Think Silicon is a prominent provider of ultra-low power GPU intellectual property (IP) designed for embedded systems, boasting over 15 years of expertise in creating highly efficient, small-scale 2D and 3D graphics, video, display, and machine learning-accelerated microprocessor cores. These innovations enable their clients to achieve low system power consumption, cost savings, and enhanced performance. Catering to a variety of ultra-low power and area-restricted devices such as wearables, home control appliances, and entertainment systems, Think Silicon's technology is utilized by leading fabless semiconductor firms, cloud service providers (CSPs), and design houses globally. Their focus on delivering high-quality, battery-friendly user interface graphics solutions allows for sophisticated graphical user interfaces and gaming-class graphics speed in cost-sensitive products.

‍

Expanding its reach into the artificial intelligence (AI) domain, Think Silicon leverages its extensive experience in ultra-low power GPU IP to offer cutting-edge capabilities similar to its achievements in the graphics sector. Their configurable GPUs are adaptable for a broad spectrum of applications, from smart IoT products like wearables and health trackers to video overlaying in data centers, emphasizing AI functionality for battery-powered edge devices. Think Silicon's approach covers the entire SoC (System on Chip) design process, from architecture planning to tape-out, supported by a skilled cross-disciplinary team. Their technology's reliability is demonstrated through rigorous validation, simulation, and code coverage measurements, accompanied by comprehensive support tools including a Software Development Kit (SDK) to facilitate software application development, ensuring devices capitalize on the GPUs' high performance and energy efficiency.

‍

A collaboration between Think Silicon and Edge Impulse will provide users with notable advantages at the intersection of the enablement of the development and deployment of machine learning models on edge devices, simplifying the process of bringing AI to a wide array of applications. Several advantages come to mind, notably:

‍

1. Enhanced Machine Learning Performance on Low-Power Devices: By combining Think Silicon's expertise in ultra-low power, high-performance computing with Edge Impulse's machine learning model optimization, users could develop and deploy more efficient AI algorithms. This means creating smarter edge devices capable of complex processing with minimal energy consumption.

‍

2. Streamlined Development Process: Developers will benefit from a more integrated development environment, combining Edge Impulse's machine learning development tools with Think Silicon's GPU and microprocessor technologies. This would simplify the process of designing, training, and deploying AI models on embedded systems, making it more accessible for developers without deep expertise in either machine learning or hardware design.

‍

3. Optimized Graphical and AI Capabilities: This partnership would lead to the optimization of graphical user interfaces (GUIs) alongside AI functionalities, enabling devices to not only perform complex AI tasks efficiently but also to enhance user interaction through rich, responsive GUIs on the same low-power platform.

‍

Despite their small size and low power consumption, Think Silicon's GPUs offer high performance for 2D and 3D graphics processing. This is made possible through architectural innovations that optimize data flow and processing efficiency, as well as specialized hardware blocks designed for specific graphics operations.

‍

In addition to graphics processing, Think Silicon's IP cores are optimized for accelerating machine learning algorithms. This involves specialized hardware accelerators that can efficiently perform operations common in machine learning tasks, such as matrix multiplications and convolutions, thereby speeding up inference on edge devices.

‍

The NEOX product line is designed to address the increasing demand for advanced graphical capabilities and AI functionalities in edge devices, which include IoT devices, wearables, automotive systems, and smart home appliances.

‍

Think Silicon plans to enter the AI market in the consumer health wearables vertical. The experience Edge Impulse has collected working with Oura and NoWatch along with our internally developed HRV processing capabilities, will give Think Silicon an edge when they enter the market.

‍