GPU LLR program
 

Is anyone working on a GPU-based LLR program? And if no one is, would anyone be interested in one?

No guarantees, since I haven't mentioned it to the person yet, but I know there's a version he's written which works for people willing to fight with their computer to get it to work, a beta in other words. He's definitely not going to be willing to release the source code, but I doubt if he'd be angry about people worrying about trojan activity.

How often do you offered such things?

Knowing someone who got a new, primefinding (in no time) program/algorithm but can't say who he is and he can't give any source or example to convince other in here, because his code is under the law of his company.

So, do us a favour:
Give examples/results or code so any anybody can see if you/your "invisible friend" got new things which are worth to discussed.

Otherwise this should be the last post of this thread.

[QUOTE=kar_bon;269676]How often do you offered such things?

Knowing someone who got a new, primefinding (in no time) program/algorithm but can't say who he is and he can't give any source or example to convince other in here, because his code is under the law of his company.

So, do us a favour:
Give examples/results or code so any anybody can see if you/your "invisible friend" got new things which are worth to discussed.

Otherwise this should be the last post of this thread.[/QUOTE]
I haven't talked to him personally, yet, that's why I'm being so secretive. I'd ask the forum, get flamed, then he'd get offended personally. Well, actually, to be honest, I believe he got offended because he gave me a quote to post here and THAT got flamed. So, technically, it wasn't the flaming I got that offended him.

I have ADD and am more of a free spirit than a lot of people on here, so the flaming was fast and furious when I used to post on here. A lot of my unnamed friend's prime-finding methods(separate from the program I mentioned in terms of using proven methods) are more statistical than 100% scientific, which I think is one of the reasons for the ruffled feathers when he used to post here. He believes 6x+/-1 can also help in figuring out where to look for primes, rather than simply being a fact about all primes other than 2 or 3.

I was simply making an offer, it'll just be typical of the forum if this gets flamed and locked.

[QUOTE=kar_bon;269676]How often do you offered such things?

Knowing someone who got a new, primefinding (in no time) program/algorithm but can't say who he is and he can't give any source or example to convince other in here, because his code is under the law of his company.[/QUOTE]
It isn't the same code. Plus, if we're talking about the same thing(and, no, this new program ISN'T patented by his company) one of the reasons he's not willing to distribute it is because it's only legal if he makes no profit and his labor is free. My brain is mis-behaving now, so I can't remember the specific type of hardware that's physically malleable when you first buy it, but if I ever have 1k to throw around, he can make me an LLR computer that can do as much as about $20-30k of normal hardware. The problem is that it takes hours for him to alter it and that labor has to be 100% free for it to be legal for him to distribute it. Also, in terms of the number, Riesel Sieve had to shut down their operations for very similar reasons. I've heard that if a US citizen finds a prime that breaks US government code, they can be imprisoned even if it's 100% by accident. Simply put, if someone in the US has threatened US security, even by accident, their rights under the Constitution become severely limited.

Edit: Riesel Sieve was shut down because of the bit-level they were sieving at could potentially threaten US codes. The prime number my friend found is kept secret for similar reasons.

[QUOTE=jasong;269680]I've heard that if a US citizen finds a prime that breaks US government code, they can be imprisoned even if it's 100% by accident. Simply put, if someone in the US has threatened US security, even by accident, their rights under the Constitution become severely limited.
[/QUOTE]
Thanks for a good laugh on a Sunday. I like those movies too.

[url]http://www.mersenneforum.org/showthread.php?t=14608[/url]

[QUOTE=jasong;269680]Edit: Riesel Sieve was shut down because of the bit-level they were sieving at could potentially threaten US codes. The prime number my friend found is kept secret for similar reasons.[/QUOTE]

So PrimeGrid is a Government office?

If not, PrimeGrid has to be closed immediatly, because their sieve depth is some bits deeper than RieseSieve ever was!

Oh, and please give us the speical length of such primes. I don't want imprisoned by finding such a prime the next time.

[QUOTE=jasong;269680]Riesel Sieve was shut down because of the bit-level they were sieving at could potentially threaten US codes. The prime number my friend found is kept secret for similar reasons.[/QUOTE]

It's time for your meds again.

[QUOTE=CRGreathouse;269693]It's time for your meds again.[/QUOTE]
Do your research, B2's computers were temporarily siezed and searched before the Riesel Sieve project was shut down. Do you think that was a coincidence.

edit: I suppose it's possible the government changed it's mind, which would explain why B2 released the data after months of protest. Note that that's plenty of time for the government to change their codes away from anything having to do with Riesel numbers.

[QUOTE=kar_bon;269691]So PrimeGrid is a Government office?

If not, PrimeGrid has to be closed immediatly, because their sieve depth is some bits deeper than RieseSieve ever was!

Oh, and please give us the speical length of such primes. I don't want imprisoned by finding such a prime the next time.[/QUOTE]
You might think it's extremely funny, but government encryption is based on prime numbers, so there's always the possibility of invading territory that affects government security. Think about it, supercomputers are based on off-the-shelf parts and the BOINC collection could easily fit in the top-20 of supercomputers if people wanted to treat it that way.

[QUOTE=ixfd64;269689][url]http://www.mersenneforum.org/showthread.php?t=14608[/url][/QUOTE]
The program I'm talking about was never published publicly and does at least 4 times as well as a Core2 processor would, as in 4 times as fast. And that's with budget hardware of a year or two ago.

Edit: Is CUDA the AMD GPU language? I think their's goes by a different name, but that's what he prefers, if I'm not mistaken.

Edit2:The program was made back when it was a real bitch to program GPUs, and it uses integers, unlike Prime95. That's why I was active in the GPU threads years ago, because I had access to an expert.

Edit3: B2 has government clearance, so they may have been able to convince him with facts that couldn't be told the admin of Primegrid. It's possible RS was shut down before it could've become a problem. Remember, RS was about to obtain a program that went beyond 64-bits, and the Primegrid project didn't start til much later so plenty of time for the government to get their ducks in a row.