Jump-start your machine learning development with Seeed Studio’s new AI platform. The J1020 is a standalone Linux box that integrates a 128-core GPU with extremely low power consumption. It outperforms a general purpose CPU for tasks that require learning new data sets and detecting or reporting features or outliers of a data stream.
What’s even more impressive is that the platform supports a plethora of communication interfaces that are the bread and butter of the embedded world. You get a number of SPI, I2C, UARTs, and I2S buses, as well as high-speed camera interfaces that support 4K. Together with a powerful Quad-core ARM A57 processor, 4 GB RAM and 16 GB eMMC, it is a great proposition for embedded projects that require a touch of AI or ML.
Seeed Studio J1020: Price and Availability
The Seeed Studio reComputer J1020 is available in a variety of flavors ranging from entry-level with 4GB and SD card to high-end with 16GB of RAM and non-volatile eMMC memory. The processor is a Quad-core ARM with an NVIDIA Maxwell GPU for the low-tier systems and a Hexa-core ARM with an NVIDIA Xavier GPU for the more expensive models. The price varies from $199 (opens in new tab) until $699 (opens in new tab)†
Seeed Studio J1020: Design
The Seeed Studio J1020 computer comes in a small box containing a universal 12V 24W power adapter. The heat sink, Jetson Nano module and carrier board are all mounted in a black aluminum housing. The device measures 13cm x 12cm x 5cm and weighs 405g. The top acrylic cover is secured with four magnets and can be easily removed by pressing a release pin on the bottom.
With rubber pads on the bottom and left side, the case can be placed horizontally or vertically. There are two display outputs, an HDMI 2.0 and a Display Port 1.4, and four USB 3.0 host ports. A USB 2.0 Micro-B connector and an Ethernet LAN connector complete the list of user-accessible interfaces.
Internally, the Jetson Nano module is secured to a carrier board with two screws and has a heat sink covering the main chip. The carrier board maps all of the Nano’s interfaces to external connectors, but also adds two CSI camera interfaces, a 40-pin header that supports up to 26 GPIOs, and a fan connector.
The Jetson Nano is an entry-level System-On-Module (SOM) with a built-in GPU that supports 128 Maxell CUDA cores. A multi-function Quad-core 1.5GHz ARM processor is paired with 4GB LPDDR4 and 16GB eMMC 5.1 flash memory. The display output supports up to two 4K displays.
The system is compatible with other development boards via 3.3V GPIO headers while supporting electrical interfaces such as SPI, I2C, I2S, PWM and UART. This makes the J1020 an extremely versatile platform to jump-start any AI experiment with interfaces to the outside world.
Seeed Studio J1020: Features and Use
Applying flow will instantly boot an Ubuntu image from the eMMC storage. Since the J1020 comes without Wi-Fi, the only way to update the OS or access it remotely is through the Ethernet port. There is no LED to indicate that the system is active and there is no push button to start or turn off the device.
A USB hub is used to add more ports to the design. This reduces USB throughput when multiple peripherals are transferring data at the same time. In addition, the four USB ports also share a single power protection. This will cut power to all four ports if one fails. If the GPIO header is connected directly to the CPU, accidentally applying a voltage outside of the CPU specifications may cause destruction.
The main target audience for the J1020 platform is AI and Machine Learning (ML) enthusiasts. Software development starts with signing up and downloading the x86-only SDK from the NVIDIA website. Applications written with C++ or Python are then compiled and deployed to the board via the USB cable.
Geekbench 5 for ARM was used to benchmark the J1020 and resulted in a respectable single-core score of 237 and 868 for multi-core. The Machine Learning benchmark used in the synthetic test was run only on the ARM CPU and not on the Maxwell GPU. During the benchmark, the J1020 used about 1W at rest and 15W during the multi-core tests.
Seeed Studio J1020: Competition
The abundance of AI and ML development kits means it’s really easy to start experimenting on a budget. Big names like Google and Microsoft have invested heavily in cloud solutions like Microsoft’s Azure AI or hardware platforms like Google’s Coral.
Google’s Coral Development board is very similar to the J1020. For just under $150, we get a board that, while cheaper, has only 1GB of RAM, 8GB of eMMC, and no case. The Coral has built-in WiFi and Bluetooth, but has fewer USB and display outputs. The technology behind the TPU is different from that of the GPU, and while the Coral is extremely fast at making conclusions, it can’t build a new model. So it is less autonomous and needs an uplink when deployed in the field.
Seeed Studio J1020: Final Verdict
The J1020 is a great board for diving deep into Machine Learning and AI experiments. It has a capable GPU with 128 CUDA cores, although these are based on the dated Maxwell architecture. The price remains higher than the competition as this is an entry-level platform, but you get a system that can run as a standalone PC. We were impressed with the idle power dissipation of 1W. The device also boots up quickly thanks to the use of eMMC, while the number of camera inputs is also much more than what the competition gives.
Passive cooling is a double-edged sword as it makes the system quiet while limiting performance with thermal throttling. In an embedded world, a passive solution is a plus as there is no mechanical point of failure.
We’ve featured the best servers for small businesses (opens in new tab)†