A couple of weeks ago we had the opportunity to interview Ken O’Neill, AMD space solutions engineer on a series of very interesting topics: The impact of technology on space explorationthe challenges it represents, the possibilities it offers and how specialized solutions are helping us address space missions with greater guarantees.
One of the most important recent missions, Nisar, has also become one of the most interesting for the technological world precisely because it uses specialized AMD solutions, and also for the possibilities and innovations that these solutions have brought to the world of space exploration.
The interview is about all this, and touches the most important points of this mission both at the level of objectives and challenges and at the level of hardware and technologies used. Before we start we want to thank Ken O’Neill for his time, and for the precision and clarity with which each of the questions answered.
Ken O’Neill, AMD Space Solutions Engineer. Image supplied by AMD.
(Mc) 1.- To put our readers in context, could you explain what Nisar is and why is it important?
(Ken) Nisar, or the NASA-Iro Syntheticaperture Radar mission, is a joint initiative of Terrestrial observation between NASA and the Indian Space Research Organization (ISRO). It is the first mission of its type to operate synthetic opening radar (SAR) of double band from space, using L and S frequencies, designed to allow An unprecedented image of the earth’s surface.
The mission will help scientists monitor and better understand natural disasterssuch as earthquakes, volcanoes and landslides, while following long -term processes such as ice layers, deforestation and reduction of groundwater. The observation aims to equip climate science, the response to disasters and agricultural monitoring with a level of precision and frequency of revision never before.
By capturing highly detailed and frequent radar images, Nisar will generate massive data volumes. This is where processing and computing on board become critical. The knowledge obtained from Nisar will inform numerous scientific research, as well as the formulation of policies and sustainability planning.
(MC) 2.- What is AMD’s role in the Nisar mission?
(Ken) AMD has provided technology for Adaptive High Reliability Computing for Nisar Radar Load. Our FPGAS radiation tolerant They will be in charge of critical signal processing tasks aboard the satellite. These devices support the compression, manipulation and optimization of radar data before transmission, essential functions considering the amount of data that are collected.
It is important to remember that Nisar collects data using two radar bands simultaneously, effectively duplicating the processing load. Historically, radar missions sent unprocessed data to Earth. But Nisar’s huge output volume requires on -board computing to filter, compress and prioritize data before the transmission.
Programmable logic (PL) devices, such as FPGAS and adaptive socs, allow real -time and high -performance processing while They operate with low energy consumptionwhich makes them ideal for environments that require high reliability systems such as space.
(MC) 3.- AMD has a wide portfolio of hardware solutions. What type of hardware did AMD choose for this mission (CPU, SOC, FPGAS, etc.) and why?
(Ken) For Nisar, the selected hardware was our FPGA AMD Virtex®-5QVa hardened device for radiation and spatial degree, chosen by its low latency, low consumption, radiation tolerance and long -term support.
The viTex-5QV offers a robust combination of programmable logic and high-speed interfaces in a compact format. It was designed to meet the strict demands of the orbital operation, where Reliability is not negotiable.
In more recent missions, we are seeing the adoption of more advanced space degree adaptive socs, such as the AMD XQR Versal™ AI Edgewhich integrates logic, signal processing (DSP), AI motors and CPU nuclei.
(MC) 4.- From the perspective of the hardware, what are the most important challenges to be overcome in this type of mission?
(Ken) The space is relentless. The hardware must be ionizing radiation resistantto wide temperature ranges and conditions without the possibility of repair. The main and most unique challenge of space is exposure to radiation and the need to mitigate both individual failures and the cumulative ionization that can degrade or interrupt electronic systems.
AMD space degree devices have hardened, advanced packaging and rigorous tests. Our adaptive socs are exposed to protons, heavy ions and gamma radiation during the tests. We also support fault tolerance at the system level through techniques such as Triple modular redundancy (TMR)which allows to continue operating even when a logical route fails.
Another challenge is longevity. Missions like Nisar They can operate for a decade or more. Therefore, we design and support components with long availability windows, consistent manufacture and reliable performance over time. In space, reliability is not a characteristic, it is a requirement.
(MC) 5.- Performance against reliability and stability: What matters more in Nisar and how does AMD add both aspects?
(Ken) In Nisar, as in any space mission, there is no commitment between performance and reliability: Both are essential. The radar payload requires an extremely high data flow and real -time processing capabilities. At the same time, any hardware deployed in space must meet the highest stability and resistance standards.
Adaptive socs and spatial degree fits combine the flexibility and processing capacity necessary for intensive tasks, such as SAR data manipulation, while offering hardened architecture, extensive radiation tests and Support for mitigation strategies of critical failures for the mission.
(Mc) 6.- What type of software integration was used in this mission?
(Ken) Nisar’s payload focuses on SAR, and AI plays a support role in optimize what data are transmitted. With a limited bandwidth for discharge, it is not practical to send all unprocessed data to the Earth, so the system must intelligently decide what is worth sending.
This is where PL devices contribute. They allow filtering, compression and signal processing on board, and the newest designs, such as adaptive socs AMD Versal ™ AI Edge, also They support inference of AI directly in orbit.
While nisar in itself Does not execute large -scale AI models on boardfuture missions using these devices can support applications such as detection of anomalies in telemetry or classification of images in real time.
(MC) 7.- What is the main objective of the Nisar mission?
(Ken) Nisar’s goal is to advance in the understanding of the changing surface of the earth. Using double -band SAR images, the mission will provide highly detailed, consistent and repeatable observations of the earth, ice and oceans over time.
These data are invaluable for climate science: They help track the movement of glaciers, measure soil moisture, monitor agricultural productivity, observe tectonic activity and detect natural disasters such as landslides or subsidencies.
Nisar’s precision and frequency make it a transforming tool for the science of the terrestrial system, providing useful information for scientists, governments and humanitarian agencies about how our planet is evolving.
(MC) 8.- How are AI and Hardware specialization impacting in this type of space mission?
(Ken) As the sensor resolution and data rates increase, It is no longer viable to send everything to Earth. The missions now require more processing capacity on the edge, aboard the satellite itself. That is where the inference of AI will become more common, for use cases such as image classification, data filtering or real time monitoring of the spacecraft.
In addition, we are seeing Telemetry abnormalities detection Using recurring neural networks directly on devices suitable for orbit. This is the type of intelligent functionality on board that will be critical for the next generation of missions.
