PR Center

News Center VisioNexT's 3 core technologies: ③ Low Power 2023-06-01 627 hits

Low Power
Introducing VsioNexT's 3 Core Technologies





Low power consumption is an important consideration in the design of computer vision system-on-chip (SoC) or processors for edge devices. These devices often operate on battery power and need to perform complex image processing tasks in real time while consuming as little energy as possible. By minimizing power consumption, the battery life of the device can be extended, allowing it to operate for longer periods without recharging. This is particularly important for applications such as drones and wearable devices where long battery life is essential. Additionally, low power consumption also helps reduce heat generation, which improves the reliability and longevity of the device.

At VisioNexT, we address these needs through our Low Power technology, which incorporates Dynamic Voltage Frequency Scaling (DVFS). DVFS enables SoC to adjust voltage and frequency based on the processing load and computational requirements of the application running on it. This significantly reduces power consumption while optimizing performance. In addition, our strategy includes power gating, which activates only the necessary processors for each application’s operating policy or use case. This technique controls power on/off for blocks depending on each use case to reduce power leakage by separating power domains.









For example, in the case of security cameras, when objects such as people, vehicles, etc. appear, the NPU is maximized to activate AI features that detect and classify them, but when the objects are not on the screen, the NPU is minimized to consume power more efficiently. With DVFS and Power Gating, depending on the use case, from standby to a complex AI function, power consumption is expected to be reduced by up to 50%. Finally, it is important to consider performance trade-offs when designing power-optimized SoC. Therefore, we focus on the precise analysis of use cases provided by customers and apply them to the early design phase of Vision AI SoC for low power.



In conclusion, Low Power technology is crucial in the current industry, especially for edge devices. Our Low Power technology is a cutting-edge solution that optimizes the power consumption of IoT/embedded/edge devices by leveraging DVFS and using power gating based on precise use case analysis. This provides longer battery life and reduces heat dissipation. VisioNexT’s expertise in Low Power technology makes it an ideal partner for companies that require efficient IoT/embedded/edge devices and Vision AI solutions.

List