Dr. Meghali Chopra is co-founder and CEO of SandBox Semiconductor. She is responsible for SandBox’s vision and strategy and oversees the development of SandBox’s software products and technologies. Dr. Chopra received her PhD in Chemical Engineering from the University of Texas at Austin where her research focused on computational algorithms for plasma process optimization. She has her B.S. with Honors in Chemical Engineering from Stanford University. Dr. Chopra is an industry expert with publications in leading peer-reviewed journals and patents in the areas of semiconductor processing and computational optimization.
Tell us about your company?
Founded in 2016, SandBox Semiconductor is a pioneer in developing AI-based software solutions to accelerate process development for semiconductor manufacturing. Our fully integrated, no-code AI tool suite gives process engineers the ability to build their own physics-based, AI-enabled models to solve challenges during process definition, ramp-up, and high-volume manufacturing.
Using SandBox’s physics-based models and machine learning tools, process engineers in the semiconductor industry can virtually simulate, predict, and measure process outcomes. Even with small sets of experimental data, SandBox’s tools can extract valuable insights and patterns, helping engineers to gain a deeper understanding of manufacturing processes and to make informed decisions about recipe adjustments. SandBox leverages expertise in numerical modeling, machine learning, and manufacturing optimization to develop its proprietary toolsets, which are used by the world’s leading chip manufacturers and semiconductor equipment suppliers.
What problems are you solving?
At SandBox, we reduce cycles of learning for next-generation advanced manufacturing technologies. To optimize a recipe, a process engineer must specify a process window for tens of process conditions including pressure, temperature, and gas flow rates. Determining the best process conditions is so challenging that oftentimes a recipe will take over two years to develop, or worse, the chip is dropped from production because the cost of the process development becomes too expensive. This technology gap and cycle time is a significant barrier to the deployment of novel microelectronic devices and imposes a substantial economic burden on semiconductor manufacturers who must make significant R&D investments to stay afloat.
SandBox provides computational modeling software that accelerates process development and enables semiconductor manufacturers to reduce costs, get to market faster, and commercialize new processes not possible before.
What application areas are your strongest?
SandBox works on leading-edge logic and memory manufacturing processes. Our users are typically performing technology development or high-volume manufacturing recipe optimization. Our technologies have been used on a range of optimization applications including feature-level, die-to-die, across-wafer, chamber-to-chamber, and tool-to-tool.
What keeps your customers up at night?
The process engineers we work with must figure out how to optimize many process conditions to manufacture billions of features across the wafer with nano-scale precision and at high throughput. These process engineers are extremely knowledgeable and arguably the single most important individuals within each of our semiconductor customers. Unfortunately, these process engineers are often over-worked as they must continually push the envelope in advancing to the next node. We developed our tools with these process engineers in mind – our mission is to provide meaningful leverage to the process engineer as he or she works to enable manufacturers to bring new microelectronics to market faster.
What does the competitive landscape look like and how do you differentiate?
Our proprietary modeling pipeline enables users to make process predictions with a small number of experimental data points. The competitive landscape for process engineer-focused computational modeling tools is very limited. Many of our customers have internal modeling groups, but our observation is that most frequently our process engineering users rightfully rely on their expertise and intuition to drive critical changes in recipe development. To that end, the most common recipe optimization approach is the process engineer’s intuition. We seek to help these process engineers in their role, particularly as the advanced manufacturing nodes increasingly push the limits of physics and chemistry in conjunction with the process engineer’s demands in a 24-hour day.
What new features/technology are you working on?
SandBox recently released a new product for its technology suite called Weave™. Weave™ significantly improves metrology accuracy and precision by leveraging advanced machine learning capabilities to extract and analyze profiles from SEM and TEM data. Process development engineers can spend up to 20% of their time manually measuring SEM and TEM images. With Weave, process engineers minimize tedious manual tasks and increase metrology accuracy, resulting in more insights, quicker experimentation, and reduced costs during process definition, ramp-up, and high-volume manufacturing. The introduction of Weave continues on our platform vision as we work to provide a comprehensive tool-suite to bring easy to use physics-based AI tools to market with the goal of enabling the process engineer.
How do customers normally engage with your company?
Customers can reach out to us at info@sandboxsemiconductor.com or through our website at www.sandboxsemiconductor.com.
Also Read:
CEO Interview: Dr. J Provine of Aligned Carbon
CEO Interview: Vincent Bligny of Aniah
Executive Interview: Tony Casassa, General Manager of METTLER TOLEDO THORNTON
Share this post via:
Intel High NA Adoption