts, indenture and other related information exist
be so large that they require weapon systems to be unavailable for which are invaluable for DMSMS case analysis. DoD CD-ROM
operational service for a significant period while being upgraded. products are available which provide supply, maintenance,
Between the planned upgrades, other options, such as bridge-buys, procurement, design, engineering, and other logistics data on items
may be necessary as interim means to mitigate microelectronics purchased / used by the government. The AFMC DMSMS Case
obsolescence for on-going equipment manufacturing while upgrade Resolution Procedures Guide and NAVSEA DMSMS Case
development proceeds. Planned system upgrades should be Resolution Procedures Guide include comprehensive listings of these
considered for new electronic systems, when the time scale for resources.
obsolescence can be predicted accurately, under circumstances of
rapid technological development, or when a lifetime buy is The Government Industry Data Exchange Program (GIDEP)
inappropriate (e.g. shelf life constraints). maintains DoD's centralized database for DMSMS and provides a
number of services to support DMSMS Management.13 GIDEP
Life Cycle Analysis and DMSMS Monitoring issues DMSMS Notices when notified by a part manufacturer or
GIDEP participant that a part or production line will be discontinued.
The foundation for effective life cycle obsolescence management GIDEP promulgates notice to representatives at subscriber activities
resides in careful integration of DMSMS program elements with in DoD, and to member organizations in private industry. These
system / equipment configuration control activities. Maintenance of notices normally contain data such as the last date of part
accurate configuration data to the piece part level is essential in manufacture, last date for order processing, and minimum order
support of DMSMS impact assessments and associated resolution quantity or buy value. GIDEP DMSMS notices are available to
analyses. At the same time, this information will also support registered GIDEP participants via the GIDEP WWW Database
visibility of potential out-year DMSMS problem areas and provide Access. Member organizations can submit parts lists through the
the basis for proactive resolution efforts. Accordingly, an effective automated parts matching service and they will be run against the
life cycle DMSMS management program will involve components GIDEP database to determine if there are any non-conforming or
from each of the following areas: obsolescent parts reported against it. A report is then returned to the
member showing any DMSMS information against their parts list.
Configuration Item Identification / Analysis: Development and GIDEP also provides for exchange of information relative to part
maintenance of current configuration item listings to the piece part manufacturing, testing, operation, and characteristic data among
level are essential to effective DMSMS program management. industry and government agencies, and may also be used as a
6 of 11
primary source of information for identifying substitute parts and EXECUTIVE AGENT FOR
redesign criteria.
DOD MICROELECTRONICS DMSMS
The DoD DMSMS Teaming Group is a formalized group of
representatives from DoD programs and industry that work together The Defense Microelectronics Activity (DMEA) has been
to share solutions to common component obsolescence problems.14 designated the Executive Agent for DoD Microelectronics DMSMS.
The Teaming Group maintains a database of current information on As such, DMEA develops and coordinates solutions to DoD's
component obsolescence and, whenever possible, explores obsolescence problems and is responsible for issues relating to IC
resolutions that will work for all programs experiencing the microelectronics DMSMS. These responsibilities include:
obsolescence problem, often reducing the cost. For example, if a
specific component used by more than one program is no longer • Develop partnerships with the military services, other DoD and
offered by either the original device manufacturer or from an non-DoD organizations, the Semiconductor Industry, and
aftermarket supplier, each affected program may determine that Electronics Manufacturing Industry to foster cooperation and
emulation is the best resolution. Each affected program could then achieve a joint resolution to current and/or potential DMSMS
share the nonrecurring engineering costs equally. issues that affect the DoD.
• Advocate and develop cost-effective technical solutions to the
Part Selection Guidelines DoD’s IC Microelectronics DMSMS problems.

When selecting components for use in new equipment designs, • Implement an "IC Clearinghouse" for IC problems and solutions
the life cycle of the components should be considered in order to in order to provide a vehicle to talk about common DMSMS
minimize the negative impact of obsolescence during product problems, and share in the solutions.
development and subsequent product support. New design activity
• Develop guidelines and strategies in conjunction with the
should focus on selecting devices that are in the early stages of their
military services and DLA that will help weapon system program
product life cycle. When selecting device manufacturers,
managers effectively manage and mitigate microelectronics
consideration should be given to those manufacturers who have
obsolescence and related issues.
established agreements with aftermarket sources for product life
cycle extension. • Provide design recommendations for mitigating microelectronics
obsolescence throughout the life cycle of a weapon system.
Examples of industry guidelines that include consideration for
component life cycles when selecting components are ANSI/AIAA- DMEA’s Flexible Foundry is being developed to support obsolete
R100-1996 Parts Management, IEC CA-AWG/1/DC Component 5-volt semiconductors that the commercial industry has abandoned in
Management Plans, and CENELEC ES 59010 Electronic Component the pursuit of newer lower voltage technologies. The program was
Policy and Management Programme. implemented after the commercial semiconductor industry made the
understandable and justifiable business decision to no longer produce
SD-18, Defense Standardization Program Guide for Part parts for the low-volume, long-product-cycle military market.
Requirements and Application, provides part acquisition guidelines DMEA’s Flexible Foundry solves this problem by licensing and
for Program Managers (PMs)/System Program Offices (SPOs) and fabricating proven industry microelectronics processes. This allows
original equipment manufacturers (OEMs). The database contains DoD to obtain its small volume requirements from DMEA’s Flexible
guidance that enables OEMs to use military and commercially Foundry while large volume orders are supplied by industry. The
specified parts for military equipment. It provides guidance on how flexible foundry provides a diverse mix of functions ranging from
the DOD and its contractors can cooperatively select devices, which personalization of device and gate arrays to full custom fabrication of
will result in the lowest cost of ownership for the DOD. ASICs.

Part Documentation RESPONSE TO DMSMS EVENTS

A process should be in place to collect, store, and retrieve Numerous resolution alternatives exist which may be used
component data needed to address DMSMS issues that will arise in singularly or in combination. Industry experience shows that
the future. For example, in cases where engineering drawings (e.g. responses to DMSMS events fall within the following major
SCDs) were used in conjunction with design disclosure packages, categories and percent probabilities:15
these drawings would be used in conjunction with evaluating
candidate replacement parts. Component data to be considered may
DMSMS Response % Probability
include the original device manufacturer's data sheet, application
data, qualification data, etc. For complex devices, functional and Replacement Part 67%
behavioral level models (e.g. VHDL) should be included to allow
designs to be cost-effectively transitioned to new technologies. Life-of-Type-Buy 20%

Bridge Buy / Redesign 12%

Emulation 1%

The following presents common DMSMS problem resolution

alternatives.16 The potential may exist to combine resolution options
to achieve cost, technical or schedule benefits. For example,
modified LOT buys (called “bridge buys”) may be made to provide
sufficient stopgap materiel while longer term design-related

7 of 11
alternatives are pursued. Therefore, throughout the case investigation carefully to ensure that the selected COTS part meets the needs of
process, the potential for integrating elements of different solution the application. For some severe application environments (e.g.
methodologies to support cost-effective resolutions should be airborne uninhabited, space applications), COTS items may not be a
considered. viable solution due to reliability considerations.

Alternate Source The Government Electronics & Information Technology

Association (GEIA) G-12 Solid State Device Committee developed
Use this option if part specifications and test, acceptance, and guidelines for assessing the suitability of plastic encapsulated
related technical data are complete and available. Aftermarket microcircuits and semiconductors for use in military, aerospace and
sources are firms that buy obsolete production lines and thereby other rugged applications. EIA Engineering Bulletin SSB-1,
maintain capabilities to reproduce selected DMSMS parts. Many Guidelines for Using Plastic Encapsulated Microcircuits and
original device manufacturers establish agreements with aftermarket Semiconductors in Military, Aerospace and Other Rugged
sources for product life cycle extension. If such manufacturers were Applications provides:17
selected during equipment design, the aftermarket source provides a
viable option when the original device manufacturer discontinues • Methods for selecting the most suitable device for the application
product. from both an equipment performance and economic perspective

When considering this alternative, manufacturer production • Means to emulate commercial buying practices by drawing upon
capabilities, tooling, test programs, etc., should be evaluated to qualification and reliability evaluation methods applied by the
ensure ability to meet original item specification requirements. For microelectronics design and manufacturing industry
some ground based systems and for space based systems in general, Emulation
particular attention must be given to ensuring the alternative source
for the device meets radiation requirements. Emulation is the process of developing replacements for obsolete
microcircuits using state of the art materiel, design and processing
In addition to or in lieu of purchasing manufacturing capability, techniques. For unavailable components; however, a risk does exist
an aftermarket supplier may procure wafer or chip product from the that emulated parts may fail to meet certain unspecified performance
original manufacturer. Some final manufacturing steps such as characteristics of the original item and thus, suitability for all
specialized packaging and testing are usually required to prepare the applications may not be guaranteed. As with aftermarket
device for application. manufacturers, price per unit for emulated items is likely to be
extremely sensitive to order quantities and this fact should be
It may be possible to make an extended buy from this supplier or considered when developing a procurement strategy for this
to negotiate a long-term parts supply agreement. If the wafer or chip alternative. At the same time, the emulation process involves creation
was produced on a QML, they may be acceptable without further of a design library supporting wafer fabrication; therefore, if the
testing; otherwise, a test and qualification program may be necessary. DMSMS item is a common or previously emulated design,
preliminary engineering costs may be greatly reduced. The emulation
Substitution process may be conducted at the IC, circuit card, or other designated
system indenture level, and is therefore often considered a subset of
This alternative involves analyzing DMSMS item characteristics redesign initiatives as discussed.
and attempting to locate a similar part with an acceptable degree of
nonconformance. A detailed cross reference and comparison of Life-of-Type (LOT) Buy
original versus substitute part characteristics must be conducted, and
an engineering deviation or waiver is generally required to support This alternative involves purchasing a supply of DMSMS items to
the change since it may require relaxing part specifications or support total demands for the projected service life of the impacted
performance parameters. It should be noted that cross-reference systems / equipment. It should be noted that LOT purchases are not
methodologies may differ for mechanical / materiel versus electronic necessarily limited to DMSMS items. For example, in the case of
items, in part due to the availability of MILSPEC/MILSTD/CID microcircuits, the only available option may be the purchase of LOT
references. For example, for electronic items the process may quantities of die, which would require additional fabrication steps
generally be expedited through immediate analysis of lower quality prior to use. Similarly, the LOT buy option may involve purchase of
parts, or by utilizing commercially available systems that cross- items or materiel essential to continued production or repair of the
reference all parts. DMSMS item. LOT buys have traditionally been a common
resolution alternative, but accurately predicting lifetime demand
Redefine Requirement to Accept Commercial Item requirements can be difficult. Other issues to consider when
evaluating this option include long term storage, periodic verification
Working through the appropriate engineering support activities, of lead solderability, etc.
redefine the requirement to accept a commercial item. This could
lead to the emergence of additional sources. The process is similar to Redesign / Design Modifications
the substitution alternative, except you are redefining the item to
accept a commercial item already available, instead of finding an This alternative involves designing out DMSMS items via
item that is similar to the DMSMS item. engineering changes at various system indenture levels, with goals of
enhancing system performance and improving reliability and
It is important to remember when selecting commercial-off-the- maintainability. As in previous alternatives, redesigns at the
shelf (COTS) items that the spectrum of those items in quality and component or assembly may involve significant risk and extensive
technical specifications is broad. The design limits, environmental system integration testing if the item in question has multiple
profiles, and life cycles vary. General categories of commercial items different applications. Moreover, depending on the scope and level of
include consumer, industrial, automotive, and specialty items. The the redesign effort, substantial nonrecurring engineering and life
characteristics of these items must be understood and evaluated
8 of 11
cycle logistics costs may accrue. Redesigns may be most appropriate or other DMSMS risk areas identified during case investigations. For
when a fairly large percentage of current or potential DMSMS parts example, conversations with manufacturers may indicate emerging
are resident within a particular component, equipment or end-item. DMSMS problems or broader supplier financial or technical
circumstances that may affect continued production operations.
• Redesign / Design Modifications: This alternative involves Alternately, a predominance of DMSMS cases involving similar part
designing out DMSMS items via engineering changes at various types or technologies may suggest general obsolescence trends. Any
system indenture levels, with goals of enhancing system such source / technology trends should be monitored in support of
performance and improving reliability and maintainability. As in life cycle DMSMS management efforts.
previous alternatives, redesigns at the component or assembly
may involve significant risk and extensive system integration NONRECURRING ENGINEERING COST FACTORS
testing if the item in question has multiple different applications.
Moreover, depending on the scope and level of the redesign To minimize the impact of DMSMS, DoD activities and original
effort, substantial nonrecurring engineering and life cycle equipment manufacturers (OEMs) must be able to incorporate the
logistics costs may accrue. Redesigns may be most appropriate most timely and cost-effective resolutions. The following
when a fairly large percentage of current or potential DMSMS nonrecurring engineering cost factors were developed to allow DoD
parts are resident within a particular component, equipment or programs to uniformly report DMSMS cost avoidance associated
end-item. with implementing the best resolution in line with program
• Redesign the Entire System: Replacing an obsolete or requirements and cost constraints.18 Original equipment
discontinued item often can extend the life of a next-higher- manufacturers (OEMs) can also use these cost factors to perform an
assembly (NHA) and / or result in enhanced performance. In economic evaluation of solutions to DMSMS events.
addition, it may be more economical to replace the item or the
NHA than to use another method to resolve the problem. Resolution Low Average High
Replacement with newer technology or replacement of a higher
Existing Stock $0 $0 $0
assembly are two common replacement options. Note:
Replacement of a higher assembly is not limited to the next Reclamation $629 $1,884 $3,249
higher assembly. For example, an entire radar unit may be
Alternate $2,750 $6,384 $16,500
replaced with a newer, more enhanced one rather than continuing
to replace board or part level discontinued items on the original Substitute $5,000 $18,111 $50,276
radar unit.
Aftermarket $15,390 $47,360 $114,882
• Redesign or Modify the Next Higher Assembly: As indicated in
Emulation $17,000 $68,012 $150,000
the previous alternative, this option is touched upon within the
context of other alternatives such as substitution. In cases where Redesign— Minor $22,400 $111,034 $250,000
replacing the DMSMS item itself is cost, time or design
Redesign— Major $200,000 $410,152 $770,000
prohibitive, consider the replacement of the next higher assembly
as an alternative. For example, replacement at the board level Life of Type Buy * * *
many be a better option than replacement of an individual chip.
* The LOT buy resolution is program-specific and should be
• Replacement with Newer Technology: With the continual calculated by individual programs
improvement of technology, many serviceable technologies
become obsolete rather than nonfunctional. They may rapidly go
out of production in favor of the newer, enhanced technology. These cost factors were determined for nonrecurring engineering
Replacing these items with the newer counterpart if it meets (NRE) in constant fiscal year 1999 dollars. NRE cost factors do not
form, fit, function and interface requirements may be an easy and include procurement and administrative labor hours (e.g. time to
cost effective solution. A review of the specifications should be identify sources of supply) and do not include costs associated with
done to ensure obstacles to use of the new technology are not developing new microcircuits using state-of-the-art technologies (e.g.
artificial (i.e. created by the limits of technology available at the ASIC replacement for several discrete devices).
time). Enhanced performance may be achievable through
exercising this alternative. When evaluating this option, design The following should be considered when using these cost
analysis may be necessary to ensure the newer technology does factors:
not introduce functional performance problems (i.e. using a
higher speed device may result in timing problems, a lower • New source qualification could add cost; however, no standard
voltage device may be susceptible to noise in supply voltage, value could be obtained because the cost could be amortized as
etc.). part of recurring cost.

DMSMS CASE RESOLUTION HISTORY FILE • If radiation hardening testing is required, the cost factors
presented in the table could increase from $5,000 (dose rate only)
The DMSMS solution should be documented for the purpose of to $52,000 (dose rate, total dose, and single-event upset) and
capturing lessons learned, use by other programs confronted with the possibly as much as $82,000 for microprocessors.
same DMSMS case, to support cost estimating efforts for DMSMS • If additional testing is required, each cost factor could increase
resolution activity and to support technology trend analysis. The from $600 (acoustic microscopy only) to $47,340 (full
history files should contain all data collected or developed during the qualification of a 100-piece lot).
case resolution process. The files should be maintained to support
follow-on analyses and to assist others in conducting related
DMSMS investigation efforts. As a corollary action, procedures
should be established for tracking prospective sources, technologies
9 of 11
 AN ECONOMIC METHOD FOR EVALUATING Henry Livingston has over twenty years
of engineering and engineering
ELECTRONIC COMPONENT OBSOLESCENCE management experience in the Aerospace
SOLUTIONS Electronics industry. He presently
manages Component Engineering at
This section of GEB1 describes a quantitative approach to BAE SYSTEMS Information and
developing a solution for component obsolescence. It is important to Electronic Warfare Systems. Henry is
have tools that will allow the engineering team to understand the Vice-Chairman of the Government
economic factors involved in the determination of the optimum year Electronics & Information Technology
for planning a redesign. Detailed models can be developed that Association (GEIA) G-12 Solid State
consider anticipated production life of all components within a Device. The G-12 Committee develops solutions to technical
subassembly. When using this methodology, the engineering team problems in the application, standardization, and reliability of solid
must assure that they have accounted for all potential obsolescence state devices. Henry received the Electronics Industries Association
events that can be projected for the production and support lifetime Engineering Department Distinguished Contribution Award for
of the product in question. The details of this methodology were exemplary leadership and outstanding contributions to the defense
derived from "An Economic Method for Evaluating Electronic electronics industry, to the Defense Department and the Electronics
Component Obsolescence Solutions" by G. Zell Porter, Boeing Industries Association in the field of solid state devices. Henry is the
Company (May 1998). third recipient of the GEIA E. J. Nucci Memorial Excellence Award
for Engineering Excellence to Henry Livingston for his many years
ACKNOWLEDGMENTS of dedicated service to G12 and the high reliability solid state devices
industry. Henry is a member of the IEEE Components, Packaging,
Members of Task Group G9906 of the GEIA G-12 Solid State and Manufacturing Technology Society, the IEEE Electron Devices
Devices Committee developed GEB1. While the members of the Task Society and IEEE Reliability Society.
Group and principal contributors are shown below, it is not possible
to include all of those who assisted in the evolution of this Bulletin. REFERENCES
To each of them, the members of the GEIA G-12 Solid State Devices
Committee extend their gratitude… The GEIA G-12 Solid State Devices Committee would like to thank
the following authors and organizations for significant contributions
Olin Bray Sandia Labs through their published works that provided the foundation for EIA
Doug Casanova Defense Microelectronics Activity Engineering Bulletin GEB1, Diminishing Manufacturing Sources
Joe Chapman Chapman Enterprises and Material Shortages (DMSMS) Management Practices…
Justine Corboy Sarnoff Corporation
Frances Crull Northrop Grumman, ESSD 1
Onkvisit, S.; Shaw, J. "Product Life Cycles and Product
Tom Dean Boeing Company Management," Quorum Books, New York, New York, 1989.
2
Don Gagnon ITT Industries Pecht, M.G.; Das D. "Electronic Part Life Cycle," IEEE
Scott Hatch BAE SYSTEMS Transactions on Components and Packaging Technologies,
Wes Hubbell Raytheon Volume 23 Issue 1, March 2000, pp.190-192.
3
Bick, E. (TACTech Inc.). "New & Improved Life Cycle Codes
Mike Jones Defense Supply Center Columbus Updates - Response To AIM Users Regarding The New &
Paul Kelley Litton Aero Products Improved LCCs." [email protected] (20 Jul. 1999).
4
Henry Livingston BAE SYSTEMS TACTech Inc. "Obsolescence Management In the Year 2000." 28
Lee Mathiesen Lansdale Semiconductor Oct. 1998. ftp://tactech.com/pub/overview.exe (23 Oct. 2000).
5
Keith Meyer NWSC DMS Technology Center Dasgupta, A.; Magrab, E.B.; Anand, D.K.; Eisinger, K.; McLeish,
J.G.; Torres, M.A.; Lall, P.; Dishongh, T.J. "Perspectives To
Charlie Minter Best Manufacturing Practices Center of Understand Risks In The Electronic Industry," IEEE Transactions
Excellence on Components, Packaging, and Manufacturing Technology, Part
John Nirschl Rockwell Collins A, Volume 20 Issue 4, Dec. 1997, pp.542-547.
6
G. Zell Porter Boeing Company Hitt, E.F.; Schmidt, J. "Technology Obsolescence (TO) Impact On
Bill Pumford GIDEP Operations Center Future Costs," Proceedings of the 17th Digital Avionics Systems
Conference (DASC), 1998, AIAA/IEEE/SAE, Volume 1, pp.A33-
Kevin Rankin Defense Microelectronics Activity
1 - A33-7.
David Robinson Defense Supply Center Columbus 7
Borky, J.M.; Lachenmaier, R.N.; Messing, J.P.; Frink, A.
Ron Shimazu Defense Microelectronics Activity "Architectures for Next Generation Military Avionics Systems,"
Vernon Singleton Lockheed Martin Systems Integration IEEE 1998 Aerospace Conference, 1998 Volume 1, pp.265-281.
8
Jack Tarsa General Dynamics Defense Systems Houston, J.A.; Concha, L.; Bohannan, R. "VHDL Design
Environment For Legacy Electronics (VDELE)," Proceedings of
Bahig Tawfellos Honeywell
the VHDL International Users' Forum, 1997, IEEE, pp.162-169;
Walter Tomczykowski ARINC Incorporated and…
Fitzhugh, G.L. "Rapid Retargeting - A Solution To Electronic
Systems Obsolescence," Proceedings of the IEEE 1998 National
Aerospace and Electronics Conference (NAECON 1998), pp.169-
176.
9
Barker, D. "Why Are People Always Talking About VHDL?" 16
Aug. 1999.

10 of 11
 http://www.ml.afrl.af.mil/ib/dpdsp/AFMC_VHDL_Article.PDF
(23 Oct. 2000).
10
Audsley, N.; Bate, I.; Grigg, A. "Portable Code: Reducing the Cost
of Obsolescence In Embedded Systems," Computing & Control
Engineering Journal, Volume: 10 Issue: 3, June 1999, pp.98-104.
11
Bray, O.H.; Garcia, M.L. "Technology Roadmapping: The
Integration of Strategic and Technology Planning for
Competitiveness," Innovation in Technology Management - The
Key to Global Leadership. PICMET '97: Portland International
Conference on Management and Technology, 1997, IEEE, pp.25-
28.
12
Minter, C. (Best Manufacturing Practices Center of Excellence).
"Diminishing Manufacturing Sources and Material Shortages
(DMSMS) in the Systems Engineering Process." Proceedings of
DMSMS Conference 1999.
13
Government Industry Data Exchange Program (GIDEP)."
http://www.gidep.corona.navy.mil/ (23 Oct. 2000).
14
Martinez, J.G.; McDermott, J.T. "DMSMS Teaming Group
Process." 1998.
http://www.gidep.org/dmsms/initiatives/teamgroup.htm (23 Oct.
2000).
15
Porter, G.Z. (Boeing Commercial Aircraft Company) "Strategies
for Obsolescence Management in the New Millennium."
Proceedings of DMSMS Conference 1999.
16
US Naval Sea Systems Command (NAVSEA). "DMSMS Case
Resolution Procedures Guide." 2 Aug. 1999.
http://www.logtools.navsea.navy.mil/SuplySup/DMSMS/dmstable
.htm (23 Oct. 2000); and…
US Air Force Materiel Command (AFMC). "DMSMS Case
Resolution Procedures Guide." 15 Jul. 1998.
http://www.ml.afrl.af.mil/ib/dpdsp/crgv10.pdf (23 Oct. 2000);
and…
US Army Materiel Command (AMC). "DMSMS Program Case
Resolution Guide." 13 May 1999.
http://smaplab.ri.uah.edu/dmsms98/papers/dmsms.pdf (23 Oct.
2000).
17
Livingston, H. "SSB-1: Guidelines for Using Plastic Encapsulated
Microcircuits and Semiconductors in Military, Aerospace and
Other Rugged Applications." Proceedings of DMSMS Conference
2000.
18
"Resolution Cost Factors for Diminishing Manufacturing Sources
and Material Shortages." Prepared for Defense Microelectronics
Activity (DMEA), McClellan Air Force Base, California, by
ARINC Incorporated. Revised May 1999.
http://www.dmea.osd.mil/dms_contrb.html (23 Oct. 2000).

11 of 11