Boost Modeling Efficiency with the MBP/MQA 2026 Update 1.0 Enhancements

Modern device modeling continues to evolve as complexity increases across advanced technologies and applications. Engineers are expected not only to achieve high accuracy, but also to reduce iteration cycles, improve efficiency, and manage diverse data types.

Within the Keysight Device Modeling Suite, the Model Builder Program (MBP) 2026 Update 1.0 addresses these challenges by introducing enhancements that streamline SRAM modeling, accelerate ML-based extraction, and reduce manual data handling effort. These updates are designed to help users focus more on model quality and insights, rather than workflow setup and data preparation.

In addition, the Model Quality Assurance (MQA) 2026 Update 1.0 introduces enhancements that strengthen aging model validation capabilities and expand simulation support with Synopsys PrimeSim SPICE.

Accelerate SRAM Modeling

SRAM modeling traditionally requires significant manual effort to define fitting targets, process circuit-level data, and align transistor characteristics with system-level behavior.

MBP 2026 Update 1.0 simplifies this process by introducing pre-defined SRAM modeling targets, supporting both DP and IV data, enabling users to:

Quickly set up key SRAM metrics, including SNM, write margin, and critical currents
Simultaneously evaluate SRAM and transistor-level characteristics, ensuring consistency across abstraction levels
Reduce manual data preparation, eliminating the need for custom scripts and preprocessing flows

With these capabilities, engineers can move from data loading to meaningful analysis much faster, while maintaining full visibility into both circuit and device behavior.

Figure 1. Overview of SRAM characteristics and transistor IV results

Leverage ML-Driven Flows for Faster Convergence

As model complexity grows, traditional optimization approaches often struggle with convergence efficiency and robustness.

MBP 2026 Update 1.0 extends its ML-based optimization capabilities by introducing three new predefined extraction flows: BSIM-BULK, PSP, and VBIC.

These flows are designed to help users:

Accelerate convergence toward optimal solutions using ML-driven optimization strategies
Reuse proven workflows across similar technologies, reducing setup time for new projects
Handle complex, high-dimensional parameter spaces more effectively

By providing ready-to-use flows, MBP enables users to adopt advanced ML optimization without needing to build workflows from scratch, significantly lowering the barrier to entry.

Figure 2. Three new ML‑driven modeling flows implemented in the MBP Python Editor

More Flexible and Intuitive Equation-Based Parameter Tuning

As equation-based modeling becomes increasingly common in subcircuit netlists, engineers need more efficient ways to control and fine-tune model behavior. Adjusting individual constants, testing small variations, and maintaining equation integrity can quickly become time-consuming and error-prone with traditional approaches.

MBP 2026 Update 1.0 addresses these challenges with an enhanced Parameter Panel designed to make equation-based tuning more intuitive, interactive, and scalable.

Within this improvement, users can:

Accelerate parameter tuning by adjusting equation constants directly through intuitive, interactive controls
Increase modeling efficiency by quickly exploring parameter sensitivity and observing immediate impact on results
Enhance tuning flexibility using delta-based adjustments (value = base + delta), enabling incremental refinement without modifying base definitions
Streamline workflows with a detachable Parameter Panel, allowing more flexible and focused interaction
Overall, these improvements transform equation-based parameters tuning from a manual, iterative process into a more efficient, interactive workflow, enabling engineers to achieve faster convergence and more reliable modeling results.

Figure 3. GUI for tunning constant and delta value in updated Parameter Panel

Small Updates, Meaningful Impact

Other feature updates in MBP/MQA 2026 Update 1.0 make result analysis more intuitive, actionable, and efficient, helping users identify issues earlier and refine models with greater confidence.

These updates empower users to:

Identify fitting quality at a glance using color-coded error highlighting in tables
Improve optimization performance by enhancing parallel execution
Enhance data processing efficiency with flexible Python plugin support
Accelerate model validation with enhanced aging QA workflows
Expand simulation support with Synopsys PrimeSim SPICE.

Conclusion

MBP/MQA 2026 Update 1.0 delivers a range of enhancements focused on making device modeling workflows faster, more intuitive, and more efficient. From simplified SRAM modeling and ML-driven extraction flows to improved parameter tuning and enhanced data analysis capabilities, these updates are designed to help engineers reduce manual effort while gaining deeper insight into model behavior.

By combining automation, flexibility, and improved usability, this release enables users to accelerate convergence, streamline complex workflows, and maintain high modeling accuracy across a variety of applications. Whether working on advanced transistor models or circuit-level validation, engineers can now achieve reliable results with greater confidence and efficiency.

Learn more

limit