D²OL Benchmark - DBench 1.1 !!!

[Cossey]

142 on a 3200+ at 260X10 mem only running at 173 though as the board only seems to give 2.70v even if i set it to 2.85v (neo2) ive got a ddr booster but it doesnt work as my psu hasnt got a -5v line. i need to get my electronics books out at set up a 5 to -5v flipper then i should be at 260 sync with my bh5

[enzoR]

2.4C@3.115 mem at 209 2-2-2-5 = 186 (HT on) useing Rod's formula

[cinders]

124 seconds - ain't bad I guess...

[enzoR]

is that with your FX@3ghz?

[shmaa]

What was your mem speed in each of those test?????????

whatcha mean?

all test were 1:1 if that's what you're asking... dividers on K7 =

[cinders]

is that with your FX@3ghz?

Hey enzoR, yep thats running at my current spec. Do you think the score is good for my setup or should I be expecting a faster time?...

I'm thinking of joining your D²OL team as my PC is sat tuned off all day whilst I'm at work...

Cinders.

[[XC] moddolicous]

Hey enzoR, yep thats running at my current spec. Do you think the score is good for my setup or should I be expecting a faster time?...

I'm thinking of joining your D²OL team as my PC is sat tuned off all day whilst I'm at work...

Cinders.

I think that is about right for your setup. Its not a bad idea joining the team. Anything helps us to the top.

[cinders]

I think that is about right for your setup. Its not a bad idea joining the team. Anything helps us to the top.

I shall be joining this week when I get time to set it all up.

[Rodzilla]

Coupla questions:

a) Sticky?

b) Anyone confirmed how this behaves in a multi-threaded/processor invironment? I.E. there a way to run 2 or 4 instances at once and combine the result? (or I suppose assuming 100% efficiency you could divide by 4, but let's stick to the real world here )

c) Does the bench account for the random behavior inherent in D2OL task assignment i.e. how some tasks/targets take longer than others?

Just wondering . . .

a) We already have too many...

b) You run the bench just like you run D2OL... multiple copies. There is no better way to do it and it is 100% real world.

c) The bench uses a real candidate to calculate your score. The candidate never changes from bench to bench so you can compare scores... This bench uses the D2OL engine and a real drug candidate. This is so we can compare scores and call it a bench! If the candidate changed we couldn't call it a bench. Think of it this way... If 3DMark Randomly put 2, 3, 4, 5, or 6 trolls in the game test of 2DMark 03 you couldn't call it a bench... as it's not performed as a control. The constant (or control) is the work to be done... the variable is you PC and it's setup/hardware.

[shmaa]

If 3DMark Randomly put 2, 3, 4, 5, or 6 trolls in the game test of 2DMark 03 you couldn't call it a bench... as it's not performed as a control. The constant (or control) is the work to be done... the variable is you PC and it's setup/hardware.

tell that to the barrel on car chase high.

[Rodzilla]

tell that to the barrel on car chase high.

What? Are you giving me crap or are you trying to be useful?

[shmaa]

What? Are you giving me crap or are you trying to be useful?

i'm giving that barrel crap... i woke up in a cold sweat many nights after having a nightmare about that barrel skipping across my 3dmark run.

[Rodzilla]

i'm giving that barrel crap... i woke up in a cold sweat many nights after having a nightmare about that barrel skipping across my 3dmark run.

Many years of therapy and lots of pills and you will be all right! ...Until then... you scare me!

[shmaa]

They gave up on the pills years ago...

back on topic.. is this what you wanted to know? Or was that question not directed to me?
http://www.xtremesystems.org/forums/...954#post769954

[Rodzilla]

They gave up on the pills years ago...

back on topic.. is this what you wanted to know? Or was that question not directed to me?
http://www.xtremesystems.org/forums/...954#post769954

Yea.. I was referring to this post...
http://www.xtremesystems.org/forums/...387#post766387

So you ran it with your CPU clock the same each time... But your ram was running at 100MHz and 220MHz... And you only had a 2 sec difference... Hmmmmmm... So not only do Mem timings not make a diff... But Mem clock doesn’t matter either?!?!!?!!??!?! That pretty much means that RAM doesn’t mean anything in D2OL... It may not even matter if you have enough to keep it from paging. (Which is probably true as the speed of your ram has no effect...)

[shmaa]

Yea.. I was referring to this post...
http://www.xtremesystems.org/forums/...387#post766387

So you ran it with your CPU clock the same each time... But your ram was running at 100MHz and 220MHz... And you only had a 2 sec difference... Hmmmmmm... So not only do Mem timings not make a diff... But Mem clock doesn’t matter either?!?!!?!!??!?! That pretty much means that RAM doesn’t mean anything in D2OL... It may not even matter if you have enough to keep it from paging. (Which is probably true as the speed of your ram has no effect...)

correct.

also note my main rig has 1 GB of ram and it runs the benchmark the same speed (when running at 2.0ghz) as one of my 2.0ghz crunchers with 256mb of ram running WinXP

[Rodzilla]

correct.

also note my main rig has 1 GB of ram and it runs the benchmark the same speed (when running at 2.0ghz) as one of my 2.0ghz crunchers with 256mb of ram running WinXP

Well that makes sense because D2OL only uses about 65MB per node... But what if you had all the physical mem used... and D2OL had to be paged?!?! hell... if it doesn't matter you can stick a 128 stick of the cheapest ram in a dedicated cruncher and it won't matter at all!

[shmaa]

Well that makes sense because D2OL only uses about 65MB per node... But what if you had all the physical mem used... and D2OL had to be paged?!?! hell... if it doesn't matter you can stick a 128 stick of the cheapest ram in a dedicated cruncher and it won't matter at all!

even with only a 128mb stick, if you run win2k and kill all unessential processes and services and have d2ol running, you still wont run into large page file usage.

[STEvil]

167 sec 2500+ @ 2500mhz 12.5x200 2x512 CH-5 @ 2-3-2-11 new win install
168 sec 2500+ @ 2500mhz 12x5x200 2x512 CH-5 @ 2-3-2-11 old win install
164 sec 2500+ @ 2594mhz 11.5x225.4 2x512 OCZ VX @ 2-2-2-11 old win install + steam running
149 sec 3000+ @ 2500mhz 9x277.7/133 2x256 BH-6 @ 178.5mhz 2.5-2-2-5 old win install


My winnie is a POS.. 2300mhz @ 1.65v max stable

[shmaa]

My winnie is a POS.. 2300mhz @ 1.65v max stable

Have you tried this solution for poor clocking winnie's?

http://www.xtremesystems.org/forums/...ad.php?t=55942

[[XC] moddolicous]

In order to run this, do I have to stop regular D2OL from running?

[Cossey]

yep needs to have the normal d2ol turned off otherwise itll slow down and give the result of trying to run 2 at once.

[[XC] moddolicous]

yep needs to have the normal d2ol turned off otherwise itll slow down and give the result of trying to run 2 at once.

Thanks Cossey. Just got it to work. This is from my main rig.

[[XC] moddolicous]

Here's my dad's P3 700 @ 800. Not much, but every little bit helps out, right. BTW, can someone run a test with 1t and 2t to see if that makes a difference?

[STEvil]

Have you tried this solution for poor clocking winnie's?

http://www.xtremesystems.org/forums/...ad.php?t=55942

Yup.

In fact, if I set the CPU voltage by VID only it will remain at 1.47v reguardless of setting.. or 1.37v at some

If I use the % increase then it all works correctly.. go figure. I used a DMM to verify the vcore was changing each time, too... I dont trust bioses at all.

So.. fast forward to now.. took the CBBID 0451 back and am now holding a CBBHD 04444XPKW 0081.. interesting combination