Rudolph Releases Back-End Manufacturing Review SoftwareOctober 1, 2009
Discover fast review and die sort classification software is able to dramatically reduce time and cost of defect review in back-end process environments.
FLANDERS, NJ -- Rudolph Technologies, Inc. (NASDAQ: RTEC), a worldwide leader in process characterization solutions for the semiconductor manufacturing industry, announced the release of Discover Fast Review and Die Sort Classification Software for back-end manufacturing. It is specifically designed to dramatically reduce the time and cost of reviewing backend manufacturing defects detected by the NSX Series automated macro defect inspection system.
Discover automatically classifies defects, either on-line or off-line, eliminating the time-intensive need to review multiple instances of similar defects. It also distinguishes nuisance defects that do not have an impact on yield from those that should receive further off-line investigation, reducing costly, labor-intensive manual review. Discover is now available on new NSX Systems or as a retrofit for existing NSX installations.
"Defect review is an expensive, time consuming and labor-intensive task," said Rajiv Roy, Rudolph's vice president of business development and director of back-end marketing. "Automated tools detect the defects, but humans must ultimately decide which are important. The problem is only made worse as the size of potentially significant defects decreases and they become more numerous, making them more time consuming to see and evaluate manually. Fast review and die sort classification automatically sorts defects into similar categories, and then identifies those most likely to impact yield and, subsequently, receive manual review. We have seen improvements in tool and review efficiency ranging from 18 to 40 percent, with commensurate reductions in time and cost of the review process."
Discover's fast review and die sort classification integrates a number of specific analytical operations to optimize the review process. Failed die may be identified based on criteria such as the number and size of defects within the die; these defects may then be included or excluded from further review. Additional failed die, and defects within them, may be designated and excluded from further review based on custom rules, such as the failing of neighboring die for certain defect types. Spatial pattern recognition can associate multiple defects in wafer scale patterns resulting from a single cause, such as a scratch or a stain. Repeating defects, found in the same location within die, and local clusters may also be identified. Nuisance defects that do not impact yield are eliminated. After classification, one or more representative defects are then selected for review and image capture. It is not unusual for the number of required image captures to be reduced by a factor of ten or more.
About Rudolph Technologies
Rudolph Technologies Inc. is a worldwide leader in the design, development, manufacture and support of high-performance process control metrology, defect inspection and data analysis systems used by semiconductor device manufacturers. Rudolph provides a full-fab solution through its families of proprietary products that provide critical yield-enhancing information, enabling microelectronic device manufacturers to drive down costs and time to market. The company has enhanced the competitiveness of its products in the marketplace by anticipating and addressing many emerging trends driving the semiconductor industry's growth. Rudolph's strategy for continued technological and market leadership includes aggressive research and development of complementary metrology and inspection solutions. Headquartered in Flanders, NJ, Rudolph supports its customers with a worldwide sales and service organization..
Most Read Articles