Forecasting & Technology Drive Spares Supply Chain
By David Lammers
The spares business at Applied Global Services is evolving, as fab managers seek greater assurance of supply and advanced technical sophistication from their parts suppliers.
|David R. Lee,|
AGS Spares Marketing
“The twenty-first century spares operation is a complex, global logistics system,” said David Lee, managing director of spares marketing at Applied Global Services (AGS). “Fabs are expensive investments, and disrupting fab production due to a missing part can be a much greater concern than negotiating discounts on individual parts.” Lee noted that what’s needed today, with burgeoning growth in global device manufacturing, is a fully optimized supply chain that can support the growing demand and added complexity of today’s and tomorrow’s semiconductor manufacturing.
“Previously, the focus in the parts business was more on the tactical issues, inventory levels and pricing, when what truly benefits our customers is a supply-chain partnership connecting the Applied Materials supply chain The spares business at Applied Global Services is evolving, as fab managers seek greater assurance of supply and advanced technical sophistication from their parts suppliers. to the fab’s demands as efficiently as possible to maintain an assurance of supply at the right cost.”
Semiconductor manufacturers, Lee said, are realizing that while part prices are important, if the parts aren’t available when they are needed, or the quality is inconsistent, the cost of fab disruptions can greatly outweigh the perceived benefits of low-cost alternative parts.
One recent offering from Applied Materials is a Forecast Parts Management (FPM) model, in which customers sit down with managers from AGS to develop forecasts of the parts and quantities that will be needed over a given period (see figure 1). “We guarantee customers assurance of supply,” Lee said, while AGS “gets much-improved visibility into the part requirements that will make our customers successful.”
Figure 1. Forecast Parts Management (FPM) is designed to help minimize customer fab disruptions and ensure the highest quality parts, at competitive costs.
In turn, Applied Materials can use the combined forecasts from our FPM customers to improve the efficiency of the entire supply chain. This means the benefits of FPM allow greater assurance of supply, with more efficient use of inventory, which translates into savings for customers.
FPM is becoming a partnership platform, where customers and Applied Materials establish an ongoing working model, built around a forecast, that puts fab performance as the top priority. Through FPM, Applied Materials helps customers minimize fab disruptions and ensure they receive the highest-quality parts, at costs that enable them to remain competitive.
Lee noted that other programs are also evolving to help customers continuously improve efficiency and performance, and achieve lower cost of ownership for their Applied Materials equipment. “For example, we’ve been spending a lot of time strengthening our Total Kit Management (TKM) solutions, which offer turnkey spares, cleaning and logistics capabilities,” he said. TKM programs provide customers with a steady flow of clean kits to eliminate kit shortages and spending spikes, and often include industry-leading advanced coatings that extend part lifetimes, reduce particles and lengthen mean time between cleans.
“We’ve seen a threefold increase in our kitting business over the past three years as customers recognize the benefits offered by our TKM product. We see broad interest in the industry for solutions like TKM, from leading-edge manufacturers to more mature fabs,” Lee said.
AGS Central Engineering
Technology demands are moving upstream as well. Tom Cho, vice president of the AGS Central Engineering organization, said when he started his career at AGS, “it was pretty easy to replicate the parts. If a supplier had issues with a dimension, or needed to qualify a new computer numerical control (CNC) machine, we would send out an engineer and go through the qualification process. The wafer process window would not be impacted if the dimensional spec of the part had a wider range in limits. Today, however, the part dimensional spec must be much tighter because even a minor deviation will impact the wafer performance.”
Now, engineers from Central Engineering are embedded in the Applied business units designing new systems. Together, they are developing an upfront understanding of components, cleaning requirements, chamber matching, and health-monitoring techniques. The cooperation between the Applied system business units and the AGS organization, Cho said, is “a perfect synergy, and the customer is a direct beneficiary.”
“If we invest up front, it helps the business unit sell systems, and helps us mature the product and understand the cost-per-wafer, chamber matching, and cleaning requirements up front, rather than AGS learning afterwards. That gives us a better chance to meet the CoO target of the customer,” Cho said.
AGS has been qualifying parts suppliers located closer to major customers, often in less costly regions. “We have been moving consumables and spares to places closer to the customer location, so we can close that loop faster. If we make it there, and consume it there, the logistics costs and cycle times are lower, and we have an AGS-tailored supply base nearby,” he said.
Flying spares from a parts manufacturer in one country, to a cleaning specialist in another, and then to a parts center in a third country, often involves unnecessary logistics and costs. “We are bringing it together, all in the same location or as close as possible,” Cho said.
On the technology side, a major push is to develop “intelligent components” that can perform self-monitoring and contribute to data collection and system-level health checks. Intelligent components are needed to match chamber-to-chamber and system-to-system, a key demand from fab managers that must ensure Angstrom-level performance.
“Even in the 90s,” Cho said, “we had robots that could do auto-calibration, auto-zeroing, internal zero checks. And mass flow controllers also had selfcalibration capabilities. We are moving to the next level, so when we integrate those components we can do integrated calibration and can perform self-help at the chamber/system level.”
RETURNING TO A STEADY STATE
The science of keeping parts clean is a far cry from an earlier era when parts cleaning involved making sure there was no residual grease on the part. Now, the goal is to return parts to the “steady state” of the original recipe. A part’s optimum surface condition might involve a monolayer of aluminum oxide, for example.
“When we clean the part, we have to rejuvenate the part to that steady state. With chamber matching we have to make every wafer, every die, feel and see the same thing. If the part requires a layer of aluminum oxide, that is the state that we recondition to,” Cho said.
GREEN-TO-GREEN IN ONE SHIFT
One project reduced the green-to-green time for chamber liner maintenance down to one shift, versus the previous 40 hours needed to season the chamber and bring the process back to the desired state. “We looked at the morphology of the surface, and recreated the mix, not just to clean, but to build up to the needed surface condition. We brought that green-to-green time down from 40 hours to one shift,” Cho said.
Much of this metrology work is done at the chamber and system levels, partly to avoid driving up the cost of individual components. AGS has the material science and metrology expertise to determine the radicals’ finite time constant, and how they react with the chamber surface. To determine the number of radicals generated by a remote plasma source, for example, AGS relies on internal chamber metrology.
“The supplier may not have the means to know how many radicals they are generating. That would be stretching our suppliers’ capabilities. We don’t want to incur more costs for the suppliers. Everyone is under a lot of pressure to drive down costs, and anything we ask of the supplier side results in added costs. We can develop additional software and do component health checks from our controller, not only for that component but calibrated for the whole system,” Cho said.
“We are really moving up the spectrum. In terms of the order of complexity, each node goes up 10X or even 100X, and we are optimizing minute things to get the maximum performance,” Cho said.
UP THE TECHNOLOGY LADDER
Cho said Applied has expanded its custom engineering operation, which works with individual customers on ways to either reduce process costs or tweak a tool’s capabilities.
While Applied has been engaged in custom engineering for more than 20 years, over the past few years the group has grown to more than 150 AGS engineers working with customers.
In some cases, Cho said, the original tool might be equipped with parts that exceed the performance needs of a particular customer. In that case, less expensive parts are sourced and tested. In other cases, major co-engineering efforts are launched to modify parts and hone a tool’s capabilities.
Lee said this custom engineering work extends to 200mm fabs. “Even though most 200mm tools are fully depreciated, cost is a still a factor fab managers must address. Having solutions to help our 200mm customers continue to run profitable operations is a critical capability we want to expand.”
SUPPORTING LEGACY TOOLS
|David A Kirkpatrick,|
AGS Parts and Repair Operations
AGS ships roughly 20,000 parts a week, many of them to support legacy tools where the original parts supplier may no longer be in business. David Kirkpatrick, a managing director in AGS parts and repair operations, said Applied Materials takes a strategic view of the key role spare parts play in keeping customers’ fabs running smoothly. “It is priority number one to have parts to ship on time, and with fill rates to support our customers,” he added.
In today’s robust environment, Kirkpatrick said some suppliers are running unprecedented backlogs. “The large increase in wafer fab equipment (WFE) last year outpaced the historical capacity of our supply chain. Ensuring available spare parts is our number one priority. Significant inventory investment and a focus on increasing supply-chain capacity is also our number one priority,” he said.
Accomplishing that goal requires optimizing spares inventory at locations around the globe. That global logistics system goes beyond the capabilities of small companies that in some cases re-engineer Applied parts, supplying them on a piece-part basis.
SERVING FUTURE MANUFACTURING
Lee said his view is that the semiconductor industry has derived so much value from advances related to Moore’s Law that it has sometimes overlooked the value-generation potential of improving business processes. “The semiconductor industry is going to have to mature in new ways. By placing our customers’ success first, and reimagining the way we supply the operational needs of semiconductor fabs, we are creating new business models that add value through assurance of supply, quality, and performance—all at competitive costs.”
For additional information, contact firstname.lastname@example.org.