redback

Thanks for pointing this out. I'd missed out on this discovery. But then again I missed out on alot of stuff^* in 2020 ....

"This 5D parameter volume is extremely large and requires large computing resources and efficient algorithms to cover. To meet the large computational cost of the searches, we utilized the distributed volunteer computing system, Einstein@Home (Knispel et al. 2010; Allen et al. 2013). Under this system, the parameter space is split into millions of smaller chunks, which can be searched by a typical personal computer within a few hours. These ‘work units’ are computed while volunteer’s computers are otherwise idle. We also ported our Einstein@Home search code from CPUs to GPUs, which has previously been done for radio pulsar searches (Allen et al. 2013). The approximately 10 000 GPUs active on Einstein@Home increase the computing speed by an order of magnitude."

"We are very grateful to the thousands of volunteers who donated computing time to Einstein@Home, and to those whose computers first detected PSR J2039–5617: J. Bencin of Cleveland, OH, USA; and an anonymous volunteer whose username is ‘Peter’."

Way to go E@H ! :-)

These are astrophysically very 'busy' systems, hosing the neighbourhood with radiation. Imagine an orbit of just a few hours between a tiny but dense neutron star vs an extended & distorted much lighter partner. This paper - glossing over the mathemagic - reads like a detective story. To be clear : one uses optical curves to hint at the gamma parameter space search, the results of which then lead to the correct folding of radio observations.  So E@H crucially connected the dots on this ! 

Cheers, Mike.

* But I did discover Spotify so it's not all bad news. :-)

(edit) Hence one could also perform a well targeted GW search on the same pulsar informed by said observations ?  Hope that it has a sufficiently quadrupolar nature. Hmmm .... yes, I think so.