Summary of PUMPKIN I500 SBC USES MEDIATEK I500 SOC FOR EDGE AI APPLICATIONS
The OLogic Pumpkin i500 is a Single Board Computer designed for AI edge computing and computer vision. Powered by the MediaTek i500 SoC, it features an octa-core CPU, Mali-G72 GPU, and dedicated AI accelerators. The board supports dual-band Wi-Fi, Bluetooth 5, various camera interfaces, and multiple audio inputs/outputs. It runs Android 10 or Yocto Linux and is certified for AWS IoT Greengrass and Microsoft Azure IoT Edge, making it ideal for smart hubs, robots, and industrial applications.
Parts used in the Pumpkin i500:
- MediaTek i500 SoC
- Quad-core Arm Cortex-A73 CPU
- Quad-core Cortex-A53 CPU
- Arm Mali-G72 MP3 GPU
- Dual-core Tensilica Vision P6 DSP
- LPDDR4 RAM
- 16GB eMMC flash
- MIPI DSI connector
- MIPI CSI connector
- 3.5mm stereo headset jack
- Digital PDM microphone connectors
- RJ45 Ethernet port
- Wi-Fi 5 module
- Bluetooth 5 module
- USB-C OTG port
- USB-A host ports
- Micro USB debugging port
- 40-pin GPIO connector
- Secure Boot hardware
- Trust Zone security module
OLogic has released more details on their Pumpkin i500 Single Board Computer built to support computer vision and AI edge computing. The hardware development kit was first announced at the beginning of last year and initially scheduled to launch around that same time but was delayed due to some reasons.
The Pumpkin i500 Evaluation Kit is powered by a powerful MediaTek i500 SoC combined with Quad-core Arm Cortex-A73 and Quad-core Cortex-A53 power-efficient 64-bit CPU solutions together with an Arm Mali-G72 MP3 GPU. The board also features a rich set of GPIO for expandability and a dual-Band Wi-Fi (2.4GHz/ 5GHz) and Bluetooth 5 that make it suitable for indoor/outdoor home and industrial applications.
Features and Specifications of the Pumpkin i500 SBC include:
- CPU: MediaTek i500 octa-core processor with quad-core Arm Cortex-A73 running at up to 2.0 GHz and quad-core Cortex-A53 cores at 2.0 GHz
- GPU: Arm Mali-G72 MP3
- Dual-core Tensilica Vision P6 DSP/AI accelerator @ 525 MHz
- Upgradable 2GB LPDDR4 RAM
- 16GB eMMC flash
- 4-lane MIPI DSI connector
- Up to 25MP camera support via MIPI CSI connector
- 3.5mm stereo headset jack
- 2x Digital PDM microphone connectors
- 8-ch audio input / 8-ch audio output
- 1080p60 H.264 video decoder
- 1080p30 H.264 encoder
- 10/100M Ethernet (RJ45)
- Wi-Fi 5 (802.11ac) 2×2 MIMO dual-band (2.4GHz, 5GHz) and Bluetooth 5 for wireless connectivity
- 1x USB-C OTG port
- 2x USB-A host ports
- Micro USB port for debugging
- 40-pin GPIO connector
- 5V DC via one USB-C port, and,
- Support for TensorFlow Lite and Object detection (MobileNet + SSD)
- Security Features: Secure Boot, Trust Zone, TF-A, OP-TEE
- Operating System: Yocto Linux and Android 10
The Pumpkin i500 SBC enables the development of AI vision products like control panels, smart appliances, points-of-sales, smart hubs, robots etc. Users can also easily deploy their workloads to the Edge through standard containers because the board is qualified for AWS IoT Greengrass and certified for Microsoft Azure IoT Edge.
Read more: PUMPKIN I500 SBC USES MEDIATEK I500 SOC FOR EDGE AI APPLICATIONS
- What processor powers the Pumpkin i500 SBC?
The board is powered by the MediaTek i500 octa-core processor featuring quad-core Arm Cortex-A73 and quad-core Cortex-A53 cores. - Can the Pumpkin i500 support high-resolution cameras?
Yes, the board supports up to 25MP cameras via the MIPI CSI connector. - Does the Pumpkin i500 include wireless connectivity options?
Yes, it features Dual-Band Wi-Fi 5 at 2.4GHz and 5GHz along with Bluetooth 5. - Which operating systems are supported on this board?
The device supports both Yocto Linux and Android 10. - Is the Pumpkin i500 compatible with cloud edge platforms?
Yes, it is qualified for AWS IoT Greengrass and certified for Microsoft Azure IoT Edge. - What video encoding capabilities does the board offer?
It includes a 1080p30 H.264 encoder and a 1080p60 H.264 video decoder. - How much storage and memory does the evaluation kit provide?
It comes with upgradable 2GB LPDDR4 RAM and 16GB eMMC flash storage. - Can I use TensorFlow Lite on this single board computer?
Yes, the board supports TensorFlow Lite and Object detection using MobileNet plus SSD.
