February 23, 2026

Aliens, nostalgia, and 12B blips

SETI@home: Data Acquisition and Front-End Processing

Arecibo is gone, but fans want their PCs hunting E.T. again

TLDR: SETI@home’s new paper explains how millions of home hours scanned radio data and flagged 12 billion possible signals, much of it from the now‑collapsed Arecibo telescope. Comments swing between nostalgic memories and a debate over reviving the project with smarter tools, making citizen science feel urgent again.

SETI@home just dropped a deep‑dive on the “front end” of its alien‑listening project, and the comments exploded into nostalgia and comeback talk. For newbies: SETI@home let regular folks lend home PCs to sift radio signals for signs of extraterrestrial tech. The paper says data from Arecibo and others flowed to 100,000+ computers, powering huge number‑crunching to flag about 12 billion possible blips, then filtering out human‑made noise. Read the study: doi.org/10.3847/1538-3881/ade5a7.

Strongest vibe? Bittersweet awe. bpoyner points out Arecibo literally collapsed in 2020—end of an era—yet there’s still a mountain of data to chew. elicash tosses a hot question: could a reboot with smarter “distributed agents” bring SETI@home back and fix why researchers paused it? The thread splits fast: dreamers chant “revive it,” realists reply “it was killed for a reason.”

Then the jokes roll. alex_suzuki: “Immediate nostalgia activated,” remembering a Pentium whirring for E.T.; others flash retro rigs and quip their 90s PCs were ghosted by aliens. Meme pitch of the day: “GPU miners by day, ET hunters by night.” Mood check: hopeful chaos, grieving Arecibo while daring the internet to listen again.

Key Points

  • SETI@home distributed time-domain radio data to over 100,000 volunteer computers from 1998–2020, providing ~10^15 FPOP s−1.
  • Front-end processing uses coherent integration to confine drifting signal power to single DFT bins.
  • The coherent search spans 123,000 Doppler drift rates within ±100 Hz s−1.
  • Analyses cover multiple signal types using various DFT sizes with frequency resolutions from 0.075 to 1221 Hz.
  • About 1.2 × 10^10 detections above power and fit thresholds were produced and passed to a backend that removes RFI and seeks persistent, ETI-consistent groups.

Hottest takes

"Arecibo collapsed in December of 2020. It sounds like they have a lot of data to still churn through" — bpoyner
"I do wonder if there's a way to bring it back with distributed agents" — elicash
"Immediate nostalgia activated. I ran this on a Pentium" — alex_suzuki
Made with <3 by @siedrix and @shesho from CDMX. Powered by Forge&Hive.