10th Anniversary Insights
The evolution of the subcontractor

By Mark DiOrio

Click here to enlarge image

The role and status of the packaging, assembly and test subcontractor has steadily increased over time. Since first being introduced in the late 1960s/early 1970s as “job shops” that provided some extra manufacturing capacity to the semiconductor industry, the subcontractor today provides various design, development, assembly, test and supply chain management services. No longer just a provider of extra capacity, many subcontractors today represent the total capacity of some well-known and very successful “fabless” semiconductor companies.

The semiconductor industry was a $145 billion industry in 2001; of that figure, assembly and test accounted for about $35 billion. Roughly one-third of assembly and test was outsourced, thus bringing the subcontract assembly and test market to approximately $12 billion. And, as outsourcing increases, subcontract assembly and test providers should see even brighter days ahead. The use of subcontract assembly and test services is fast becoming an industry standard. But how did the subcontract assembly and test industry migrate from the lowly job shop shadows to such a prominent role in our industry? I give credit to a few companies and a lot of good people.

The Early Days
In the early 1970s, the early providers of subcontract services were a couple of Stanford Microsystems and Dynetics factories located in the Philippines. Anam/Amkor started at about the same time. They were in low-cost labor regions, and they provided mostly assembly capabilities for those in search of extra capacity and cost reduction.

At this same time, big multinational corporations (MNCs), such as Motorola, National Semiconductor, Fairchild and Texas Instruments, really pushed new packages and packaging technologies. They retained control by keeping their developments in-house and using their own factories. Such MNC factories were also placed in low-cost labor regions, but process controls and overall quality were better emphasized. Many of these MNC factories provided improved quality because they ran limited process flows with reduced variables. Thus, the MNC assembly and test engineers in these factories developed essential skills in the high-volume, low-cost assembly and test of high-quality, very reliable packages. The only consideration to use a subcontractor at this point was if capacity could not be met.

A Real Industry Emerges
The 1980s witnessed a growth in subcontract assembly and test companies. This was largely fueled by the semiconductor industry's overall growth and the need to support it with manufacturing services. During this time, fabless semiconductor houses emerged. Companies such as Linear Technology, Altera, PMC/Sierra and many more were starting up with limited capital and the desire to outsource their non-value-added manufacturing needs. Many MNC factory engineers and their production and operation counterparts – all very skilled – started to migrate to new opportunities available at the “subcon” factory. Thus, a new generation of subcontract assembly and test began to emerge, one with more training in manufacturing engineering and with more of a focus on quality.

Then companies like ASE, Lingsen, SPIL and Hana began to emerge. Provided the packaging changes were incremental in nature (an increase in lead count or package size), subcons could do many of the same things as an MNC but at a lower cost. New subcon start-ups, such as Swire (now an AIT factory), differentiated themselves as state-of-the-art subcons with a strong focus on quality and turn-around time. MNC factory workers who migrated to the subcons facilitated informal technology transfer.

Subcon as a Partner
The 1990s really gave credence to the subcon outsourcing model used by semiconductor companies. Subcons no longer just manufactured designs, they became increasingly more involved in package design and development. Fabless companies depended on packaging subcontractors to be the primary factories to manufacture components. Design, engineering and marketing offices of the subcon became established in markets near their customers and attracted highly skilled technologists and managers.

MNCs with factories later realized that subcons could assemble and test as well, if not better, than they could, and in many cases at a lower cost. No longer just a provider of extra capacity, a subcon's ability to perform plus the extraordinary product margin performance of the fabless companies forced many MNCs to re-think their manufacturing overhead structures. Thus began the sale of MNC factories to subcon companies.

Key Drivers
There have been many people who have made the subcon industry what it is today, and a few of these people have had far-reaching and positive impact, including:

Jason Chang – The chairman of ASE made the company into what is today perhaps the largest of all semiconductor pack aging and test companies in the world.

The Kim family – The founders of Anam/Amkor made it the most prominent and successful name in subcontract assembly and test.

John Boruch – He was instrumental in the growth and prominence of Amkor, and a significant key to the overall semiconductor industry's acceptance of the subcontract assembly and test house.

Ralph Duceour – He started up ASE and then left to found PT Astra in the early-1990s. The company is now called AIT and is a top-10 subcon on the rise.

Victor Batinovich – He founded Swire and ASAT, two very successful subcontract assembly and test companies; both companies enhanced the quality perception of the subcon by the semiconductor industry at-large.

What's Next?
With more than 80 subcons around the world, the next decade will present new challenges. The package has become a limitation in device performance. Advanced packaging of large silicon, high I/O, high-speed devices requires a paradigm shift. Packaging engineers will become more involved in design and electrical performance aspects of the die and package together.

How will the subcon meet these challenges? Does the subcon have the ability to come up to this new technology plane? By all experiences, the answer is “yes.” And companies that get there the fastest will be successful, and they may need more capacity. Thus, we will also see consolidation among them. Perhaps the subcon of tomorrow will even be … a wafer fab! AP

Click here to enlarge image

Mark DiOrio, chief executive officer, can be contacted at MTBSolutions Inc., 1630 Oakland Road, Suite A102, San Jose, CA 95131-2450; 408-441-2173; Fax: 408-441-9700; E-mail: [email protected].


Easily post a comment below using your Linkedin, Twitter, Google or Facebook account. Comments won't automatically be posted to your social media accounts unless you select to share.