These are lots that are temporarily suspended during testing due to a concern about yield, the failure rates of certain bins or the quality of testing. Obviously, any Lot on Hold will cause greater inefficiency in a testing facility. Ideally, you want all lots to be tested and, if final test, sent on to QA and finished goods as soon as possible.

How are “Lots on Hold” typically managed?

The testing facility analyses the data in conjunction with the owner of the lot (typically a fabless customer or IDM, or another division within their own company) and a decision is made to re-test, release or scrap the lot. Meetings are arranged per customer on a weekly basis or more often than that depending on how close the working relationship is between the two companies.

The decision time can be days, even a week or more depending on the issues encountered. Multiply this by hundreds of lots a quarter in a typical fairly large testing facility and the size of the problem to both sides is clear.

A new way of managing “Lots on Hold”

By building the rules into yieldHUB, datalogs exhibiting potential testing and quality issues appear on a web-based dashboard which is shared by the fabless company and the test house. Tools are at hand to analyse the datalogs very quickly. If required, special yieldHUB licenses are provided to the fabless company or IDM to make sure they only see their own data.

This makes signing off any lot much faster. Days to make decisions are reduced to minutes. No need for meetings except to confirm any unusual outstanding issues. But most issues can now be resolved quickly and efficiently. Experts who know the product can now make the decision on-line from thousands of miles away from the test facility, seeing exactly what the test facility is seeing.

Integrating yieldHUB (via its new API) with the testing facility’s MES system enables the most efficient Lots on Hold system available.