The chip delivers  5.7 tera operations per second (TOPs) of peak INT8 performance optimized for deep learning, a nearly 25X increase compared to the prior generation.

The Kintex UltraScale XQRKU060 FPGA provides on-orbit reconfiguration with radiation tolerance across all orbits.

A portfolio of ML development tools supporting industry standard frameworks, including TensorFlow and PyTorch, enable neural network inference acceleration for real-time on-board processing in space with a complete “process and analyze” solution.

The on-orbit reconfiguration capabilities, together with real-time on-board processing and ML acceleration, allows satellites to update in real-time, deliver video-on-demand, and perform compute “on-the-fly” to process complex algorithms.

The ML capabilities are suitable to a variety of problems spanning scientific analysis, object detection, and image classification – such as cloud detection – enabling improved processing efficiency and reduced decision latency both in space and on the ground.

As protocols and applications progressively change, the adaptive compute architecture of the XQRKU060 allows unlimited on-orbit reconfiguration to enable customers to perform last-minute product updates prior to launch, as well as after it has been deployed in orbit.

The device has  2,760 UltraScale DSP slices and provides up to 1.6 TeraMACs of signal processing compute, more than a 10X increase compared to the prior generation, as well as dramatic efficiency gains for floating point computations.

The increased compute capability in space is paired with massive I/O bandwidth from 32 high-speed transceivers (SerDes) that can run up to 12.5Gbps to deliver 400Gbps aggregate bandwidth.

THe chip has 40x40mm ceramic packaging capable of withstanding vibrations and handling during launch as well as radiation effects in harsh orbit environments.

The architecture features an innovative design for single event effects (SEE) mitigation thereby meeting the industry requirements for all orbits, including low earth orbit (LEO), medium earth orbit (MEO), geosynchronous orbit (GEO), and deep space missions.

 Flight units of the 20nm RT Kintex UltraScale space-grade XQRKU060-1CNA1509 FPGA will be available in Xilinx Class B and Class Y test flows per the MIL-PRF-38535, beginning in September of this year.

Mechanical samples and prototype units are available now.