Caution urged for people-less fabs
Chief Editor
300 mm technology is changing the industry. While it`s not necessarily causing an Industrial Revolution in semiconductor wafer manufacturing, it`s a historical event. But it brings with it a set of challenges to be borne by everyone in the industry.
Some experts foresee a people-less fab, or maybe the better-known “lights out” fab, as one possible goal of 300 mm semiconductor manufacturing facilities. As more minienvironments, SMIF technology and automation are built into fabs, the one component that cannot be dismissed is probably the most important of all: the human being.
I believe the people-less fab will occur, but it is years away. There are too many “what ifs” that need to be addressed. What if there is a problem with a minienvironment? What if the automation software has a virus? What if there is a problem with the automation hardware? I`m sure these questions are being addressed today. Afterall, SMIF and minienvironment technologies are not new. Technology that encapsulates the process tool to maintain a Class 1 environment is vital. Let`s face it, when 300 mm wafer fabs come on-line, minienvironments will play an important role in keeping costs down. They hold the promise of huge cost savings if they can be proven to deter airborne molecular contamination and allow wafer manufacturers to decrease their room classes yet keep yields high.
As evidenced in this month`s Special Report, “Factory automation revolu- tionizes cleanrooms,” (page 19), wafer fabs that use minienvironments and SMIF technology and have a Class 100 or higher — that is, Class 1,000 or Class 10,000 cleanroom — may be able to save big money.
In the process of our interviews with various industry experts, we discovered that automation seems to be under-utilized at semiconductor fabs. With fab personnel clad in bunny suits walking the fab floor with boxes of 200 mm wafers — one manufacturer described it as “a bee buzzing from flower to flower and pollinating it.” There must be a better way to move wafers.
For 200 mm fabs an inherent architectural problem exists — the rare use of unicassettes. Although there are many ways to put a carrier on a process tool, experts say the only feasible technology for 300 mm wafers is the unicassette. With 200 mm fabs rarely using this technology, those factories will have more difficulty incorporating automation.
When a manufacturer spends billions of dollars on a semiconductor fab, costs savings are critical. Automation and minienvironments seem a common sense way to save money. Upgrading 200 mm fabs may prove expensive, but the 300 mm fab will surely incorporate much more automation, more use of minienvironments and spearhead the newest cleanroom technology. Can the people-less fab be far away? No, but proceed with caution.