EO·Atlas
Preview build / EO·Atlas v0.9, content still landing
analysis-service ยท products

Antelope DPU

Compiled from public sources on 2026-06-10. Not independently verified by KP Labs.

Does KP Labs own this listing? Claim and verify it →

KP Labs Antelope data processing unit for onboard satellite AI and data handling.

The Antelope DPU is an onboard data processing unit developed by KP Labs for CubeSat-class satellites, built around a Xilinx Zynq UltraScale+ MPSoC (quad ARM Cortex-A53 up to 1.5 GHz, dual Cortex-R5, integrated FPGA) delivering 160 GOPS of compute. The unit ships in a PC/104 form factor with 8 GB DDR4 ECC memory, 4 GB SLC NAND Flash, and an optional SATA SSD, and carries a broad interface set including CAN, I2C, SPI, RS422/485, UART, LVDS/GTH transceivers, USB 3.0, USB 2.0, 1 GbE Ethernet, and GPS PPS. Input voltage ranges from 6.5 to 14 V; per-mode power consumption is not published on the product datasheet.[1]

The unit runs 64-bit Linux or bare-metal operating environments and is compatible with KP Labs Leopard and Lion DPU product lines as well as the Simera Sense HyperScape100 hyperspectral camera.[1]

The KP Labs product page and the IOD mission page use different designations for the same hardware: the product page uses Antelope DPU while the mission page uses Antelope OBC.[1][2]

Flight heritage was established on the Antelope In-Orbit Demonstration, flown on a D-Orbit ION Satellite Carrier launched on SpaceX Transporter-9 in 2023, funded by the Polish National Centre for Research and Development (NCBR). During eight weeks of continuous LEO operation, the unit ran RandomForest and Telemanom algorithms for onboard telemetry anomaly detection and sustained uninterrupted ML inference.[2]

Sources
Cite https://eo-atlas.org/products/analysis-service/antelope-dpu Markdown twin → Field definitions →