A researcher from the SETI Institute described to a packed audience at GTC 2025 how SETI had successfully trialed a novel method to identify interstellar radio waves which, theoretically, can also be used to identify communication from intelligent extraterrestrial life.
Luigi Cruz, a staff engineer at SETI, the world’s foremost organization looking for signs of intelligent life on other planets, described how his team used NVIDIA Holoscan and AI to accurately identify radio signals emitted by a far-off pulsar.
Using 42 different but synchronized antennas located at the Allen Telescope Array in Hat Creek, California, the SETI team identified radio signals emitted by a pulsar nestled in the Crab Nebula—which lies about 6,500 light years from Earth.
“The universe is very large and mostly empty,” Cruz said. “We need superhuman means—which is what AI basically is—to search the data in creative ways.”
The SETI researchers used NVIDIA Holoscan and a bespoke neural network to continuously process in real-time gigabits worth of radio data the pulsar emitted.
The results proved the efficacy of a method the SETI team has worked on for more than a year—one which uses AI and accelerated computing to identify patterns in various kinds of interstellar signals.
In his talk, Cruz described how NVIDIA GPUs have dramatically streamlined how researchers look for interstellar signals.?
Traditionally, astrophysicists used telescopes to scan the universe and then collect and save that data to hard drives. Only later, when time and manpower permitted, would researchers go back to look for potential signs of intelligent communication in the cached data.
Today, the SETI team uses Holoscan and the NVIDIA IGX Orin with RTX A6000 GPUs to run real-time inference on radio signals collected from space.
Rather than having to wait and sift through massive amounts of data looking for potentially coherent patterns, the AI screens the raw-sensor data in real-time, discarding most of it as irrelevant but also pulling out data that could be evidence of interstellar communication.
Before last month when the SETI team turned their telescopes towards the Crab Nebula, it didn’t know if its AI could reliably discern radio waves from deep space and identify patterns. What they were looking for, in essence, was a proof of concept that its AI could find patterns in interstellar data—in this case, radio waves from a well known pulsar.?
The AI passed the test with flying colors. It identified the pulsar’s giant radio pulses, reconstructing the data exactly as it was supposed to.
Now that he knows the SETI model can discern interstellar radio waves, Cruz and his colleagues hope the AI can identify other patterns from signals emanating from deep space—potentially even ones originating from intelligent extraterrestrial life.
