A new ability to pinpoint sources of fast radio bursts | Hacker News Reader

A new ability to pinpoint sources of fast radio bursts

news.berkeley.edu

74 pointsgmays10 comments

CHIME is great because it is so cheap. It's parabolic cylinders lying on the ground pointing fixedly straight up. The construction and mantainence cost per meter^2 of effective aperture is the lowest it possibly can be. It's basically a handful of large "half pipe" looking things next to shipping containers packed full of GPUs and power equipment. The only downside to this parabolic cylinder design is reflections down the long axis of the cylinder but that can be mitigated.

It's a (nearly) filled aperture with lots of redundant baselines which allows for easier calibration and seeing through the galactic foreground. But being so (relatively) closely spaced it's resolution is limited. Adding outriggers fixes this for when they want to get a position and I'd bet they put one at angles to the main array cylinder direction.

There was a nice talk[1] about CHIME over at Perimiter which went into the hardware and software details (at around 21 minute mark).

[1]: https://pirsa.org/18040069

Would it be beneficial to anyone if this design was scaled up (more or longer half pipes) or many of these were created and dotted around the world?

It would, but a pilot program of (much smaller/cheaper (no gpus for imaging)) STARE2 deployments around the world for fast transient radio burst detection would probably be a better start. https://www.caltech.edu/about/news/magnificent-burst-within-... https://arxiv.org/abs/2001.05077

>"When I saw the data, I was basically paralyzed," says Christopher Bochenek (MS '18), lead author of a new Nature study on the STARE2 results, ... "At the radio frequencies we observe with STARE2, the signal was much stronger than what CHIME reported. We had caught the FRB head-on."

Surely 256 GPUs along with custom FPGA boards aren't cheap either? But I guess it's a relative cost thing.

I'm surprised those are right next to it, but then, I can imagine they need to do a lot of filtering because the firehose of data is too much to send over the internet unfiltered.

Certainly not at the time. Although buying the GPUs that CHIME uses now would be a lot cheaper. What I meant is that where cost could be minimized it was. Relative to other radio telescopes of it's effective aperture it was/is spectacularly cheap.

That's a coincidence, Australian astronomers published a paper on a new interferometry-based system for fast radio burst detection today too:

https://www.csiro.au/en/news/All/News/2025/January/Australia...

https://www.cambridge.org/core/journals/publications-of-the-...

I saw this, this morning and thought OP was linking to the same. Odd.

> emanating from somewhere in the northern constellation Ursa Minor.

I've heard about some great publishing corporations out of Ursa Minor. Guide research?

Reply on news.ycombinator.com