Artificial intelligence technology, as one of the most advanced productive forces in the “next few decades,” is gradually penetrating into various industries, driving the rapid development of the digital economy. It is expected that the compound annual growth rate of AI-related solution markets will remain at 65% over the next decade, reaching trillions by 2027. As an important new infrastructure in the digital economy era, the level of development of artificial intelligence infrastructure has become one of the important indicators for measuring a country’s scientific and technological innovation capabilities.

Using open-source hardware products as a foundation has become an important technological trend globally. Hardware openness usually refers to the free publication of schematics, bill of materials, and related source code files, allowing developers and companies of varying levels to design, manufacture, and produce devices based on the schematics and bill of materials. Open-source hardware mainly has the following characteristics:

Modular hardware design and open technical documentation greatly reduce the entry barrier for hardware. For software developers or students lacking hardware knowledge, it is also easier to quickly utilize the underlying hardware functions for solution development.

Open-source hardware typically uses standardized modules and interfaces, including the design of peripheral devices and driver software, which also follow corresponding standards, making hardware-level customization and expansion more flexible and effectively adapting to specific application scenario needs.

The open technical design documents and bill of materials can effectively reduce hardware design and manufacturing costs, thereby further lowering solution development costs and the overall solution’s TCO (Total Cost of Ownership).

The open-source hardware community provides rich technical support, including documentation, discussion forums, and Q&A communities, which can further lower the development threshold and continuously iterate and optimize the functionality and reliability of open-source hardware.

Traditional CPU architectures are not suitable for the high-density parallel computation required by deep learning model inference workloads. To achieve lower latency inference, it often requires adding additional acceleration cards (such as dedicated GPUs) to the system for deep learning inference. However, at the same time, additional acceleration cards will lead to increased costs and energy consumption.

Traditional open-source hardware usually adopts specific embedded operating systems and programming languages, requiring users to develop and compile on personal computers before transferring to the open-source hardware for deployment. For non-embedded developers or beginner students, there is a demand for more user-friendly operating systems, interactive interfaces, development environments, and deployment testing processes.

Aiming at the globalization trend of open-source hardware and artificial intelligence technology, DFRobot has launched the LattePanda 3 Delta based on Intel® processors. LattePanda 3 Delta is a single-board computer that supports a complete Windows operating system. While maintaining the usual pocket size of LattePanda, its thickness is only 16mm, and its CPU performance is improved by 2 times compared to the previous generation, with GPU performance increased by 3 times. Thanks to such powerful performance, LattePanda 3 Delta can be used to easily build edge computing, image recognition, and other projects. After nearly 8 years of accumulation and development, LattePanda has been widely used by global makers and developers and has also penetrated into more and more industrial fields.

🔵 Equipped with the next-generation Intel® processor, leveraging advanced processes and architectures, this processor can provide excellent performance at low power consumption. While meeting computing power requirements, it supports compact hardware design and longer battery life, suitable for integration into various embedded systems.

🔵 Using integrated graphics to accelerate AI inference, significantly improving inference speed compared to CPUs, without the need for additional GPUs or NPUs. This supports AI loads at low cost, low power consumption, and small size. Combined with the OpenVINO™ toolkit, developers can quickly deploy various AI inferences on the LattePanda 3 Delta development board and achieve optimized performance.

🔵 Onboard ATmega32U4 coprocessor, providing rich GPIO interfaces. Combined with the rich peripheral modules and development libraries provided by DFRobot for the LattePanda product line, developers can easily build various systems and call control interfaces.

🔵 Running a complete version of the Windows system supports a wider software ecosystem. Software that can run on ordinary computers can also run smoothly on the LattePanda 3 Delta motherboard, making debugging, porting, and deploying software more convenient.

🔵 Providing mature documentation, toolchains, and examples, making the development process simpler.

This year, autonomous driving cars have become a research hotspot for the application of artificial intelligence technology. Through cameras and various sensors providing information and deep learning-based computer vision algorithms, autonomous driving systems can obtain information about various obstacles, traffic signs, and drivable areas that appear in the car’s field of vision.

Based on the Donkey Car open-source project, the Mushroom Cloud Maker Space has incubated and developed the Drifting Donkey community in China, focusing on AI education and sports platforms, serving AI geeks and RC car enthusiasts. Based on the LattePanda 3 Delta as the core single-board computer, the Drifting Donkey community has launched the latest generation DonkeyDrift X2. Benefiting from the significantly improved CPU performance and AI inference capability of the N5105, this DonkeyDrift X2 can achieve rapid recognition and response without the design of additional acceleration cards or AI modules, greatly enhancing the stability and smoothness of autonomous driving.

Currently, with the popularization of live broadcasting technology, the demand for live broadcasting equipment is also increasing. Ordinary computers or Mini PCs are relatively large, inconvenient to move or adjust positions, and are not suitable for flexible deployment in live broadcast rooms. The LattePanda 3 Delta, with its compact size and powerful performance, can effectively help locate all equipment on-site, measure distances, and obtain camera and lens source data, achieving excellent and smooth live broadcast effects that can meet the needs of the live broadcast industry. Compared to ordinary computers and general single-board computers, the LP development board has the following advantages:

💠 Strong performance: Equipped with the Intel® Celeron® N5105 processor, it can ensure high synchronization of data, achieving consistent depth of field, perspective, and exposure, improving live broadcast quality. The integrated graphics performance is strong, supporting the encoding and output of 2 channels of 4K 60Hz video, achieving higher quality live broadcasts.

💠 Good compatibility: Supports running a complete Windows system, compatible with most media playback and display software, facilitating the use and connection of live broadcast equipment.

💠 Rich interfaces: Supports connecting various devices and sensors, can directly obtain lens and camera source data from the machine, plug and play, very convenient. It reserves a slot for 4G/5G module cards, eliminating the reliance on Wi-Fi networks, and device location is no longer limited.

💠 Compact and portable: Its size is only that of a pocket, making it easy to place on photographic equipment. Simply adding a support and a shell makes it easy to move.

In the context of the Internet of Everything, smart terminal devices are important carriers of IoT and AI. As an important tool for human-computer interaction, they are increasingly applied in industries such as office, medical, and self-service, such as facial recognition visitor machines, medical carts, and self-service payment machines. Such integrated devices often have complex functions and require core development boards with good performance in various aspects and large displays that are easy to operate to present a good human-computer interaction experience.

LattePanda 3 Delta’s excellent performance and compact size make it have the following advantages as a development board for such smart terminal devices:

🔀 High performance: LattePanda 3 Delta is equipped with a powerful Intel Celeron N5105 processor and 8GB RAM, providing high-speed computing power and smooth running experience, enhancing user experience when operating smart terminal devices.

🔀 Multifunctional: LattePanda 3 Delta supports various peripherals, such as touch screens, cameras, solid-state drives, etc., which can meet the functional requirements of smart terminal devices in various human-computer interaction scenarios.

🔀 Easy to develop: Since LattePanda 3 Delta is based on open-source hardware and software, supporting Windows and Linux operating systems, device manufacturers can customize and modify according to project requirements, making secondary development easy.

⏩ Good display effect: 1920×1080 high resolution, 15.6-inch large size, has a sufficiently large display area and high-definition display effect, meeting the scene requirements of smart terminal devices.

⏩ Simple connection: The Type-C touch screen only needs one cable to connect to the Type-C interface of the LattePanda 3 Delta, supporting the transmission of audio, video, and touch data simultaneously, while providing reverse power to the LattePanda 3 Delta. This brings convenience to the installation and maintenance of smart terminal devices.

Click “Read the original text” to download the white paper “DFRobot Launches Development Board for Edge Computing and AI Based on Intel® Processors”