Recently, the research team led by Professor Xie Guangjun and Professor Zhang Zhang from the School of Microelectronics at Hefei University of Technology has made a series of remarkable strides in the realm of artificial intelligence chip design. Concentrating on the development of high-performance Analog-to-Digital Converter (ADC) chips tailored for audio applications, including intelligent speech recognition, the team has introduced an innovative technique known as "Bidirectional-Advancing Data Weighted Averaging (Bi-ADWA)." This groundbreaking approach has enabled the successful design and fabrication of a continuous-time Delta-Sigma modulator chip, boasting key performance indicators that rival international standards.

The research findings have been published in IEEE Transactions on Circuits and Systems I: Regular Papers (TCAS-I), a premier journal in the integrated circuits field. Utilizing a 180 nm CMOS process, the chip achieves an impressive dynamic range (DR) of 102.5 dB and a peak signal-to-noise-and-distortion ratio (SNDR) of 95.8 dB within a 20 kHz audio bandwidth. Furthermore, it exhibits a spurious-free dynamic range (SFDR) as high as 109 dB, substantially enhancing the quality of audio signals. Operating at a mere 1.8V voltage, the chip consumes only 0.58 mW of power and occupies a core area of 0.8 mm². Its comprehensive performance metric, known as the Schreier Figure-of-Merit, reaches an exceptional 177.9 dB, positioning it at the forefront of international standards.

In addition to these achievements, the team has also made significant progress in the field of memristor-based in-memory computing chip design. Relevant research results have been published in esteemed journals such as IEEE Transactions on Very Large Scale Integration (VLSI) Systems. These accomplishments not only underscore the team's profound expertise in analog and mixed-signal integrated circuit design but also provide essential chip technology support for applications like next-generation intelligent speech front-ends and portable high-definition audio devices.