- Lead Time:
- 14 weeks (FM)
- Lead Time:
- 10 weeks (EM)
- Processor:
- Quad-Cortex-A53 (Gen 1)
- Processor:
- 8-core Arm Cortex-A78AE (Gen 2)
- GPU:
- 2D/3D Vivante GC7000 Lite GPU and VPU (Gen1)
- GPU:
- 1024-core NVIDIA Ampere (Gen 2)
- RAM:
- 4 GB LPDDR4 SDRAM (Gen 1)
- RAM:
- 16 GB 128-bit LPDDR5 (Gen 2)
- Mass:
- 1.4 kg (Gen 1)
- Mass:
- 1.5 kg (Gen 2)
- Size:
- 96 x 96 x 50 mm^3 (Gen 1)
- Size:
- 113 x 100 x 70 mm^3 (Gen 2)
- Supply Voltage:
- 5V
- Idle Power Consumption:
- 3 W (Gen 1)
- Idle Power Consumption:
- 9.3 W (Gen 2)
- Typical Power Consumption:
- 7.5 W (Gen 1)
- Typical Power Consumption:
- 20 W (Gen 2)
- Peak Power Consumption:
- 22 W (Gen 1)
- Peak Power Consumption:
- 30 W (Gen 2)
- Operating Temperature Range:
- -25C to 85C
- Storage Temperature Range:
- -40C to 85C
- Operating System:
- Linux Based
- Software:
- Basic software package included
FeatherEdge, a compact Data Processing Unit tailored for AI applications in orbit, redefines space efficiency. Its 0.5U size and low power design ensure seamless compatibility with diverse satellite platforms. By processing onboard sensor data directly and transmitting only crucial information, FeatherEdge slashes downlink costs and significantly bolsters response times for critical events in orbit.
At its core, FeatherEdge integrates a robust AI computing module with Sidus Space’s radiation tolerant motherboard, operating as a fully integrated Linux system. Capable of trillions of operations per second and boasting precision thermal management within its enclosure, it minimizes reliance on external thermal controls. This device combines cutting-edge computing prowess with space-grade reliability, delivering a complete AI payload for unparalleled on-orbit high-performance edge computing capabilities.
Specification | FeatherBox GEN 1 | FeatherBox GEN 2 |
CPU | Quad-Cortex-A53 | 8-core Arm Cortex-A78AE |
GPU | 2D/3D Vivante GC7000 Lite GPU and VPU | 1024-core NVIDIA Ampere |
AI Performance | Coprocessor: 4 TOPS | Coprocessor: 100 TOPS |
Microcontroller | PIC12 8-bit | Rad-tolerant ARM Cortex M7 32-bit |
RAM | 4 GB LPDDR4 SDRAM | 16 GB 128-bit LPDDR5 |
Storage | 40 GB SLCNAND Flash(ECC); 16 GB eMMC | 680 GB pSLC NVMe SSD (ECC) |
Operating System | Linux Based | Linux Based |
AI/ML Model Type | TensorFlow Lite | TensorRT |
Connectors | Micro-D | Nano-D, M12, Micro-B, USB |
Ethernet | 1x GbE | 3x GbE |
USB | 1x USB 3.1 Gen 1 (5 Gbps) | 4x USB 3.2 Gen 1 (5 Gbps), 2x USB 2.0 |
UART | 2x RS-422 | 2x TTL, 2x RS-232, 1x RS-485 |
Other I/O | 2x SPI, 2x I2C, 8x GPIO | 2x SPI, 2x I2C, 1x CAN, GPIOs |
Power Consumption | 3W idle, 7.5W typical, 22 W peak (tens of microseconds) | 9.3 W idle, 20 W typical, 30 W peak |
FeatherEdge
Sidus SpaceFeatherEdge, a compact Data Processing Unit tailored for AI applications in orbit, redefines space efficiency. Its 0.5U size and low power design ensure seamless compatibility with diverse satellite platforms. By processing onboard sensor data directly and transmitting only crucial information, FeatherEdge slashes downlink costs and significantly bolsters response times for critical events in orbit.
At its core, FeatherEdge integrates a robust AI computing module with Sidus Space’s radiation tolerant motherboard, operating as a fully integrated Linux system. Capable of trillions of operations per second and boasting precision thermal management within its enclosure, it minimizes reliance on external thermal controls. This device combines cutting-edge computing prowess with space-grade reliability, delivering a complete AI payload for unparalleled on-orbit high-performance edge computing capabilities.
Specification | FeatherBox GEN 1 | FeatherBox GEN 2 |
CPU | Quad-Cortex-A53 | 8-core Arm Cortex-A78AE |
GPU | 2D/3D Vivante GC7000 Lite GPU and VPU | 1024-core NVIDIA Ampere |
AI Performance | Coprocessor: 4 TOPS | Coprocessor: 100 TOPS |
Microcontroller | PIC12 8-bit | Rad-tolerant ARM Cortex M7 32-bit |
RAM | 4 GB LPDDR4 SDRAM | 16 GB 128-bit LPDDR5 |
Storage | 40 GB SLCNAND Flash(ECC); 16 GB eMMC | 680 GB pSLC NVMe SSD (ECC) |
Operating System | Linux Based | Linux Based |
AI/ML Model Type | TensorFlow Lite | TensorRT |
Connectors | Micro-D | Nano-D, M12, Micro-B, USB |
Ethernet | 1x GbE | 3x GbE |
USB | 1x USB 3.1 Gen 1 (5 Gbps) | 4x USB 3.2 Gen 1 (5 Gbps), 2x USB 2.0 |
UART | 2x RS-422 | 2x TTL, 2x RS-232, 1x RS-485 |
Other I/O | 2x SPI, 2x I2C, 8x GPIO | 2x SPI, 2x I2C, 1x CAN, GPIOs |
Power Consumption | 3W idle, 7.5W typical, 22 W peak (tens of microseconds) | 9.3 W idle, 20 W typical, 30 W peak |
- Lead Time:
- 14 weeks (FM)
- Lead Time:
- 10 weeks (EM)
- Processor:
- Quad-Cortex-A53 (Gen 1)
- Processor:
- 8-core Arm Cortex-A78AE (Gen 2)
- GPU:
- 2D/3D Vivante GC7000 Lite GPU and VPU (Gen1)
- GPU:
- 1024-core NVIDIA Ampere (Gen 2)
- RAM:
- 4 GB LPDDR4 SDRAM (Gen 1)
- RAM:
- 16 GB 128-bit LPDDR5 (Gen 2)
- Mass:
- 1.4 kg (Gen 1)
- Mass:
- 1.5 kg (Gen 2)
- Size:
- 96 x 96 x 50 mm^3 (Gen 1)
- Size:
- 113 x 100 x 70 mm^3 (Gen 2)
- Supply Voltage:
- 5V
- Idle Power Consumption:
- 3 W (Gen 1)
- Idle Power Consumption:
- 9.3 W (Gen 2)
- Typical Power Consumption:
- 7.5 W (Gen 1)
- Typical Power Consumption:
- 20 W (Gen 2)
- Peak Power Consumption:
- 22 W (Gen 1)
- Peak Power Consumption:
- 30 W (Gen 2)
- Operating Temperature Range:
- -25C to 85C
- Storage Temperature Range:
- -40C to 85C
- Operating System:
- Linux Based
- Software:
- Basic software package included
FeatherEdge, a compact Data Processing Unit tailored for AI applications in orbit, redefines space efficiency. Its 0.5U size and low power design ensure seamless compatibility with diverse satellite platforms. By processing onboard sensor data directly and transmitting only crucial information, FeatherEdge slashes downlink costs and significantly bolsters response times for critical events in orbit.
At its core, FeatherEdge integrates a robust AI computing module with Sidus Space’s radiation tolerant motherboard, operating as a fully integrated Linux system. Capable of trillions of operations per second and boasting precision thermal management within its enclosure, it minimizes reliance on external thermal controls. This device combines cutting-edge computing prowess with space-grade reliability, delivering a complete AI payload for unparalleled on-orbit high-performance edge computing capabilities.
Specification | FeatherBox GEN 1 | FeatherBox GEN 2 |
CPU | Quad-Cortex-A53 | 8-core Arm Cortex-A78AE |
GPU | 2D/3D Vivante GC7000 Lite GPU and VPU | 1024-core NVIDIA Ampere |
AI Performance | Coprocessor: 4 TOPS | Coprocessor: 100 TOPS |
Microcontroller | PIC12 8-bit | Rad-tolerant ARM Cortex M7 32-bit |
RAM | 4 GB LPDDR4 SDRAM | 16 GB 128-bit LPDDR5 |
Storage | 40 GB SLCNAND Flash(ECC); 16 GB eMMC | 680 GB pSLC NVMe SSD (ECC) |
Operating System | Linux Based | Linux Based |
AI/ML Model Type | TensorFlow Lite | TensorRT |
Connectors | Micro-D | Nano-D, M12, Micro-B, USB |
Ethernet | 1x GbE | 3x GbE |
USB | 1x USB 3.1 Gen 1 (5 Gbps) | 4x USB 3.2 Gen 1 (5 Gbps), 2x USB 2.0 |
UART | 2x RS-422 | 2x TTL, 2x RS-232, 1x RS-485 |
Other I/O | 2x SPI, 2x I2C, 8x GPIO | 2x SPI, 2x I2C, 1x CAN, GPIOs |
Power Consumption | 3W idle, 7.5W typical, 22 W peak (tens of microseconds) | 9.3 W idle, 20 W typical, 30 W peak |