Adding the 4th V in Semiconductor Big Data Dimensions
Posted By: Irteza on October 10, 2016
The semiconductor manufacturing industry has witnessed drastic changes over the time and these changes have made semiconductor manufacturing one of the most complex and highly advanced industry where each product being manufactured goes through a number of specialized steps in the supply chain before reaching the end consumer.
Each step in this process generates huge amount of manufacturing data that is collected and stored by the engineering team for analysis. This manufacturing data is analyzed to drive critical decisions directly linked to production planning and operations, which impacts product quality, manufacturing yield and production ramp-up. With the increase in the number of semiconductor devices being produced annually, the data being generated is increasing manifolds too, making it relevant and necessary to deploy semiconductor big data solutions.
Big Data in semiconductor manufacturing and testing was initially defined in the following three major dimensions 1.)Volume (tons of data being produced and stored in the manufacturing supply chain), 2.) Velocity (speed with which the data is being generated and stored) and 3.)Variety (various data types and formats being generated across different steps of the manufacturing process by different vendors) but recently the fourth dimension has been added which gives the rationale behind deploying big data analytics for semiconductor manufacturing. The fourth dimension or the fourth V is Value (the analytics or insights), which is made possible by semiconductor industry big data analytics solutions like yieldWerx. It also explains the benefit of capturing and storing huge amount of data points that cannot be processed and analyzed with the traditional solutions.
As per various industry report the data volume is doubling every two years and will cross 40 trillion GB data mark in the next 3 years. This huge amount of data and a strong focus on product quality, yield and operational efficiency makes it important to analyze and improve product and equipment related issues. Similarly, on the customer end there is an increased expectation of faster delivery of high quality and reliable products. The yield management solutions allows you to access and store all the data in a common and standard data format, which can be correlated from multiple steps in the manufacturing process irrespective of the geography of the facility or the vendor from where the data is sourced. This makes it possible to have a standardized data for all the business units as well as the management tiers. The insights and reports can be customized depending on the need of the user, where by the user can be someone in engineering, product planning, operations or someone in the top management thereby deriving real value for the users of big data analytics.
These solutions not only allow the engineering staff to provide valuable analytics, but also the ability to control operation at the manufacturing floor by sending out alarm messages if inline control parameters are out of control limits. In some cases the solution is integrated with the production line and manufacturing execution system (MES) thereby providing support to real-time decision making and improving the operational efficiency. As a result, the semiconductor manufacturers are able to effectively manage their test floors, optimize their production planning, reduce test time, improving yield and throughput, thus ramping the products faster to market – providing the real value of big data analytics in semiconductor manufacturing.
For more information on how yieldWerx can help improve production yield, reduce test time and ramp products faster to market, sign up for a free 15 days trial period.