The Next 25 Years - Part II
23 August 2007
In the second in this short series, Peter Grundy from The SMART Group's technical committee gives us his personal view of what the future has in store for the electronics manufacturing industry
Oh, for a crystal ball that is not cloudy! Our electronics manufacturing world is now mature and is growing due to stimuli such as the increase in electronic modules in cars but this maturity suggests that there is less Revolution and more Evolution. It is not so much a case of “can electronics be used to make this product viable?” but is more a case of “can extreme electronics make this electronic product more functional, more densely packed or physically smaller?”
This means that all the weaker equipment suppliers have been either shaken out of the industry or they have died and we have a range of suppliers that are stable, strong and can react to evolutionary needs. Today, if you visit any of the major exhibitions around the world, there is very little revolution. 20 years ago, almost every stand had some revolutionary product or idea on display.
So, our basic future premise is one of stability with evolution. Possibly, one or two of the major suppliers may change direction or buy each other out but, in general, the suppliers we have today will probably be around for the next few decades and this applies to all aspects of electronics manufacture from raw materials and components to capital equipment and test. There may be structural changes to come, and a previous example would be the move by component manufacturers away from direct sales to sales via distribution, but most of the names we know now should still be there in 25 years.
We are demanding ever faster and more powerful data handling which has put huge emphasis on integrated circuit design and yield to the point where 450mm wafer diameters, or bigger, are necessary to accommodate the memories and processors of the future. As has been mentioned many times before, a silicon wafer plant operating with 450mm diameters costs billions of dollars and no single company can justify the returns within a sensible time frame whist allowing for the vagaries of the market to alter volumes of production and sale. Therefore, the trend for several companies to combine their resources and build jointly operated plants will continue and may spill over into other arenas such as those where capital equipment manufacturing companies operate.
It is likely that there will be increased use of modules employing stacked die configurations and the direct attachment of silicon die to silicon die via alloy bumps or stud bumps. Perhaps much of the control and signal circuitry that today uses discrete components will be incorporated into IC’s and this will lead to the increased use of Application Specific IC’s (ASIC’s). ASIC’s have a subtle benefit in that it is very difficult to assess how the designer has created the circuit and the ASIC therefore offers a degree of Intellectual Property protection and this may become more important over the coming years.
Possibly, the increased use of modules will eventually mean less placement of passive components which could have an effect on the passive component manufacturers and also the assembly equipment manufacturers. If this trend develops into the large-scale use of modules built on silicon, there may be a need for more specialised equipment to assemble these modules and the equipment manufacturers are likely to evolve into it. Most of them are capable of doing this now, they need the market impetus. Conversely, if the possibility of using large scale silicon stacks does not emerge, then there will be a need to continue to develop modules with huge numbers of passive components in smaller spaces. Therefore, we might see the need for smaller case sizes than 01005. Some passive component manufacturers are already experimenting with the 0050025 case size but the assembly equipment we know today cannot handle such tiny devices and a whole new range of specialised equipment might be required but only if the economics permit.
Increased use of silicon does not necessarily sound the death knell for passive manufacturers. They are likely to develop specialised networks or other devices that utilise their skills.
In the past, we all thought that Flip-Chip would take over the world by 2000 but it has not made the anticipated impact. Yes, Flip-Chip is being used in lots of places and industries, but it has not dominated in the manner expected. Therefore, what is written here is a best-guess fit to current ideas and trends.
Is there a Revolution coming?
So far, we have considered what we will call “standard electronics”, the use of components assembled on to a substrate of one kind or another. There is a lot of work being put into the development of printable electronics whether the product is a battery, a fuel cell, clothing or whatever. These developments show great promise for the future and will allow many manufacturers of equipment and products the chance to migrate into new arenas.
Much of the work is still in the University and research lab world but soon there will be commercially available products. If the cost-performance benefits are good, then printable electronics will take a share of the market.
Using today’s crystal ball, in summary, we can estimate that the electronics manufacturing world will continue to evolve into more functional and probably smaller packages but may have a revolution if printable electronics become a major player in the market. Military and Defence related products are not the only technology drivers today, as they once were and portable data products such PDA’s or mobile phones will continue to stimulate evolution. Our changing energy requirements will force the development of newer forms of electronics such as printable electronics in fuel cells.
There will always be a need for relatively unsophisticated products such as mower or washing machine controls that are built in huge numbers at low cost but any Revolution will come via entirely new needs.
Contact Details and Archive...
Most Viewed Articles...