Some 2.26B results...Nice mem bandwidth

**Bushboy** · 10-05-2002 02:40 PM

I got a brand new malay 2.26B in a trade for a 1.8a and slapped it on an ABit It7 and Oced straight away. Mem bandwidth is great and I'm almost at 3.1ghz with only 1.65Vcore..

And the mem bandwidth, this Corsair has maxed out at DR460 because even with 3.2v selected in BIOS it undervolts all the way down to 2.85v! (can anyone help me with that) But still this will do for now.

I know 10 burn-ins isnt a lot but I just wanted to show it wasn't a benchmark and reboot deal, and I'm not patient enough to do more.

**mike.elmes** · 10-05-2002 02:56 PM

Bushboy,

Seems you have ironed out a few of your wrinkles with your new board.
Nice o/c. Now get a C1 processor in there and break some records.

mike

**Bushboy** · 10-05-2002 03:00 PM

Yep, and thanks mike for your help with the previous problem, I finnaly got this IT7 working well

. I was hoping I can get enough money for a C1, which I think I will soon after I'm done having my fun with this 2.26B. I'm still trying to find a way to reduce the Vmem undervolt, I know my RAM can go higher with more than 2.85v.

BTW how much you selling that vapo for I'm considering a vap/promethea...drop me an e-mail if it's still up for grabs

**UaZa** · 10-05-2002 07:26 PM

on my it7max:

thats with geil pc3500

:P

**Bushboy** · 10-05-2002 07:37 PM

Nice score there, I was hoping to get around 240-250 FSB, I think this Corsair could do it if I could give it more than 2.85v. Is it a PSU problem maybe, I don't think my currect one gives my IT7 enough juice.

**hallowen** · 10-06-2002 09:03 AM

Way to go, nice score

**QuadDamage** · 10-07-2002 03:33 AM

because even with 3.2v selected in BIOS it undervolts all the way down to 2.85v! (can anyone help me with that)

if you have PSU with adjustable +3.3V rail, use the pot in it and crank it up. +5V/+12V rails will drop so use second pot. i had similar probs with XMS 3000@2.9V, but managed to feed it with 3.05V, thanks to AOpen 400w PSU.

**Bushboy** · 10-09-2002 01:44 PM

Thanks, care to share where you got the PSU from, as I haven't found it, or if you could recommend me one that you know has the adjustable pots...I know some dont.

**Bushboy** · 10-09-2002 02:43 PM

Ahh but of course the OCing doesn't stop

...

Finally got 6000MIPS

Just when I thought it wouldn't go further...

**Bushboy** · 10-10-2002 02:15 PM

and a pcmark score...I have no idea how good this is, CPU is only at 3ghz though (176FSB).

**Bushboy** · 10-10-2002 04:50 PM

How bout this CPU benchmark, a few mhz makes a difference I guess! I tested it three times and it came out the same each time.

**mike.elmes** · 10-10-2002 05:26 PM

What's curious....not that I am doubting.....is that I have seen a big discrepancy in Sandra CPU benchmarks. I believe there to be a bug of some sort as I saw some benchies last week by PiLsY
where their was a 500-600 mips difference.

**Bushboy** · 10-10-2002 06:01 PM

Yes I saw that one too and thought this was the case rather then the extra mhz I put on but I ran it several times and it still came out more or less the same. I'll test it again at that speed later and see what I come up with just to make sure.

**mike.elmes** · 10-10-2002 07:09 PM

OK ....I did 3 runs on Sandra CPU . I scored 6263 mips at a speed of 3089.89 mhz.

I guess there may be some differences between the different versions of Sandra. My Sandra is the 2002 standard. Perhaps your sporting the pro version.

I saw some wild fluctuations in a thread last week by PiLsEY.

Probably variation within the various versions....say that 10 times real fast.

**Bushboy** · 10-10-2002 07:23 PM

Hmmm I've rerun and what happened is it started at 6200MIPS, then kept going up and by the 4th run it was pretty much what I have in the screeny. So I rebooted and tried again, but this time it started at 6500MIPS then gradually went down and stayed at 6400 apart from once where it was 6301. So basically at 3.1ghz it ranged from 6200-6600...a lot of fluctuation I agree, and yes this is the pro version.

**Klosterbruder** · 10-11-2002 03:10 AM

i have benched sandra with 3,1 Ghz too but my score was only 5900.

**Bushboy** · 10-14-2002 01:27 AM

that's odd, it seems a bit low. Perhaps the 1.6s have lower scores even at the same speeds. I've now been running 5 CPU benches each time and taking the most sensible or average one to be fair with fluctuation. I got my TEC on there with some chilled water and this was the last pic I took before the HDD mysteriously corrupted, I'm working on it now...

EDIT: I have my water chilled sub zero now and have put in a new MBM screeny...this was NOT the same screeny as after benching but at the same speed and Vcore. THis is an idle temp but I just had to post it

**UaZa** · 10-14-2002 06:12 AM

thats really wierd!!
here is my cpu bench at 3.1

cpu temp

**Klosterbruder** · 10-14-2002 07:49 AM

hmm, uaza these are the same scores like I have. OK my CPU working at his limit and produce some errors, thats perhaps the reason for this scores. But you have 1000 Mips more than we have thats really a lot. I know guys with 3.3 GHz who have around 6400 Mips. Your Multimediascore is same I have, but the other ones is to huge.

**Bushboy** · 10-14-2002 09:23 AM

Are both you guys using windows 2000 or other, because it's no the XP design even though you can chage it to windows classic. I thought perhaps it was XP, so I went into my windows 2000 partition and the scores were not as high but still about 6700 after a few runs. I don't understand really why it's so high

**Klosterbruder** · 10-15-2002 07:03 AM

I made with XP, too.
ok what is your score after the first run? Cause my benches are at the limit and so i could do only one, perhaps it increases if you do it more often. I've heard of somebody who get over 24k 3d marks, cause everytime he repeat the bench, he get 2k more, perhaps you have something similar.

**Bushboy** · 10-15-2002 10:06 AM

I'll do some more benching but usually I run about 5 CPU benchmarks and post the average one. Yesterday I took my TEC off and at 3.1ghz my score in XP was about 6500 every time...I put it back on with sub zero CPU temps, and it was back up to 7000+ MIPS. Perhaps the cooling plays a major part in arithmatic benchmark too.

**mike.elmes** · 10-15-2002 11:13 AM

Bushboy,

I believe your onto something here....When I had my Vapochill on a tbird I used to score much higher on some Sandra scores.

I think I remember....it's called thermal acceleration do to temperature. I will be able to prove this once my prometeia arrives. In the mean time you have essentially proven the same thing...run your computer at a set speed, on air, then cool sub zero and run computer at the same speed....I can guarantee you,the colder you run,the higher your Sandra CPU benchmark will be. I would expect this would also benefit 3dmark scores.

comments welcome,

mike

**Klosterbruder** · 10-15-2002 11:27 AM

hmm i think, if the temp is lower, the eletrical resistance (don't know if this is the correct English word) is lower and the electrons are faster, that's perhaps the reason for higher scores.

**Bushboy** · 10-15-2002 02:31 PM

I'm not sure of the science of it, but Mike you were right about 3dmark! At 2.73ghz and card slightly OCed to 275/560 (I think) I used to get 12000 on air/WCIng. Now with TEC and -15C it's at 15K!and BTW this screeny was taken right after the benchmark was finished and CPU usage went down, I didn't wait for it to cool down at all!

Thread: Some 2.26B results...Nice mem bandwidth

Thread Tools

Search Thread

Rate This Thread

Display

Some 2.26B results...Nice mem bandwidth

Posting Permissions