News
STMicro Advances PiezoMEMS Development in Singapore
STMicroelectronics, in collaboration with A*STAR and ULVAC, has launched Lab-in-Fab 2.0 at its Ang Mo Kio campus, focusing on lead-free piezoelectric MEMS. This initiative accelerates the transition from prototyping to manufacturing, aiming to achieve a sustainable alternative to lead-based materials by 2027. The Lab-in-Fab model enables seamless integration of research and production, enhancing time to market for MEMS applications in consumer electronics and medical devices.
Launches
3D Chip Stacking Method Created to Overcome Traditional Semiconductor Limitations
Researchers from the Institute of Science Tokyo have developed a novel 3D chip integration method named BBCube, which employs advanced bonding techniques to enhance memory bandwidth and reduce power consumption in high-performance computing applications. This innovative approach addresses traditional semiconductor limitations by stacking processing units directly above DRAM, significantly improving power supply integrity and efficiency.
Charts
Smart Factories Need Intelligent Components
At the MES & Industry 4.0 conference, industry leaders discussed the pivotal role of data in the evolution of smart factories. Key insights highlighted that achieving data integrity is critical for manufacturing efficiency. Notable advancements in AI integration into manufacturing execution systems (MES) were presented, demonstrating the trend towards intelligent factories capable of real-time data-driven decision-making.
Research
Machine Intelligence on Wireless Edge Networks with RF Analog Architecture
A groundbreaking study from MIT and Duke University introduces MIWEN, an RF analog architecture that facilitates deep neural network inference on power-constrained edge devices. By eliminating local weight memory, this innovation significantly reduces energy consumption and enhances real-time inference capabilities in low-power environments, paving the way for more efficient edge computing solutions.
Insight
SambaNova Shifts To Inference, Courts Cloud Customers
In an exclusive interview, SambaNova’s CEO Rodrigo Liang discussed the company’s strategic pivot towards inference workloads amidst a shifting market landscape. With a focus on open-source models and a robust cloud deployment strategy, SambaNova aims to support enterprises in leveraging AI for enhanced operational efficiency. Liang emphasized the importance of customization in AI models to deliver superior value and performance.
Stay tuned for more updates as we continue to bring you the latest in semiconductor innovations and trends!