Optimizing Reliability Certification Data Collection In Order To Accelerate The Production Process Of New Product

Integrated Circuit(IC) was invented only about 50 years ago in 1958 by Robert Noyce,one of the founders of Intel Corporation.But the rapid development of IC industry and the impacts to the world are unbelievable.Along with continual growing of the application scope and the importance to our lives,people are more and more caring about the quality and reliability(Q&R) of IC products.Especially,as one of the top level products in designing and manufacturing technologies,how to make sure CPU can well meet the Q&R requirements from customers or end users became one of the most difficulties that the manufacturers must solve.Every semiconductor device manufacturer will use its own appropriate way when the company decides to develop a new product.The company I am working for is always following a standard model to develop our products.This model covers the whole life of the product,from market research through production and eventually to product discontinuance.There is also a related certification and qualification system of Q&R to support the model.However,the ruthlessness of modern market competition requires the device manufacturers must try their best to shorten the throughput time(TPT) of new product development.When the product has similar function and pricing with competitors,the manufacturer will have to bring its product to market earlier than the competitors to get higher percentage of market share and earn additional profits.It's like a situation of "quick fish eats slow fish".Meanwhile,the shorter new product development TPT can help reduce the cost of development.Through analyzing the new product development qualification system,I found although the current product qualification methodologies can well meet the Q&R commitments to customers,there are a number of redundant experiment and data collection during qualification.This is very obvious especially in a complex product family qualification.Basing on the knowledge on the product and risk assessment,I believe the current methodologies can be optimized to accelerate the product to production progress.This thesis consists of four chapters.I will find out the disadvantages and the areas need to be improved by analyzing the current qualification system.After that, the optimization or improvement directions will be proposed.And a series of well designed experiments on a mobile CPU will follow to validate the proposals.In the last of this thesis,I will share the actual product qualification status with the implementation of new optimization methods.And prospect the future challenges of product qualification.