As part of a collaborative research effort, our team has developed an entirely novel approach for determining the mean and limits of functional data analysis, called the PMM (Pairwise Midpoint Mean) method. We have shown that the newly developed method is both much faster and far more accurate than earlier state-of-the-art techniques.
Building upon the PMM device developed by our team, which efficiently represents the mean behavior from a set of electrical devices, we have also developed a method to define the bound region. These bound devices will enable users to easily identify faulty electrical devices and assess potential aging deterioration. Additionally, this approach can provide insights into the expected lifetime of advanced electrical devices, making it a valuable tool for maintaining the reliability and performance of sophisticated systems.
Precise mean characteristics curves for a range of active and passive electronic components, such as diodes, transistors, resistors, inductors, and capacitors, can be produced with this tool. Device characteristic curves can come from observed I-V sweeps obtained in the lab across a range of processes and temperatures, or they can come from SPICE simulations. With the use of this tool, semiconductor designers can produce fabrication recipes or refine their designs to increase device yield and reliability.
As research into High Electron Mobility Transistor (HEMT) devices continues to expand, enhancing the performance of these semiconductor devices has become increasingly critical. HEMTs are essential for high-frequency and low-voltage operations, but they face significant challenges due to the severe thermal issues that arise during high-frequency operation. These thermal limitations can hinder device performance and reliability. Optimizing the thermal behavior of HEMT devices is crucial for advancing industrial and high-tech applications. Our work focuses on mitigating thermal spikes and designing more efficient heat transfer systems for HEMT devices, which will play a key role in improving their overall performance and enabling further advancements in the field.
Under a joint research collaboration, our team has developed a novel method to evaluate the Xyce Simulation software. This is done by analyzing some benchmarks circuits including Compensated CMOS Operational Amplifier, CMOS Operational Amplifier, Continuous Time State Variable Filter, Leapfrog Filter, Single State Common Emitter Filter and Differential Amplifier using the software.
Our team has introduced a novel approach to optimize data processing latency, speed, and power by developing custom Linux kernels and drivers to support PCIe communication between the FPGA and HOST, as well as FPGA to SSD.
Under a joint research collaboration, MetroScientific performs designing, implementing, testing, and maintaining firmware for FPGA & involves working with digital logic, programming languages, and hardware description languages to create efficient and reliable FPGA designs.
© 2026 Onera-AI Technologies. All Rights Reserved.
