Experts At The Table: 450mm Fab And Facilities Challenges

Second of three parts: Unsustainable methods; supply chain evolution; separate consortiums; infrastructure issues; standards.

popularity

Semiconductor Manufacturing & Design sat down to discuss future 450mm fab and facilities challenges with Gerald Goff, director of the project management office for fab design and construction at GlobalFoundries; Joe Cestari, president of Total Facility Solutions; Ivo Raaijmakers, chief technology officer of ASM International; and Michael Brain, senior director of the Fab Solutions Business Unit at Brooks Automation. What follows are excerpts of that conversation.

By Mark LaPedus
SMD: What are some of the problems and challenges facing the fab facilities segment?
Goff: The single wafer clean technology that has been introduced is driving the facilities requirements at an unsustainable rate. We have seen an explosion in the use of UPW (ultrapure water). We have seen an explosion in the use of sulfuric acid and hydrogen peroxide. The consumption values there, with single wafer cleans, have risen to the point where they are no longer sustainable. There are even questions about the supply chain for sulfuric acid to meet that demand for full production. So something is going to have to change in that regard. We are going to look at new chemistries and start to move away from sulfuric. Or we must look at minimum ways of recycling it more efficiently than we do today as a result of the transition from batch cleans to single wafer cleans.
Cestari: If you roll back in time, all you had to worry about was getting hazardous waste into a container and then sending it off for disposal. Or waste was remediated back into the system. Today, because of the new hazardous chemistry and the sheer volume of waste produced compared to what it used to be, we can’t continue to package waste and send it off to an offsite facility. We need to be able to neutralize that waste.

SMD: What about other issues?
Cestari: I am believer in a real supply chain and not just driving down the materials costs. Some level of collaboration on everything—from the component level to some pieces of the equipment to some level of standardization—is what we all need. In addition, if you look at our supply chain, there is still a lot of redundancy. There is some overlap. And there is still a lack of standardization. Everyone thinks their creative little nuance is a competitive advantage. Some of them might be. But at the end of the day, we have to enable the people who make devices. Otherwise, collectively, we are not going to be successful.
Brain: From the facilities and AMHS point of view, I think there is going to be a lot of pressure to be space efficient. The tools are getting larger, and so, the cost of the facilities is going up. And we are going to have to find ways to not just grow the facilities by the same factor as the area of the wafer. That will translate to equipment and AMHS design. But at the same time, in a transition, we never want to change more than 30% or 40% of what we are doing. We are trying to keep the risk at a minimum. So for the most part, AMHS will look the same, just bigger. Meanwhile, tool buffering is one area that been investigated at 300mm and will become more prevalent at 450mm to take some of the load off from the centralized stockers and transport systems. In software, it looks like the standards will be moved across, which will help equipment manufacturers. This is so we don’t have to redo the standards for GEM450 the way we did it for GEM300. And, there is going to be a lot of pressure for different types of fabs to do things their own way. And to some extent, we may be coming back around to where large companies can afford to do a few things their own way to attain a competitive advantage, including designing specialized equipment or unique configurations of AMHS.
Raaijmakers: At 300mm, there was a lot of fab automation that was added, compared to 200mm. We won’t see such a significant shift in 450mm, which makes it a bit easier from a software and hardware point of view. The 450mm transition is a more straightforward scale up from 300mm.

SMD: The Global 450 Consortium (G450C) has recently set up a separate fab facilities consortium. The group is called the F450C. What’s the purpose of that?
Cestari: You can argue, and it’s true by fact, that facilities cost is a small portion, as compared to the investment in equipment. Yet, many times, the facilities industry is the critical path in converting assets to cash to produce product. But the industry can do a better job. The F450C is generally focused on getting all the right stakeholders together so that when we design and build tomorrow’s facility for 450mm, it will be optimized for cost, yield and the environment. For example, we’ve invited people like Swagelok at the component level to F450C. In the fab, the feeling is that they are a valve supplier. So we will cut their price by 20% and we will be OK. On the other hand, we need to get knowledge and ideas from every level of the supply chain. We have to do a better job of getting involvement at all levels and really creating a supply chain, not a linear relationship, where we are transferring cost reduction from the top to the bottom.

SMD: What about infrastructure issues?
Goff: At GlobalFoundries, we are kind of here on our own (in New York state). We don’t have a common infrastructure to share with other IC manufacturers or foundries. The problem with that is when we look at the competitors in Asia, particularly Taiwan, where you see these technology parks with basically uninterrupted power being provided by the government. And then they have Air Products or Air Liquide coming in and putting in these gigantic gas separation farms in the science park. Then, they have plumbing to transport the gases over to the fabs like a utility. That is a competitive advantage for them. So clustering of like operations, where we can have multiple fabs, or even competitive fabs, would be some way of helping to reduce the overall infrastructure costs, versus how we do it now. We have Intel based out in Oregon. We are based in the east. It seems to be something we need to think about as an industry in how we can become a little more competitive in the world market by clustering together.

SMD: So are you saying the U.S. needs a fab policy?
Goff: That’s exactly what I’m suggesting. I’ve spent a decade of my career offshoring. Now, I’m proud of the fact that in the last half decade of my career, we are trying to onshore everything that we’re involved in from a manufacturing perspective. The competitive nature of that is a challenge, especially where we are located. I would love to see some infrastructure development programs at the federal level. As this has gone to the state level, (the fabs) get put in these little isolated areas, where developing the infrastructure is a significant challenge. If there was federal funding available that would level the playing field—and put a (central science) park together, and incentivize the companies that want to do manufacturing of this magnitude in the United States. That would be a significant benefit to all those that would be willing to do that.

SMD: What about standards?
Brain: I am a big fan of standards. They have enabled us to get as far as we have as an industry because we can collaborate and work toward a common goal. As a testament to how well the standards were done at 300mm, many of them will move to 450mm. That will allow a lot of the equipment and software to be leveraged forward, which will pull a lot of the risk out of the transition.