IDMs are typically large companies with multi billion dollar sales. They have their own internal foundries and sometimes internal test facilities. The challenge for these companies is managing a huge number of products in manufacturing across multiple product lines.

The scalability of yieldHUB databases means that these companies can centralise all their data and integrate it for powerful correlations right across manufacturing.

Data Management Challenges for an IDM

A prospect was being supplied by several wafer fabs, using several assembly houses and test subcons. Engineers in Asia, the US and Europe worked together with tight schedules to release new products. Yield issues occurred but often they were not aware of these until too many wafers were started, resulting in high scrap levels.

The company found that they were reacting to yield problems rather than preventing them. The engineers in Asia were using their own tools for analysing data ftp’d to them from Europe and the US. It was a tedious task trying to correlate performance across manufacturing stages which might prevent issues in the future. Releases were often delayed through absence of data.

In May 2010 they purchased yieldHUB predecessor called FloorVision, then yieldHUB in 2014, replacing ad hoc tools.

From Hours to Minutes

Now around 100 users from Fab, Probe, Test, Quality and Reliability have the production data they need in one place. They can share analyses in one click with remote colleagues, reducing decision times for releasing a product from one or two days to just minutes. They can correlate final test and wafer probe performance in seconds and can set personal alerts for any deviation from expected product performance in production.

They have told us that we are close to “revolutionising” their business. The engineers in different countries are getting to know each other a lot better too with greatly enhanced near real-time collaboration. yieldHUB typically saves them 1-5% of Yield on high running products.

Ongoing Work

Medium to large companies like the one in this case study will have new formats every so often and new needs for data analysis. We provide immediate support for any queries from our customers or to fix any bugs. We treat every company the same, be they large or small. We monitor the data pipelines closely and make sure any processing anomalies are checked by our technicians immediately. A larger company will typically ask us to develop or tweak a parser every few months. If a company needs a particular type of report or analysis type we work with them and they are assured of a cost-effective solution to their particular need at the time.

For larger companies we would install yieldHUB software on-site and not host it for them. We would also provide support for a mirror server.

A typical server configuration has multiple cores and at least 48 GB of RAM.

We use LAMP technology and provide data export also for JMP users and for spreadsheets.

However core analysis can be done using our web application without a need to download any data.

More information can be provided under NDA.

Can yieldHUB Help You?

Contact yieldHUB today! Our global sales and support team will be happy to help.