科技回声

Given the number of machines Luke Durant is running <a href="https://www.mersenne.org/report_recent_results/" rel="nofollow">https://www.mersenne.org/report_recent_results/</a> 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].<p>[1] <a href="https://www.mersenne.org/primes/" rel="nofollow">https://www.mersenne.org/primes/</a>

Earlier Hacker News discussion, after the announcement that a (then secret) number was a probable prime, yet to be verified prime: <a href="https://news.ycombinator.com/item?id=41858024">https://news.ycombinator.com/item?id=41858024</a>

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.

Since the announcement a week ago I've started contributing to the project myself. It's fun!

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.

Since the announcement a week ago I've started contributing to the project myself. It's fun!

GIMPS Discovers Largest Known Prime Number: 2^136279841 – 1

6 条评论

GIMPS Discovers Largest Known Prime Number: 2^136279841 – 1

6 条评论