Divide and Conquer?

10/01/2002

The packaging world has seen profound changes over the decades and we may be faced with another one shortly. One of the most intriguing scenarios that I have heard in a long time states that packaging as we know it will be carved up and claimed by wafer fabs and contract manufacturers.

This idea was put forth by Jim Walker of Gartner Dataquest at SEMICON West, and also by Tom Di Stefano of Decision Track at MEPTEC's recent wafer-level packaging (WLP) conference. They both stopped short of predicting that this would happen, but neither of these gentlemen are slouches when it comes to gauging the prevailing winds in the packaging industry.

The basic idea is that the two biggest growth areas in packaging are WLP and system-in-package (SiP) technology. It remains unclear where WLP will be centered — it is conceivable that wafer fabs will take on that challenge. On the other end of the spectrum, contract manufacturers are constantly looking for areas of growth, so they may try to expand in the direction of SiP since it is conceptually similar to surface mount technology.

If both of these things happen, traditional packaging companies will be left holding just the diminishing middle ground of packaging. So, is it all over for us? My vote is no, for a number of reasons.

At MEPTEC's WLP conference, Richard Groover of Amkor pointed out that the tidy little box on flowcharts labeled "wafer-level packaging" actually is filled with some dirty processes that do not fit into a wafer fab very well. These include backgrinding, laser marking and dicing. The larger set of processes that includes these will stay in packaging houses, according to Groover. I have to think that companies like Amkor and ASE aren't going to sit by idly and watch their industry get split up by outsiders, so you can expect them and the other subcontractors to stay at the forefront of WLP and SiP.

The other main thought that leaves the future of packaging intact in my mind hinges on the vagueness of "conceptually similar." Although putting chips on a substrate inside a package seems about the same as putting packages or chips on a board in a surface mount line, these process flows use significantly different equipment and technology. It would be a huge and risky jump for a company to try to merge the two, or even take on both. It would be a great news story if someone does try, but it is hard to imagine much of the SiP capability being extracted from the current packaging companies.

This definitely is a trend that we will be following for you, but I don't think we'll need to change the name of our magazine anytime soon!

Thanks for reading,

Jeffrey C. Demmin
Editor-in-Chief
[email protected]