GUGUO CLOUD

  Introduction to Computing in Memory Computing in memory, also known as in-memory computing or computational memory, is a new paradigm that aims to reduce data movement and increase data processing speed by performing computations directly in memory. In contrast to traditional computing architectures that rely on a separation of memory and processing, computing in memory systems leverage the inherently parallel processing capabilities of memory devices to achieve high-throughput and low-latency data processing. Advantages of Computing in Memory Applications Computing in memory has several advantages over traditional computing architectures. One of the biggest advantages is the reduction of data movement, which significantly reduces power consumption and improves overall efficiency. Additionally, computing in memory enables higher bandwidth data processing, faster data transfer rates, and lower latency. These advantages make computing in memory particularly attractive for applicati...

  The principle of producing feedback is that when a microphone and a speaker are used simultaneously, the sound signal captured by the microphone is amplified by the speaker and then output again, forming a feedback loop. Due to the existence of the feedback loop, some specific frequencies of the sound signal can become very strong after multiple amplifications, forming a howling noise. The principle of howling suppression is to use digital signal processing technology to monitor the audio input and output signals in real-time, and identify and eliminate signals that may cause howling. Specifically, the howling suppression algorithm will perform frequency analysis on the input and output signals, and when the strength of the signal in a specific frequency range exceeds a certain threshold, it will recognize that there is howling noise. Then, the howling suppression algorithm will use a certain algorithm model to eliminate or reduce the signals that may cause howling, thereby elimi...

  With the development of TWS technology and intelligence, TWS smart earbuds will play an important role in wireless connectivity, voice interaction, intelligent noise reduction, health monitoring, and hearing enhancement/protection, not just as standard equipment for smartphones, but even becoming an indispensable part of the human body. Noise reduction, hearing protection, intelligent translation, health monitoring, bone conduction+bone sound pattern, anti-loss, etc. will be the key technologies for TWS earbuds. Among them, noise reduction is the top priority, and the two mainstream active noise reduction modes on the market are ANC and ENC. Active noise reduction is achieved through the collaboration of hardware (chips, sensors, microphone arrays, etc.) and software algorithms. Currently, TWS earbuds mainly have two active noise reduction methods: ANC and ENC. (1) ANC noise reduction technology generates reverse sound waves equal to external noise through the internal noise redu...

For decades, headphones have been a crucial tool for people from all walks of life. From listening to music and podcasts to gaming and taking calls, headphones have become an indispensable accessory for many individuals. However, with advancements in technology, it's time to take headphones to the next level. This is where the concept of computing in memory comes in. Computing in memory (CIM), also known as in-memory computing, is a new technology that merges memory and processing capabilities into one device, offering significant speed and efficiency improvements over traditional computing architectures. The ability to integrate processing power and memory into a single device has enormous implications for the headphone industry. The technology opens up new possibilities for real-time audio processing, including noise reduction, equalization, and spatial audio, leading to a significant improvement in the overall listening experience. CIM technology allows for the integration of ad...

  WTM2101 is a speech recognition and wake-up chip based on computing in-memory technology, with a core consisting of a set of low-power processors specifically designed for speech recognition. Compared to traditional speech processors, WTM2101 adopts a series of optimization measures for speech wake-up and recognition scenarios, significantly reducing power consumption. One of the most important optimization measures is the full phoneme algorithm model. Traditional speech recognition algorithms require a large amount of acoustic and linguistic knowledge to build complex acoustic and language models to achieve high-precision recognition. The full phoneme algorithm model uses a simpler speech unit, namely a single phoneme, to avoid complex acoustic and language modeling, thereby greatly reducing the complexity and power consumption of the algorithm. Offline speech wake-up and recognition technology has been widely used in smart homes and smart speakers. The balance between AI speech...

  As consumers increasingly demand diversified and intelligent functions from smart wearable devices, the smart wearable industry needs to invest more in research and development and production costs to achieve product feature stacking or performance improvement. Constrained by the small form factor and cost control, the demand for high computing power and low power consumption, as well as the contradiction with the traditional von Neumann architecture, has become increasingly prominent. In this context, with the mass production of the world's first computing in-memory SoC chip WTM2101 in 2022, many manufacturers have begun to seek a breakthrough in this new type of computing architecture and have successfully developed and launched products, entering the lives of consumers. 1.Chip model: WTM2101 2.Chip type: ultra-low power AI SoC chip 3.Application: intelligent voice and intelligent health 4.Package: WLCSP (2.6x3.2mm²) 5.Power consumption: 5uA-3mA 6.AI computing power: 50Gops 7.M...

  Currently, most TWS devices use traditional dual or triple microphone ENC-based algorithms for voice enhancement and noise reduction. These algorithms can pick up sound directionally and reduce noise from other directions, but they also have significant drawbacks. Traditional noise reduction algorithms can cause some damage to human speech when directional pickup is achieved through phase manipulation, resulting in a muffled sound. They cannot effectively reduce noise from the same direction as the human voice, and they are not effective in high-noise environments, such as subways, trains, cafes, and roads. NN-based environment noise reduction algorithms are more versatile and can analyze collected audio directly using the characteristics of deep learning to separate human speech from noise, improve signal-to-noise ratio, reduce environmental noise, and preserve human speech. NN environment noise reduction can work alone with a single microphone, without requiring custom tuning a...