Daniel Babbs

Effect of Reduced Equipment Downtime Variability on Cycle Time in a Conventional 300mm Fab

The semiconductor industry is pursuing further cycle time and cost reductions through an initiative termed 300 mm Prime. One key area of focus is improving process tool availability. Equipment downtime-particularly unscheduled down time due to a failure-causes disruptions to production, loss of productivity, and higher costs. This paper focuses on fab productivity improvements resulting from the reduction of tool downtime variability. Preventive maintenance can reduce the risk of unscheduled downtime. Discrete-event simulation is used to investigate the relationship between downtime variability and overall cycle time in a conventional 300 mm fab with batch tools. Variance reduction is modeled by reducing the standard deviation of the time to repair a failed tool and by increasing the portion of tool downtime reserved for regularly scheduled preventive maintenance. Also, the effect of reserving capacity in the form of additional tools is assessed and compared to the effect of reduced downtime variability.

Effectiveness of Small Batch Size on Cycle Time Reduction in a Conventional 300mm Factory

The semiconductor industry is increasingly challenged to improve manufacturing productivity while keeping pace with technology advances and maintaining cost per function goals to meet the demands of their served markets. The task of improving factory productivity is influenced by shrinking device geometries and manufacturing methods which reduce overall cycle time. Given the complexity of semiconductor manufacturing, production ramp time and the required capital investment; the decision of which path to pursue must be balanced against the business objectives of each device manufacturer. This paper focuses on productivity improvements applicable to existing factory architectures and more specifically the reduction of the move batch size, more commonly known as the transport lot size, and its effect on the overall factory cycle time. Multi-variable, discrete-event factory simulations are used to illustrate the relationship between transport lot size and the overall cycle time using a conventional 300 mm factory architecture with batch tools.