Please help me with my Computer Science project...Check out my new, cross-platform compatible benchmark!

[WarCon]

redpriest_ I think some people just store irrelevant trivialities for the sole purpose of regurgitating them in order to show superiority. I hadn't ever heard that irregardless is considered a non-word because of the negative prefix and suffix. Never much thought about it, but I sure have heard and seen it used commonly.

Personally, I don't give a plug nickel for that kind of "knowledge."

Edit: Placed my period inside the quote for maximum goodness..........

 

[adlep]

So, what other algorithms can I use to make it better?

 

[adlep]

I can not let it die just yet...
I am still improving the code.

 

[Chumpman]

Originally posted by: WarCon
redpriest_ I think some people just store irrelevant trivialities for the sole purpose of regurgitating them in order to show superiority. I hadn't ever heard that irregardless is considered a non-word because of the negative prefix and suffix. Never much thought about it, but I sure have heard and seen it used commonly.

Personally, I don't give a plug nickel for that kind of "knowledge".

When ending a sentence with a quote, the quotation mark should always be on the outside of the period.

 

[Sideswipe001]

LOL

This thread has become English class.

 

[Budman]

Benchmark complete.
It takes 60.985 seconds to sort 98304 randomly generated integers.


It seems as it's only using 50% of my cpu,maybe need to have it hyperthreading enabled.

 

[adlep]

Budman

It seems as it's only using 50% of my cpu,maybe need to have it hyperthreading enabled.

? ;confused;
It uses 100% of my P4 2.53 Ghz without HT, so you are saying that you do have HT enabled?
P.S. I dont care that much about English, this thread is about SortMark...
Please stick to the topic, or create your own English thread...
Thanks

 

[KF]

>You're really pretty stupid. The reason I mentioned it was so that in your normal writings or speech,
> you don't come off looking like an idiot. I was just trying to help you. However, judging from
>your reply, I could never give you enough help to make that happen.

I suppose imperfect composition is a fault. A very minor fault. There are worse. A person could have bad manners. He could be obnoxious. He may have a personality disorder. It may be that he simply lacks due consideration for the feelings of other people; it may be his upbringing was lacking; or he may simply enjoy tolling forums to ridicule others like his teachers ridicule him.

>You're really pretty stupid.

This is poor composition. You used a contraction: You're. "Pretty "is used in a non-standard way. So is "really."

>The reason I mentioned it was so that in your normal writings or speech, you don't come off looking like an idiot.

"Come off" is idiomatic. "Normal" is used in a non-standard way. You are making yourself look like a dim bulb.

I often wonder where people are going with word aversions like "irregardless." Every word was not a word until it became a word. Historians of this sort of thing can point out instances where non-standard usage has supplanted a formerly standard usage. Living languages are expanded and revised continuously. There is no original or true English. There never was. We vote for a particular form by our own usage. Grammarians and language prudes have been overruled over and over again, as common usage has worked its way into formal usage. People who are unaware of this show their ignorance while they imagine they are exhibiting their knowledge. Proper English is a slippery, undefined concept, like "proper" hair style.

I really don't know how "ir" got attached to "regardless" or why people gravitate to it. Or why it irks purists no end.

>However, judging from your reply, I could never give you enough help to make that happen.

Really? In his reply, he seems to have understood perfectly. He explained it without a laborious and pendantic quotation.

 

[Budman]

Originally posted by: adlep
Budman

It seems as it's only using 50% of my cpu,maybe need to have it hyperthreading enabled.

? ;confused;
It uses 100% of my P4 2.53 Ghz without HT, so you are saying that you do have HT enabled?
P.S. I dont care that much about English, this thread is about SortMark...
Please stick to the topic, or create your own English thread...
Thanks

Look at my system rig,yes I have a 2.4C @ 3ghz with HYPERTHREADING ENABLED.

 

[adlep]

Look at my system rig,yes I have a 2.4C @ 3ghz with HYPERTHREADING ENABLED.

Then I guess only one half of your processor is working with SortMark
Pardon the bugs, this is still work in progress..

 

[Budman]

Originally posted by: adlep

Look at my system rig,yes I have a 2.4C @ 3ghz with HYPERTHREADING ENABLED.

Then I guess only one half of your processor is working with SortMark
Pardon the bugs, this is still work in progress..

Shouldnt matter that much,but if it was optimised with HT it would be a tad faster.

 

[KillaBong]

73.4 seconds
Windows XP Pro
Athlon Tbred 1733 @ 2300 (200x11.5)
Epox 8rda+
512mb Kingston PC3000 @ 400mhz

 

[kponds]

56.63 seconds on:

P4 2.0Ghz OCed to 3.2Ghz ($105 from Anandtech forsale forums)
Albatron PEV PRo ($55 from Anandtech forsale forums)
512MB Samsung True PC3200 @ 2.66 Multiplier (DDR 425) ($88 - I can't believe I paid full retail for this)

 

[PhoenixOfWater]

P4 1.4Ghz
Intel Mobo
512MB Rambus PC800
FreeBSD 4.8
131.156 seconds (in X(KDE 3.1))
130.012 seconds (out of X)

P4 1.6A@2.4Ghz
Albatron PX845PEV Pro
512MB DDR400
Win Xp Pro
76.281 seconds

AMD 1600+(1.4Ghz)
Asus Mobo
512Mb SDRAM PC133
Win XP Pro
135.015 seconds

P3 Cel Tualatin 1Ghz (100FSB*10)
Asus P3C-E
256MB Rambus PC 400 (don't ask why a P3 have rambus)
FreeBSD 4.8
192.07 seconds (out of X)

AMD 1.3Ghz
Jetway Mobo
128MB SDRAM PC133
Slackware Linux 9.0
183.61 seconds(Underclocked to 1Ghz (100FSB*10)in X Gnome 2.2)
182.23 seconds(Underclocked to 1Ghz (100FSB*10)out of X)


Ideas:
1. have it output to a text file and beable to keep adding to the text file like
183.61 seconds 06/21/2003 2:25AM
182.23 seconds 06/21/2003 2:28AM

2. Make a Dos ver and make a bootdisk that will run thats at startup that way ppl can try OS Vs. OS on the same computer with out have a dual boot you could also make a Linux boot disk and do that samething

Just ideas

if you need any more info on the computer I tested on just PM me
Hope this info helps out
Later

 

[adlep]

THX Phoenix!
Next release will incorporate a data prefeching, to make a greather use of CPU cache.....

 

[PhoenixOfWater]

Originally posted by: adlep
THX Phoenix!
Next release will incorporate a data prefeching, to make a greather use of CPU cache.....

NP
if you need any more help just ask
You can sigh me up as a beta tester if you want

 

[adlep]

Thank you all for the replies.....

 

[alpineranger]

You shouldn't regard this as a cpu benchmark, for several reasons, including the following:

Bubble sort on a very large array is inherently memory bound. Obviously the whole data set will not fit in the cache. That said, the working set over a small time interval will, and the memory access patterns are nice and predictable, which makes this less of a problem than it might be in other algorithms. Still, data can be read and written from the main memory only so quickly, and the cpu will spend a great deal of time spinning its wheels waiting for the memory.

Cross platform benchmarks are typically not very reliable because of architechtural differences that can be taken advantage of (to the detriment of other architechtures). Additionally, even if you try to be neutral and code in a high level language in an architechture neutral way, you will be using optimizing compilers that well, optimize for particular processors (depends on the compiler). Additionally, some compilers optimize better than others.

 

[adlep]

You shouldn't regard this as a cpu benchmark, for several reasons, including the following:

Well I agree with you partially.
1. I will try to implement the data prefeching element to the benchmark, so it will actually make a use out of the CPU cache. Right now I know that the benchmark is not using the cache very efficiently. In fact I think it is not using the cache at all.
Once again data prefech is more effective on G4 and Athlon and less effective on Pentium 4 therefore the next revision will give all of the platforms an opportunity to show their architectural strength rather than just the Mhz strength.
2. I am compiling this program on a g++ compliant, generic, freeware Dev C++ compiler, on Windows. Therefore I don't think that it will optimize the code better than generic gcc or g++. I would say all of them will compile the code in a similar way because they come from one gcc source. I am not using anything special, such us MS Visual C++. Just generic, similar, plain vanilla compilers.
3. If you will look at the results....Well, they seem to depend primely on the CPU speed and not on the memory speed, therefore how you can say it is not a cpu benchmark?
The CPU speed seems to be a critical factor here...
Take a look. Similar platforms, but diffrient results:

56.63 seconds on:

P4 2.0Ghz OCed to 3.2Ghz
Albatron PEV PRo
512MB Samsung True PC3200 @ 2.66 Multiplier (DDR 425)

and

P4 1.6A@2.4Ghz
Albatron PX845PEV Pro
512MB DDR400
Win Xp Pro
76.281 seconds

In both cases, the memory speed is similar, yet the difference between CPUs used is 800 Mhz.
Therefore I would argue that for some odd reason SortMark likes fast CPUs? :Q