What is ur minimum time to generate prime in 20000000 ?

mohiul alam prince · Post by **mohiul alam prince** » Mon Feb 28, 2005 10:09 am

Hi
A1 i have also mesured my time by ur policy and it takes 1.421 sec(in p4)
but in uva it takes 1.523 sec.

MAP

CodeMaker · Post by **CodeMaker** » Tue Mar 08, 2005 4:53 pm

Hi , my code can get prime in the range 0<=N<=200000000. But it takes 27 second, plz tell me what is your time.

Raj Ariyan · Post by **Raj Ariyan** » Fri Mar 11, 2005 9:25 am

Hi Codemaker,
Are u ok ? 27 sec ?????? Is your time generation code is ok ? If u use simple sieve of erathosthenecs then u can get those prime 2.812 sec. If u generate prime by bit then u time will be small enough. Think abt it.

mohiul alam prince · Post by **mohiul alam prince** » Sat Mar 12, 2005 1:26 pm

Hi
I think code maker's time is 27 sec for (prime in the range 0<=N<=200000000) and for 20000000 his time will be about 27 / 10
i mean 2.7 sec.

Thanks
MAP

CodeMaker · Post by **CodeMaker** » Sat Mar 12, 2005 4:53 pm

Actually for 20000000 my time is 1.3 to 1.4 seconds not 27/10=2.7 seconds. To store that huge amount, I had to do some conversions, for low limits, those conversions will be cutoff and that saves a lot of time. so, what about my question, can anyone let me a batter time plz. i am enjoying playing with seive for some days, i thought about using bits to make the limit even 100 times larger, but i will not be able to reduce the upper bound of seive to work on that limit, even seive will fail to produce any outcome....

A1 · Post by A1 » Tue Mar 15, 2005 9:40 am

Now I am happy with my run time. I have add simple if condition in my sieve prime function and it decrease my runtime significently. Now my code run 2.23s in P(III 600Mhz) and 1.34s in P(IV 2.4Ghz)

for 20000000(20 million)
and for 200000000(200 million) in P(IV) it takes 16.5.. s

CodeMaker · Post by **CodeMaker** » Tue Mar 15, 2005 2:55 pm

Hi A1, My time is for P4 1.8 GHz. Still I have some plan to reduce it a little more.

mohiul alam prince · Post by **mohiul alam prince** » Sat Mar 19, 2005 12:14 pm

Hi
now my prime generation time is 000.760 for (1 to 20000000)
my pc is p4 1.7GHz

MAP
Thanks

CodeMaker · Post by **CodeMaker** » Wed Mar 23, 2005 7:40 pm

I reached near about 0.9 ( fairly saying - just under 1 sec)

You have done very well prince...

mohiul alam prince · Post by **mohiul alam prince** » Sun Apr 03, 2005 11:09 am

Hi
Now my prime generatoin time is 000.600

thankx
MAP

CodeMaker · Post by **CodeMaker** » Sun Apr 03, 2005 3:07 pm

You know how u did it, I give up!!! and one thing, I never use anything like MACROS, i try to speed up using normal syntax and algo. And for that , that's my limit. I can't go further....i would like to know about ur process. Well done!! atlast.

mohiul alam prince · Post by **mohiul alam prince** » Tue Apr 05, 2005 3:32 pm

Hi
You are right i have used MACROS. But My algorithm is secret. I have
solved 10311 Goldbach and Euler problem by using my algo and get ranked 9.

Thankx
MAP.

CodeMaker · Post by **CodeMaker** » Tue Apr 05, 2005 7:03 pm

there is a great chance that u will not get the credit of discovering a new algo if u don't publish it

,
ok no matter....keep it up. still u r 9th best

mohiul alam prince · Post by **mohiul alam prince** » Wed Apr 06, 2005 1:20 pm

Hi codemaker
Don't mind I will publish my algo at that time if i reduce my time below 000.200 second because there is a man who can generate prime with in below 000.200 second for (1 To 20000000).So think my
prime generation time is not good at this time.

Thanks
MAP