Group aims for quicker chipmaking fixes
A group of companies including chipmaker Advanced Micro Devices is mounting a new effort to create smarter semiconductor-manufacturing equipment.
The group, which also includes domainLogix, ILS Technology and Oceana Sensor Technologies, said this week that it will work to create software and industry standards intended to help make possible--for the first time--online diagnoses of chip-manufacturing equipment.
The trade group Semiconductor Equipment and Materials International (SEMI) will oversee the three-year project, which has a $10 million budget, funded in part by a $5 million grant from the U.S. government's National Institute of Standards and Technology. The remaining $5 million will come from the group's members.
The group's main thrust will be to create a secure method of communications that would allow manufacturing companies to share data online. Using this security framework, chipmakers could share data with the companies that produce their manufacturing equipment and with their own customers.
Ultimately, the group hopes its efforts will result in speedier diagnosis and resolution of problems, which could reduce manufacturing equipment downtime and lead to lower manufacturing costs for products such as microprocessors.
It's no surprise that AMD is in the thick of the new effort. The Sunnyvale, Calif.-based company, often seen as playing second fiddle to Intel's industry lead, has been quietly remaking its image of late, moving away from that of mere chipmaker and toward the role of technology provider.
"AMD wants to be on the leading edge of the development of these systems," said Charles Clark, director of the company's global manufacturing systems.
Currently, when a piece of chipmaking equipment acts up, the chipmaker needs to dispatch a technician--one of its own or a representative from the company that built the machinery--to the factory floor to look at the problem firsthand. The proposed software and standards would allow those workers to figure out and fix the problem from a distance.
The approach is similar to corporate tech support methods that allow remote diagnosis of problems with PCs and servers. Instead of making a trip to an employee's desk, a technical specialist can use a software application to peer into the computer in question online, discover problems and attempt repairs.
At the moment, no software exists that provides remote diagnosis for chipmaking equipment, said Clark.
"The object here is to provide an application that would enable a secure mechanism that would allow (equipment) buyers and companies to communicate with each other" through online means, Clark said. Chipmakers already have a system in place that allows PC makers and other customers to place secure online orders for chips, he said.
Establishing secure online communications is the first step to sharing data, but when exactly that will happen remains up in the air. The proposed software application eventually will be made available to manufacturers, but terms of its release have not yet been decided upon.
Such technology could also improve manufacturing yields through early and automatic awareness of potential problems such as a machine gradually becoming less precise in its operation.
The diagnostic technology won't replace a person repairing a broken or worn part, but AMD and analysts say that creating smarter manufacturing equipment will reduce the cost of manufacturing its chips--cost savings that could then be passed along to consumers.
"We see that as a major lever in the future of reducing costs," Clark said.
"Overall, if equipment stays running longer and you find problems early, you save money," said Kevin Krewell, senior analyst with MDR-Instat. "Anything you can do to improve your yields lowers cost of manufacturing."
Since it launched the Athlon chip in 1999, AMD has created and licensed technologies such as Hypertransport, which provides a high-bandwidth data link between chips. Hypertransport is now freely available and governed by an industry consortium, and AMD plans to use it in its own chips beginning in 2002.
In addition, AMD will make use of its expertise in Advanced Process Control, an automated system that uses data collected from manufacturing equipment throughout the different stages of the chipmaking process to adjust certain formulas or equipment. The leap between this technology and online diagnosis isn't that large, AMD said.