Celera Incorporated
Automating Custom Analog IC Development

Although the chip industry has scaled massively over the years, the challenges such as cost, time, resources, and risk needed to develop new products remained persistent. Especially when it comes to analog/mixed integrated circuits (IC), R&D inefficiency poses a significant roadblock for manufacturers. This primarily happens owing to the ineffectiveness of the piecemeal ‘tools’ approach for analog. The approach tends to be much smaller and has lower integration, relying more on customization and parametric performance to deliver value. That is why analog R&D has lagged in productivity and resulted in long schedules and higher expenses. Moreover, analog requires highly specialized skills, talent, and costs a disproportionate share of the R&D budget and cycle time than the overall industry. A more analog-specific and turnkey approach is needed as analog functions become even more critical in sensors, actuators, motor controls, and power management in the era of connected and autonomous devices—a software-powered development service, not just a tool.

Enter Celera Incorporated.

Celera Incorporated envisions a new paradigm in chip development to support the unprecedented demand for new semiconductor solutions, i.e., a third-generation - Silicon 3.0. “We dramatically improve the R&D efficiency and reduce the risk with automation for analog,” says Abid Hussain, Co-founder and CEO, Celera. “The Silicon 3.0™ generation is inevitable. To deliver on the promise of a $1 trillion per year and beyond, the industry must develop silicon much faster and more efficiently.

Celera’s automated development platform delivers up to 100x faster chip design, 10x faster time-to-samples, and 5x faster time-to-revenue. We are focused on dramatically increasing silicon development throughput and efficiency with software automation resulting in faster custom silicon realization with lower upfront risk and cost.”

He further notes that as there is no guarantee of success in a highly competitive chip market, the bulk of ROI risk is tied up in modeling, IP, synthesis, layout, and verification parts of the development flow. Celera’s solution applies intelligent expert system AI software to each part of the development flow. The software acts as an extremely fast, tireless, and error-free robotic engineer to make design decisions under the algorithmic guidance of the platform developers’ chip design experience built into the software.

This methodology is implemented in the entire flow from modeling to verification and delivers a fab-ready chip design from a customer specification input. Delivered as a SaaS platform, it is an end-to-end service approach to the IC development flow from spec to silicon, including modeling, synthesis, physical layout, and verification. This automated software architecture enables scaling far beyond any legacy methodology. “Using our platform in a self-service model for development, white-labeled third party supply chain partners can handle silicon fabrication, assembly, and testing as well as fulfillment services,” adds Hussain. Further, the platform delivers over 25 percent TCO improvement during the product lifecycle for OEMs while securing more control of their supply chain, inherent exclusivity, and IP protection. In many use cases, clients dramatically reduced the one-year development time to a few weeks with lessened risks using Celera’s software.

Also, Celera’s solution provides clients the ability to fan out and scale to do as many designs as they want without being limited by their infrastructure. It is only limited by the number of instances clients wish to use the software. With an aim to become the first SaaS platform for semiconductor R&D, Celera plans to commercialize its platform into the traditional semiconductor and the emerging OEM industry for more custom Silicon. “We believe that R&D acceleration must happen in all segments, especially analog semiconductors. With the persistent global shortage of skilled talent and ever shrinking product cycles, this field requires an automated software platform,” concludes Hussain.