Wang Shaodi
CEO
WT
Witmem Technology
This company develops in-memory computing chips that enable efficient deep learning operations for AIoT applications like wearables and smart devices. Their chips offer high performance at microwatt to milliwatt power levels, reducing computing costs for enterprises.
- Artificial Intelligence
- Hardware
- Semiconductor
Updated 6 months ago
Funding
Funding rounds are not available yet.
Founders
Product
Problem
Traditional computing architectures require frequent data movement between memory and processing units, leading to significant energy consumption and limiting the performance of AI applications, especially in edge devices with constrained resources. This bottleneck hinders the deployment of advanced AI functionalities in applications like wearables, mobile devices, and IoT devices.
Solution
知存科技 (Witmem) addresses the limitations of traditional computing by developing in-memory computing (CIM) chips that physically integrate memory and processing, significantly reducing data movement and improving energy efficiency. Their CIM technology enables neural network acceleration with significantly reduced power consumption, allowing for the deployment of AI at the edge. The company's chips achieve high performance at microwatt to milliwatt power levels, making them suitable for battery-powered and small form-factor devices. By minimizing data transfer, Witmem's CIM chips enhance computational efficiency and enable more complex AI tasks to be performed locally on edge devices.
Target Audience
The primary target audience includes developers and manufacturers of AI-enabled devices, such as wearables, smart devices, IoT devices, and mobile devices, who require high-performance, low-power AI processing at the edge.
Features
- Proprietary in-memory computing architecture that reduces data movement and power consumption
- WTM-2 series chips optimized for high-efficiency edge AI applications in battery-powered devices
- WTM-8 series chips designed for low-power, high-performance computing vision in mobile devices
- witin_mapper compiler toolchain for mapping quantized neural networks to the WTM2101 MPU accelerator
- Complete software development kit (SDK) and integrated development environment (IDE) for chip development
- Development boards available for WTM2101 AI SOC testing and application development
- Support for various AI applications, including speech recognition, AI health monitoring, image processing, and spatial computing