September 1996
2. TECHNOLOGY DRIVER APPLICATIONS
3. TECHNOLOGY DEVELOPMENT TRENDS
E. Bihler STP P. Collander Nokia J. Exposito SGS Thomson H. Hentzell IMC R. Honegger ESEC A. M. Howard TTS J. Klerk Philips K. Kurzweil Bull J. Peeters Alcatel W. Radlik Siemens H. Reichl FhG-IZM M. Roughton Lucas M. Salagoity Solectron U. Schaefer Bosch C. von Scheele Ericsson H. Forster EC M. Hugen EC D. Broster EC A. de Baas EC C. Maloney EC F.J. Ibáñez EC
1. INTRODUCTION
The objectives of the Packaging and Interconnection (P&I) workshop were to identify (a) areas of priority for R&D activities and (b) infrastructural actions aimed at improving the industrial scenario of P&I in Europe. Presentations were made by representatives from a wide range of European industries for which P&I is an important area of technology in which to remain up to date and competitive.
The workshop was organised in three groups of presentations, the first covering applications, the second covering manufacturing technologies and the third equipment and materials. Fields of applications included cellular communications, telecommunications, automotive and consumer products; from this section of the workshop emerged the driver applications for new technologies.
Although there are clearly some areas of technology needing further research and development, many of the common concerns and requirements for the future expressed by the presenters at the workshop were related to issues associated with manufacturability and cost reduction.
A significant driving force in the field of manufacturing, particularly in the communications industry, is the requirement for achieving a fast time to market, consequently, there was a repeated demand for much greater application of concurrent engineering during the development of products to ensure new technology is included at an early stage in the development programme. Also related to minimising time to market is the need for fast customisation and for ease of upgradability (using software rather than hardware changes) to minimise the need for a large range of models.
In general, industry is especially concerned to maximise flexibility of designs and components to allow the same units to be used in as wide a range of applications as possible. Maximum flexibility of manufacturing lines using automated equipment with self adjustment and self control are needed to enable rapid changes in production. The forgoing objectives all need to be achieved whilst maintaining a high level of quality and reliability and taking full account of ecological sensitivity.
Strengths within Europe include: major IC manufacturers over a broad range of products, key customers for advanced products, European assembly equipment suppliers covering all process steps and a good partnership between IC makers/users/equipment suppliers. To ensure effort is focused on the topics pertinent to these strengths in the marketplace, there was a demand in the workshop for technology roadmaps to be compiled specifically for the European situation, rather than depending on reference to the roadmaps produced for the USA and Japan.
2. TECHNOLOGY DRIVER APPLICATIONS
2.1 Communications
Mobile cellular communications together with PCs, is one of the major drivers for Packaging and Interconnection technology with a world-wide turnover of between 35 and 40 Billion $US in 1996. There is an increasing focus on data transmission, but this is not expected to result in such high volumes as personal telephones for voice communication. However, commercial margins for mobile terminals are being reduced over the last few years which makes these products less profitable when compared to more costly high performance communication products. The personal phone is being developed into a personal communicator embodying facilities for fax, internet access and notebook in addition to voice all within the size constraints of the phone.
The big market is still seen as communications between people which will be dominated by voice transmission until image transmission becomes viable. To meet these objectives, faster transmission (at higher frequencies) and lower energy consumption are required. The major impact on advanced P&I is the demand for ever increasing complexity in a smaller, lighter, cheaper (30% reduction per year) and lower power consumption form with improved electromagnetic compatibility and reliability together with maximum environmental friendliness. Overall, there is the demand for low cost with high performance and reliability coupled with a short time to market.
Associated with the production of mobile communications equipment is the supply of strategic components and technology which include battery technology, small size displays, high density organic boards and integrated radio frequency components.
There is an overall continuing increase in the use of microelectronics technology in automotive applications. Typical target products include engine management, anti-theft, ABS/brake by wire, cruise control/anti-collision/auto-steering, dynamic control of brakes/transmission/steering/suspension, zero emission, climate control and information centres; navigational control; and communications. The associated technology demand is for increased flexibility and lower costs whilst improving reliability under higher operating temperature conditions.
2.3 Consumer
In the mainstream, consumer packaging follows that of telecommunications. The demand is for single chip or multichip modules based preferably on standard packages for minimum cost. The need for concurrent engineering is of particular importance with respect to the very large increase in power handling required in systems (peak power dissipation rising from approximately 15 Watts in 1995 to a predicted 45 Watts in 2000). Problems may be avoided by improved PCB thermal performance if this is designed into the system at an early stage. Whilst multi-chip solutions are seen as a way of achieving further miniaturisation, they are also being demanded in products such as the Intel Pentium-Pro to reduce parasitics and crosstalk for high speed data transfer.
In order to enable rapid changes in production and high levels of production flexibility, there is a need for a high degree of assembly line front-end and back-end integration such as is being developed currently by major European packaging equipment supplier.
3. TECHNOLOGY DEVELOPMENT TRENDS
3.1 Communications
Currently using surface mounting technology with 0.4mm pitch packaging, the requirement is for shrinks in the silicon area, lower ASIC I/O counts and lower power consumption to satisfy the market demands for further size and weight reductions. Faster transmission speeds are demanding higher frequencies and integrated radio frequency stages. Frequencies are generally up to 2 GHz with some links at 18 GHz and a suggested usage of the 50 GHz band for local links, a potentially large application area. Package trends are from QFP (Quad Flat Package) to BGA (Ball Grid Array) for infrastructure equipment and to CSP (Chip Scale Package) and flip-chip for phones. PCBs are moving towards flex, build-up with the need for integrated passives. Device mounting needs complementary technology to SCP such as Few Chip Packages (which reduces the KGD problems) with up to 5 chips per package. 3-D chip cubes are considered a useful solution for infrastructure memory requirements. For some applications, such as the global positioning system, advantage is being taken of military communications technology experience for consumer products.
For switching systems applications, the driving force has moved from technology to services. R&D priorities are for high density PCBs and MCM-Ls incorporating metal cores for thermal dissipation, increased component integration (embedded passives), improved high frequency performance, micro-via technology (laser drilling, etc.) and flip-chip on board, reduced size through fine pitch/low profile packages.
3.2 Automotive
Requirements are for low cost flexible leadframe few chip modules for higher power and high reliability, reliable mixed wire bonding processes for high temperature applications and the development of mold resins for operating temperatures > 175C. The single chip packages format required include: logic ICs of different integration technologies in standard package formats, power stages and interface ICs in specific power packages and mechatronics sensors and interface ICs in selected plastic packages. Few chip packages (2-3 chips) in standard package format are considered of interest for automotive electronics. There was a specific desire expressed for a European supplier of low pincount (20 to 64) power packages. Other requirements include reliability in harsh environments (>200C, 3KHz, 50g), low cost bare die with first time assembly yields > 98%, power electronics thermal management, high reliability substrate technologies, smart components/embedded electronics in housings, and pressure/gas/magnetic sensors. To meet the demands of the harsh operating environment associated with the strong automotive industry, there is a need for a source of high volume, low cost LTCC substrates in Europe.
3.3 Consumer
There is a significant technology trend towards miniaturising packaging and using sub-half micron chip technology in the consumer industry. Pressure exists to develop gold ball bonding capabilities to reduce further the minimum die-pad spacing capability from 80m to 50m by 1999. For minimum cost, multi-chip packaging should be based on existing package outlines. In order to handle the significant increase in power dissipation, concurrent engineering is needed to ensure maximum use is made of improved thermal dissipation PCB technology and thus avoid the use of special packaging techniques to obtain the necessary reduction in thermal resistance. Access to KGD and CSP is critical for the consumer industry to be able to achieve levels of miniaturisation which in-development products are demanding.
3.4 Single Chip Packaging
The traditional driving forces for SCPs are increased silicon complexity (higher numbers of I/Os, larger die, higher speed, signal integrity and higher power dissipation) and overall system miniaturisation (smaller size, lower thickness and ability to accept multi-chips). Against the predicted trend for dramatically increased I/O counts, is the move towards higher levels of integration, both in single chip designs and through the use of multi-chip modules: both of these result in fewer than predicted package I/O counts.
The world-wide package technology trend has been from peripheral contact through array contact to chip size packages (< 100 I/O as yet) and flip-chip (embryonic). R&D is shifting from the development of very fine pitch peripheral packages to area array packages responding to the demand for higher miniaturisation (in both x-y, and z dimensions), higher reliability and lower cost. The new challenge is in manufacturing to maximise the exploitation of the technology through flexible, fully controlled production lines which are susceptible of being competitively exploited in Europe.
3.5 Multi-Chip Modules
With a significant commitment towards the use of MCM technologies in portable phones made recently by Siemens, this technology is finding its way in high volume production markets. Concurrent engineering is considered particularly important if multichip modules are to be taken full advantage of design, test, system architecture, devices and packaging. Surface mount technology is now running out of headroom and multichip modules offers higher density boards/substrates which include ICs as bare die or micropackages. The introduction of MCM technology requires users to adopt new working procedures, resulting in a technology step that manufacturers need to master. Factors that should be taken into consideration when introducing multichip module technology include existing technical commitments, incremental improvements of more traditional approaches and an ever decreasing cost target.
Developments in MCM-L substrate technology have resulted in finer pitches and fewer layers bringing it close to MCM-D capabilities. The cost of substrates is now considered to be dependent not on the technology but on the maximum size of panel which can be handled in production. The sizes of MCM-D panels are now approaching that of MCM-L. MCM-C has a problem going to large panels but it will be the technology of choice for harsh environments applications (automotive). In substrate technology development therefore, large panel lithography is considered a critical issue.
Two package standards are emerging; (i) a plastic QFP with an MCM-D inside and used like an single chip package (suitable for portable phones) and (ii) for low cost (automotive applications), a BGA containing an overmolded MCM-L. A common feature of the two styles is that users are presented with a familiar package and does not have to be aware that the device is a multi-chip assembly.
The cellular phone is an ideal target market for multichip module technology because the of the strong constraints in miniaturisation and the high volume requirement which would force the reduction in cost of the technology. The multichip module technology is progressing towards maturity and is already coming down in price; communications and automotive being the segments in which the technology offers more rewarding opportunities in Europe. It is predicted that in the year 2000, MCM-L will be the dominant technology with MCM-D focused on high frequency systems and MCM-C on robust automotive systems.
It is clear from the first year of Europractice-MCM that the promotion of interest in multichip module technology is highly dependent on educating potential users in the technical and commercial advantages offered. Even though the technology has been shown to be capable of solving some of the current problems of size, speed, electromagnetic compatibility, etc., future activity needs to be targeted at producing industrial demonstrators aiming at providing practical evidence of the capabilities and cost/performance viability of multichip module technology.
3.6 PCB technology
For mass markets, cost is the challenge, whereas for niche markets it is performance. For mass markets, in comparison with the Far East materials and overheads are similar, but labour and depreciation (because of high level of automation) costs are higher in Europe. The cost reduction strategy is to increase panel size and the level of product integration whilst maintaining a high yield. The technology advances needed to achieve the above are: finer pitch track metal (change from etched copper to additive printing), denser and smaller vias (change from mechanical drilling to laser or photolithography), thinner dielectric (change from glass/epoxy to non-woven), denser chip connections (change from peripheral to area array) and denser external connections (change from pin connectors to optical or voice input).
The trend from a business perspective is to form strategic alliances of designers and manufacturers of chips, PCBs, assemblies etc. within Europe to regain the concurrent engineering advantages of the vertically integrated companies of the '80s.
PCB mounting technology is concerned today with QFP (0.4mm pitch is probably the lowest for high volumes), BGA and CSP formats but moving towards flip-chip and 3D over the next 5 years. It is essential that a low cost, reliable flip-chip technology is available in Europe to support this evolution. Help is needed to identify the most appropriate (reliability/cost) flip-chip technology for the European needs.
3.7 Materials, techniques and equipment
R&D should be directed towards consolidating the ball grid array, chip scale packages and flip-chip processes to ensure reliable manufacturability at low cost. Topics to be covered include; materials characterisation (modelling and measuring), low cost additive technologies, mechanical solder transfer techniques, low cost substrate materials, low cost fine-line multilayer techniques, process compatibility to PCB technologies and environmentally compatible materials and processes. MCM technologies are application oriented (thin-film on PCB, thin-film/thick-film combinations, etc.) and thus require a well characterised portfolio of materials and technologies from which the most appropriate may be selected with confidence for any specific application.
For the development of new highly automated assembly production lines, an essential requirement is a close relationship between users and equipment manufacturers to ensure alignment of R&D roadmaps. Priority areas are equipments for ball grid arrays, direct mounting (flip chip, chip on board), known-good die and chip size packages which are reliable and highly automated with 'self adjusting/self controlled' processes. Current development topics are fine pitch (80m) wire-bonding, low temperature wire-bonding, soft solder dispense and bonding. Overall, there is a strong requirements for manufacturing line flexibility to accommodate changes in product.
3.8 KGD & CSP
Activity and interest in Known Good Die is growing significantly with programmes at major semiconductor companies accross the world (Intel, Texas Instruments, Motorola, Samsung, GPS, Mietec, TEMIC, etc.) . In Europe, manufacturers are placing an ever increasing emphasis on the need for a supply of KGD. In general there is keen interest in digital, but little interest in analogue or discrete devices. In the USA, burn-in is used, and in Europe, electrical stress. CSP is generally in-house in Japan as yet with no large scale production.
Currently, bare die sales account for 2% of total production. This is forecasted (National Semiconductors) to grow in the period 2005 to 17-35%. The areas of use are automotive, telecommunications and portable communications. It remains unclear whether or not a substantial part of this market would be served by CSP packages, particularly those standing harsh conditions.
Requirements are: capability in harsh environments, standard delivery formats, capability for 100K to 1000K devices per type per year, and cost equal or lower than equivalent packaged part.
4. RECOMMENDATIONS
Packaging and interconnect technologies may represent up to 50% of a total system value. The presence in Europe of major players in every area of the product manufacturing chain, from semiconductor suppliers to assembly houses and equipment manufacturers, should allow the build-up of strong vertical partnerships capable of ensuring the development and introduction of innovative and cost-effective P&I techniques. Within this context the following recommendations were made:
Esprit Home Page | TCS Home Page
The URL for this page is /esprit/src/tcspkgrp.htm
It was last updated on 25 October 1996, and is maintained by John Magan - john.magan@dg3.cec.be