Given the number of machines Luke Durant is running
https://www.mersenne.org/report_recent_results/
it must cost a lot of money, given those are GPU nodes. I wonder what's the story there. Also I think up till yesterday the name was not shown on his submissions, it was only shown as ANONYMOUS.
Bad luck, there just happens to be no prime number for a long time. The new prime exponent p is 136M, the old largest one was around 82M. So the overall effort to get from 82 to 136 was around 1.75 times the effort it took from 0 to 82 [calculated as (136/82)^2-1]. So it was more effort to find the 52nd than it was to find all 51 before.
See here for the relatively unusual large gap: https://www.mersenne.org/primenet/
Given the number of machines Luke Durant is running https://www.mersenne.org/report_recent_results/ it must cost a lot of money, given those are GPU nodes. I wonder what's the story there. Also I think up till yesterday the name was not shown on his submissions, it was only shown as ANONYMOUS.
The first Mersenne prime since 1996 not to be discovered on x86 hardware, but on an NVidia A100 [1].
[1] https://www.mersenne.org/primes/
> discovered [...] on an NVidia A100
Running an OpenCL (not CUDA) program [1], that runs just as well on AMD as on Nvidia GPUs.
[1] https://github.com/preda/gpuowl
Earlier Hacker News discussion, after the announcement that a (then secret) number was a probable prime, yet to be verified prime: https://news.ycombinator.com/item?id=41858024
There are a lot more details available now worth discussing. This thread shouldn't be deprioritized as dupe, it isn't.
I wonder where Luke Durant got the money from to fund the effort. He's a former NVIDIA employee, I wouldn't be surprised if this was funded by NVIDIA.
Wonder why it took so long to find a new one. I get the time complexity is a bit more than O(p^2) by a Wikipedia search, but still, 6 years is a lot.
Bad luck, there just happens to be no prime number for a long time. The new prime exponent p is 136M, the old largest one was around 82M. So the overall effort to get from 82 to 136 was around 1.75 times the effort it took from 0 to 82 [calculated as (136/82)^2-1]. So it was more effort to find the 52nd than it was to find all 51 before. See here for the relatively unusual large gap: https://www.mersenne.org/primenet/
In some ways, it's right on schedule: https://twitter.com/deepakvenkatesh/status/18483562170366856...
Since the announcement a week ago I've started contributing to the project myself. It's fun!