Here is a whitepaper
It's not a whitepaper, it's a blog post.
Since it is such a large number GPU's become very slow at attempting it.
What about ASICs? This does not look like a problem that would be very hard to design ASICs for and just skip over GPUs entirely.
Hehe call it what you like but I have written many open source inventions as "blog posts" since blogs are a great place to self publish.
But no factoring large numbers via sieving is something that is done commonly and there is lots of literature on it. The most efficient thing is CPU's. The best "ASIC" currently would probably be built using Intel X series I9 on a mini itx motherboard with very little ram.