Re: [PrimeNumbers] Re: Ruler, compass and prime power
- On 11/1/2011 9:09 PM, djbroadhurst wrote:
>Looks like my Xeon is older and slower than your Xeon. ;)
> David Cleaver wrote:
> > SIQS elapsed time = 2243.3906 seconds.
> A very old Msieve v.1.20 linux executable took 43.0 minutes,
> using MPQS, on an anonymous grey box with these vital statistics
> CPU Information:
> Intel(R) Xeon(R) CPU E5620 @ 2.40GHz
> CPU speed: 2400.20 MHz
> CPU features: RDTSC, CMOV, PREFETCH, MMX, SSE, SSE2
> L1 cache size: unknown
> L2 cache size: 256 KB
> as revealed by the -m option of LLR.
> What might a similar inquiry reveal about your box, please?
> Best regards,
Intel(R) Xeon(R) CPU E5335 @ 2.00GHz
CPU speed: 1994.94 MHz
CPU features: RDTSC, CMOV, PREFETCH, MMX, SSE, SSE2
L1 cache size: 32 KB
L2 cache size: 4096 KB
I am also using Windows XP Pro 64-bit version for my OS. Actually, I have two
Xeon 5335's (each a quad core) in this computer. But, 6 of the cores are busy
running gmp-ecm, so I only ran yafu with a single thread. I know msieve can use
multiple threads too, I think it would be interesting to see how much faster we
could factor this number using multiple threads. However, my gmp-ecm won't be
at a stopping point for at least another 137000 seconds, so I won't be able to
run experiments until then.
- --- In firstname.lastname@example.org, David Cleaver <wraithx@...> wrote:
>I saw that, thanks David!
> Ben, I've been trying to tell them how awesome Yafu is! You can see my message
> If anyone needs any factoring utilities, yafu should be first on the list. Then
> some combination of yafu/msieve/ggnfs to factor larger numbers. I tried to
> spell it all out in the above post. Hopefully I didn't misrepresent any info
> about yafu. Please correct me if I was wrong. If anyone has any questions,
> feel free to ask on this list.
> -David C.
> P.S. For full disclosure, I helped contribute a small amount of code to yafu. :)
Since this is a list dedicated to primes, I'll also just mention briefly that yafu has one of the fastest sieve of Eratosthenes implementations I'm aware of, for generating lists of primes in arbitrary ranges up to 10^19. Maybe that is useful to folks here too.