How fast can your puter ID prime numbers? (1 Viewer)

[Scott L]

Here's a simple little benchmark that doesn't take nearly as long as 3dmark and focuses on cpu power. Go here and click on PLGEN Binary on the side bar. Download and run the simple binary and it'll give you the time in the end.

It seems to get quicker each run (was told this is because of something called branch prediction) so post your first time. I got 83 seconds what'd you get?

 

[Shawn C]

128 seconds.....but why is he spending the time to print the results? The times would be alot faster if he wasn't drawing the results and writing to the file.

 

[Charles J P]

It will be interesting to run this on various computers to see what kind of results I get. I have a PIII laptop at 1GHz, an Athalon 1600+ at home and my brother has an athalong 1.2GHz. It will be interesting since both our processors are technically 1.2GHz core and mine is just supposed to perform equivilant to 1600Mhz. I will be ticked if his is faster or something.

DAMN! My first try took 336 seconds! 357 the second time so mine went up. Of course I'm running a lot of other apps.

Oops, I just closed SETI and it ran in 164. Still twice as long as Scott. What are you running?

 

[Dennis Reno]

582 on a PII 400mhz. Hmmm, maybe I'll close SETI and give it another try.

527 w/o SETI running. It will be interesting to give it a try on my P4 2.4ghz machine at home.

 

[MikeAlletto]

179 seconds on a 1.3Ghz P4. Thinking about taking the code and changing it so it runs on linux on our dell server here at work...hmmmm

 

[MikeyWeitz]

It will be interesting to run this on various computers to see what kind of results I get. I have a PIII laptop at 1GHz, an Athalon 1600+ at home and my brother has an athalong 1.2GHz. It will be interesting since both our processors are technically 1.2GHz core and mine is just supposed to perform equivilant to 1600Mhz. I will be ticked if his is faster or something

That is not correct, the Athlon series is not only different then the XP series of processors (even their running temps vary), your XP1600 stock speed is 1.4ghz and had very little to do with your broters 1.2 Athlon.

 

[Scott L]

I have a p2.4b overclocked to 2.7, if I run the test again I get quicker times but that's the first time I got. Is SETI really processor intensive?

 

[Charles J P]

Thanks for the correction Mikey. I was speaking from memory which obviously was faulty. So is the XP series not even an Athalon chip? I was under the impression that it was still in the athalon family. Either way, its good to know that the core speed is faster than 1.2ghz

Scott, SETI supposedly uses your "unused" cpu cycles. However, I dont know how sophisticated the program is as far as determining when to release CPU to other apps. Obviously it was giving this benchmark app enough CPU to run, but as you can see, it was using some power that could have been dedicated to making it run faster.

 

[NickSo]

160-something on my XP1700+ i think...

I closed it too fast and forgot what i got lol...

 

[Bill Slack]

Isn't all of this integer math though? That won't exactally equal 'real-world' (in the types of application where it's important, anyway...) though, perhaps some of the distrubuted client focus more on int (small, accurate, whole numbers) calculations than floats (large, estimated numbers)...

 

[Steven K]

Im totally with Bill... start dealing with Floating Point numbers and THAT will be a good test of processing power.

If there is interest in this, let me know and I'll write a benchmark tool based around floating point calculations.

Also, according to this guy's description of the app, he is calling some function he wrote (called IsPrime()) through a nested loop (although he does claim this is inefficient). I don't know what he means by this, I take it to mean that he is referring to recursive calls to this function. Indeed this is VERY inefficient!

Every time a function is called, it gets "pushed" onto the stack. When the function returns, the memory used by the function (variables, etc...) is popped back off the stack. When you write a recursive function, your function calls itself from within- therefore, it doesn't return right away. Sooooo... this memory is not being popped back off of the stack. So after not too long, you will run into a stack overflow.

The default stack size (for Visual C++ anyway) is 1 MB. You can change this value higher if you like, but doing so will mean that your application will need more memory to run. When you run out of stack space, bad, bad things happen

 

[Kris McLaughlin]

158 on an Athlon 1800XP+

running winamp, icq, outlook 2000, photoshop, dreamweaver, 2 x notepad, 2 x IE6.

That was fun, I don't have the patience for those long-winded benchmark programs.

now get back to work, Kris!

 

[DeepakJR]

136 Seconds on an Athlon 2200+ and 512 DDR Ram, i would say its pretty good.

L8rz,
Deepak Jr.

 

[Rob Landolfi]

I saw this in the middle of upgrading my system, so I figured I'd give it a try for kicks:

1997 Sony VAIO PII 233 MHZ: 1108 seconds (!!!)
Newly purchased ABS PC P4 2.8 GHZ: 77 seconds

My PII has served me well, and I even added my old hard drive (4.3 GB) to my new system as a "slop drive". The old Sony was on nearly 24/7 for over 5 years and only recently started getting a bit unstable. Have a great weekend!

 

[Scott L]

Can anyone beat the Virginians? Is it possible???

:wink:

 

[Travis Olson]

199 seconds on a P3 1ghz.

 

[Camp]

157 on a 1.4 Athlon Thunderbird running Outlook, multiple IE windows, and WMP

 

[JohanK]

79 seconds on a P4 2.53.

 

[Gabriel_Lam]

72 seconds on a dual 2.8Ghz Xeon system with 768mb DDR. Apparently, doesn't seem like this program takes advantage of dual CPU's.

 

[Scott L]

Looks like we have a winner, nice setup Gab. Might be best to close this thread before dual Xeons are obsolete! :wink: