Blog

Big Inflections in the Semiconductor Industry Have Created Big Yield Problems

The semiconductor industry is currently experiencing major inflections in device architectures and process technologies (e.g. finFET and 3-D NAND). These inflections have presented very difficult and unprecedented challenges for the industry, resulting in low manufacturing yield. StreamMosaic’s founder has conceptualized specific ways to meet these challenges and improve manufacturing yield by using predictive analytics and process control. StreamMosaic’s founder has deep subject matter expertise in the semiconductor industry, and proven track record as an inventor and innovator of production-proven semiconductor process control solutions.

Read More

Are You Leveraging the Full Potential of Your Data? Put Your Massive Datasets to Work for You

Data has always played a role in semiconductor and electronics manufacturing. In the semiconductor industry, data was initially collected manually to track WIP (work-in-progress). In many instances, the analysis was performed in a very manual “ad-hoc” fashion by domain experts.

Over the years, data collection has become much more automated. But even with this improvement in the ability to collect data, it has been estimated that more than half of it is ever processed. And of this data that is processed and stored, most of it is never again accessed.

Moving forward, data volume and velocity is increasing rapidly, as more sensors are added and data collection rates on process tools are increasing by several orders of magnitude.

Recently there has been lots of interest and progress in the fast-growing field of Predictive Analytics. Thanks to the emergence of the ability to harness massive parallel processing architectures (Big Data) and the advancement of machine learning algorithms, we are now able to gain insights and make predictions using massive amounts of data at speeds that make such approaches relevant and realistic. At StreamMosaic, we will bring to bear the latest in machine learning and predictive analytics technologies to get the most out of your available data.

At StreamMosaic, we have the expertise in both Predictive Analytics and the Semiconductor and Electronics Industries to implement solutions that increase yield and decrease costs for our customers.

Read More

Continuing Moore’s Law with Predictive Analytics™

For decades, the semiconductor industry has been driven by Moore’s Law and the planar transistor architecture. This allowed for a predictable, self-sustaining roadmap for transistor cost scaling and well-defined interfaces where each individual process/layer could follow its own technology trajectory independently. But the days of scaling the traditional planar transistor have reached their end. High-K Metal Gate was able to reduce random dopant fluctuations (RDF) to extend the planar transistor technology to the 28nm Node. For the 16nm Node and below, entirely new device architectures are needed to limit RDF. New device architectures such as FinFET limit RDF by reducing dopant concentrations in the channel. But in order to achieve acceptable yield levels with these new architectures, very tight process specifications must be achieved. Thus, better process control and integration schemes are needed more than ever. The processes now also require more complex integration and can no longer be developed independently of each other. FinFET’s and 3-D NAND’s three-dimensional architecture, as well as more complex relationships between process steps, have changed the way that process variability affects device performance and yield. At StreamMosaic, we take a holistic approach to process control by integrating processes to improve yield.

Read More

StreamMosaic’s Value Proposition

The Value Proposition of StreamMosaic is to increase yield and lower costs by developing and implementing Big Data Analytics and process control solutions for the semiconductor manufacturing industry.
At StreamMosaic, we strongly believe that the predictive analytics solutions we bring to bear will have an industry-changing impact on the way semiconductors are manufactured. With a strong subject matter expertise in semiconductor manufacturing and core competencies in data mining and machine learning, we will close the gap between semiconductor manufacturing and Big Data Analytics.

Read More