Super-Computer Competition!

[starbucksaddict]

They're for a "mini beowulf"
the rules stipulate only one motherboard
and I figure for programs that would run this Frankenstien
dual p3s would score better than single athlons
(thats also why I didn't choose the p3xeon board (500Mhz max))


Flame not, lest ye be flamed.

 

[Kelledin]

Hmmm does an SBC count as a motherboard? A lot of people classify the two as the same basic part.

Kelledin

bash-2.04$ kill -9 1
init: Just what do you think you're doing, Dave?

 

[Grizely1]

Solid state drives are about $40,000 for only 1.9GB!

---------
Grass is a beautiful weed

 

[Kelledin]

Actually, somebody makes a PCI "solid state drive" that holds up to 8GB PC100 SDRAM as its "media." It stays "non-volatile" by having its own power source.

The baseboard with no memory costs ~$5000, I believe.

Kelledin

bash-2.04$ kill -9 1
init: Just what do you think you're doing, Dave?

 

[Pettytheft]

There is also a 3.1 gig Solid State drive that uses a Scsi interface. This can also be placed inside the system. Last time I checked they ran for about 27k. They also have a small hard drive attatched to them to back up it's own data.

<A HREF="http://www.dirtcheapdrives.com/tech/quantum/scsi/specs/ru5320g-yf_spec.shtml" target="_new">Quantum Rushmore</A>


Blah, Blah Blahh, Blahh, blahh blah blahh, blah blah.

 

[Guest]

How about running this intensive code to test performamce for winnner system. It is written in C and will make your CPU go wild. Also check for CPU computation accurecy.
#########################
#include <stdio.h>
#include <math.h>
int main (int argc, char** argv) {

long number1, number2, counter;
double result;
if (argc < 3)
{
printf ("usage : %s number1 number2\n",argv[0]);
exit(1);
}
else
{
number1 = atol (argv[1]);
number2 = atol (argv[2]);
result = 0.0;
}

for (counter = number1; counter <= number2; counter++) {
result = result + sqrt((double)counter);
}

printf("%lf\n", result);

return 0;

}
######################
You need to compile it and run it to compute from 1 to 1000000000

My Linux 7.1 Sparc-64bit-500 MHZ machine did it in 1 minute and 13 sec. and came out with this result 21081851083600.558594 which is the exact answer, I think!
See what yours will do!

_______________________________________________~
Software is like Sex, its better when its Free!<P ID="edit"><FONT SIZE=-1><EM>Edited by nabil on 05/30/01 06:46 PM.</EM></FONT></P>

 

[starbucksaddict]

I, personally, don't consider an SBC to be a motherboard
because it has no I/O slots (that whole mother thing)

Flame not, lest ye be flamed.

 

[Grizely1]

PCI has the highest transfer rates that you could use to connect a solid state drive to, doesn't it? that would be the ultimate gaming computer, PCI 1.9GB Quantum Rushmore Ultra 60 microsecond seek time solid state harddrive OOH SWEET..... <b>*UPDATE OF ULTIMATE GAMING RIG!*</b>

---------
Grass is a beautiful weed

 

[Kelledin]

Errr...something's up with your CPU or with your program, if that's the answer you're getting back. Possibly you're passing printf an invalid format specifier (but the format specifier is correct on my system, so...?) Either that or my reasoning is somehow off (see my reasoning below).

See, the square root of the maximum counter in all that is 10,000. So the square root of any number before the final counter will be less than 10,000. But let's be generous and assume that every square root is indeed 10,000.

This means we will have 100,000,000 square roots, all equalling 10,000. Which, added together, will produce a total answer of 1,000,000,000,000.

1,000,000,000,000 < 21,081,851,083,600 (obviously)

And we know that the <i>real</i> total will be less than 1,000,000,000,000, seeing as we were so generous in our square-root estimate above. In reality there will be many, many square roots below 10,000.

So what gives?

Oh, my system completed the whole thing in less than ten seconds. The fact that the 80387 and higher FPU's have a built-in FSQRT instruction may explain that...

My system got an answer of 666,666,671,666.612061.

Kelledin

bash-2.04$ kill -9 1
init: Just what do you think you're doing, Dave?

 

[Kelledin]

From that standpoint, I guess some older Compaq DeskPro's didn't really have motherboards either...their "motherboards" plugged into an ISA riser card module.

Kelledin

bash-2.04$ kill -9 1
init: Just what do you think you're doing, Dave?

 

[Kelledin]

64-bit PCI outdoes even Ultra160 SCSI...bump it up to 66MHz (which is done on some mobos) and it even outdoes Ultra320! WOOOO-HAAAAA!

Kelledin

bash-2.04$ kill -9 1
init: Just what do you think you're doing, Dave?

P.S. Wow, I posted more than I realized. Who says you can't go from addict to forum patriarch in a single thread?

 

[Raystonn]

Hey sure thing.. you go ahead and purchase the components everyone has mentioned and I'll assemble them and run your benchmark code on it.

(btw all, sorry for being away for a while... 3 day weekend and all)

-Raystonn

= The views stated herein are my personal views, and not necessarily the views of my employer. =

 

[Raystonn]

Picture your standard 'full tower'.. with room for 5 or so 5.25" drives, maybe another 5 or so 3.5" drives, etc. Definately not more than 3.5' or so... It should look pretty much like a normal system from the outside.

-Raystonn

= The views stated herein are my personal views, and not necessarily the views of my employer. =

 

[Raystonn]

A single full tower. It shouldn't look too far from a normal system from the outside. You've got to fool Uncle Moneybags into thinking you just got a normal computer and that's how much they all cost.. (honest!)

-Raystonn

= The views stated herein are my personal views, and not necessarily the views of my employer. =

 

[Raystonn]

"is a Tbird 1.33 overclocked to 2.5 Ghz allowed ? combined with a Geforce3 @800/600Mhz"

Only if it's physically possible within 1 full tower. The complete system will sit on a standard desktop at standard room temperature. If this has been done, just show some URLs as proof and get a majority of forum readers to agree that it is indeed possible.


"I mean.. who cares man.."

To quote Darth Vader, "I find your lack of faith disturbing..."

-Raystonn

= The views stated herein are my personal views, and not necessarily the views of my employer. =

 

[Raystonn]

"palomino"

Just remember that if it is not available by the end, this will not be going anywhere. I recall reading somewhere that Palominos were not due out for desktops until around the time Northwood is released. Has that changed?

-Raystonn

= The views stated herein are my personal views, and not necessarily the views of my employer. =

 

[Raystonn]

That was originally the plan. But alas, I was forced to scrap it in order to remain gender neutral. Drat.

-Raystonn

= The views stated herein are my personal views, and not necessarily the views of my employer. =

 

[Raystonn]

If there's a motherboard available somewhere that supports it, sure.

-Raystonn

= The views stated herein are my personal views, and not necessarily the views of my employer. =

 

[Raystonn]

Those are fine. So long as they fit in something that's about the same size as a standard full tower.

-Raystonn

= The views stated herein are my personal views, and not necessarily the views of my employer. =

 

[Raystonn]

Now you're talking!

-Raystonn

= The views stated herein are my personal views, and not necessarily the views of my employer. =

 

[starbucksaddict]

"I recall reading somewhere that Palominos were not due out for desktops until around the time Northwood is released. Has that changed?"

No, but palomino for servers is to be "available" in "June"
if its not, I'll go back to planning my Beowulf.
its not as if this hypothetical situation is is any more
believable than release date rumours.



Flame not, lest ye be flamed.

 

[Guest]

Dude, Sorry but I think you'r reasoning is not very accurate. The answer I gave you is the right one and it is documented on the net somewhere. I ran the same program under a PIII-500 intel and it was off by 2 but the sparc was a closer to what have been posted on the net from years back. I suspected that my 32bit PIII-500 intel is wrong...
It does not matter what the code is doing cause obviousley it can't be done by hand or even using a scientific calculator. But I know it took my 32bit Intel PIII-500 5 minutes and 22 seconds, where my Sun Sparc 64bit 500 took 3 minutes and 21 seconds and produced a closer resut...See below:
[PIII-500 with 512 L2 cache]

[nabil@viper nabil]$ time ./sigmasqrt 1 1000000000
21081851083598.382812
real 5m29.723s
user 5m29.640s
sys 0m0.000s
[nabil@viper nabil]$

-------------------------
[Sparc UIIe-500 with 256 L2 cache]

$ time ./sigmasqrt 1 1000000000
21081851083600.558594
real 3:21.1
user 3:20.7
sys 0.0
$
----------------------------
Then I ran the job parallel in 3 computer cluster(3 CPUs) computing at the same time and got this

[2 PIII-500's and 1 Sparc-500]

21081851083601.093750
real 1m51.967s
user 0m0.010s
sys 0m0.010s
---------------------------------

What exactly are you using to compute this ??? What kind of system and OS are you using ???

_______________________________________________~
Software is like Sex, its better when its Free!

 

[Grizely1]

If I ever win the lottery, thats the hard drive for me.

---------
Grass is a beautiful weed

 

[kurokaze]

do we score points on what we manage
to cram into this system?



Intel Components, AMD Components... all made in Taiwan!

 

[ksoth]

I copied that code to Visual C++ 6.0 and tried to compile it, but I get "undeclared identifier" errors for "exit" and "atol". I'm not much into programming yet, so why won't this work?

"Trying is the first step towards failure."