Particle Benchmark

[El_Mago_Peppins]

good

ps. i use vista 64bit

pps. have you read my last post (about the cmd line)?

[function]

Yes.
Chart corrected.

[El_Mago_Peppins]

ok i whait for news about it

ps. can you remove from chart the bar that go up from the score? XD

[function]

New version;

http://www.yazilim.com.tr/particle-bench.zip

v0.3 - 06.11.2009
Command line parameter support
Full load core engine(benchmark run in timecritical mode) (Therefore new scores may be a little high)
Adjustable thread count
Adjustable benchmark repeat
Minor bugfixes

Command line parameters:
-n Starts in benchmark mode
-c x Benchmark repeat count, default x = 1
-p x Thread count, default x = 4
-t x Custom demo text, default x = function
-d x Particle density factor(for demo), default x = 5
-a x Anialiased pixels on/off, 1 = on, 0 = off, default on
-b x Benchmarking M particle count, default 1, (changes are not recommended)

[El_Mago_Peppins]

very very good

my questions were:

custom text RESOLVED
antialiasing on/off RESOLVED
benchmark RESOLVED
particle density 1-9 RESOLVED

and other like:

window resolution (for more cpu stress) NOT RESOLVED
thread count (more thread support, 8 or more) RESOLVED
score output (very usefull) NOT RESOLVED
benchmark option: more or less particles (default is 2.000.000) RESOLVED
benchmark loop number RESOLVED
x64 version NOT RESOLVED

i'm whaiting for other option

the score output is most important

another option for this is when you running the benchmark on more loop, only the best score show up at the end of it, and not the last one, is possibile?

so...

score output (named with the local hour or with the ms score: txt or image?)
window resolution command
only the best score show up after more benchmark loops
x64 version
increase/decrease the "ball" at the center of the window with the mouse scrolling up/down (maybe the big ball can grab more fast the particles, specially in a larger window)

after all of this feature/command the final release must have a graphical real time configuration and, if you work hard on it, cuda support.

ps. what is the render mode?
pps. wht's the max number of threads?
ppps. in the bat file add the line @echo off after the first, this close the cmd window


this bench has taken me

[function]

Score output can be txt file, but only for 4-5 number text, is it required?
Window resolution freely changeble form all corners, similar to all windows.
x64 version is coming soon.
For adjustable mouse particle, wait for next version.
I'm not planning cuda support at shot date. For now, it's cpu multimedia benchmarker. May be added next major version update.
Render mode expresses two situation; paint particle or not paint. In default paints. In benchmark mode does not paint.
Thread number is theorically unlimited, but pratically mostly ideal simultaneous working thread number equal to physical cpu cores number.
"@echo off" useless to hide cmd. or not work on my pc.

Originally Posted by El_Mago_Peppins

very very good

my questions were:




i'm whaiting for other option

the score output is most important

another option for this is when you running the benchmark on more loop, only the best score show up at the end of it, and not the last one, is possibile?

so...

score output (named with the local hour or with the ms score: txt or image?)
window resolution command
only the best score show up after more benchmark loops
x64 version
increase/decrease the "ball" at the center of the window with the mouse scrolling up/down (maybe the big ball can grab more fast the particles, specially in a larger window)

after all of this feature/command the final release must have a graphical real time configuration and, if you work hard on it, cuda support.

ps. what is the render mode?
pps. wht's the max number of threads?
ppps. in the bat file add the line @echo off after the first, this close the cmd window


this bench has taken me

[El_Mago_Peppins]

Originally Posted by function

Score output can be txt file, but only for 4-5 number text, is it required?
Window resolution freely changeble form all corners, similar to all windows.
x64 version is coming soon.
For adjustable mouse particle, wait for next version.
I'm not planning cuda support at shot date. For now, it's cpu multimedia benchmarker. May be added next major version update.
Render mode expresses two situation; paint particle or not paint. In default paints. In benchmark mode does not paint.
Thread number is theorically unlimited, but pratically mostly ideal simultaneous working thread number equal to physical cpu cores number.
"@echo off" useless to hide cmd. or not work on my pc.

for the txt, if you can is another option
window resolution yes is changeable from all corners, but after you have changed it you don't know the resolution (it's only for compare cpu's fps)

so i whait for this:

window resolution command
*only the best score show up after more benchmark loops
*add decimal number support for benchmark (like 0.1 (100.000 particle) and so on)
*more time setting for benchmark (maybe the default is 100?)
x64 version
increase/decrease the "ball" at the center of the window with the mouse scrolling up/down (maybe the big ball can grab more fast the particles, specially in a larger window)

*this is most important!

ps. for the score output, is possible an html like the benchmark of far cry 2?

BTW. this is my score with cpu at 100% (8 threads)





edit. i segnale an bug:

when you open the program with the exe you see on screen the cmdline, the last is: Benchmark repeat count, the default is 1 but on the screen it show 2, the same of the demo particle, example:

if you add the command -b 3 (3 milion of particle) on the screen you see: Benchmark repeat count 3 but isn't 3, is only 1! please fix this bug

[Eliminator]

Originally Posted by CaptNKILL

I just ran it again and got 2704...



A 4Ghz Q9550 is no slouch.

It also has 50% more cache than an i7.

thats pretty interesting... i wonder what the reason behind this is... the i5/i7 will smoke any core 2 quad in other apps at the same clock speed... my Q6600 gets a measly 4610 at 3.6GHZ

anyone wanna give cinebench a try?... im curious to see what you guys get... i got 61 seconds (14370 pts) in the 64bit multiple cpu render test

[function]

http://www.youtube.com/watch?v=evHd2TrM6wE HD

http://www.yazilim.com.tr/particle-bench.avi HD 96 MB

You should watch it!

[balex99]

Thanks for the benchmark. Just ran it on my new Q9550 @ 3.23ghz. Got 3971

[balex99]

Ran it a few more times, back to back. Lowest score was 3384

[El_Mago_Peppins]

function, have you resolved the bug?