32m Low Clock Challenge

[s7e9h3n]

only a prob if you are in for 3.6ghz challenge.. if you can clock B1 higher than B2 to a pt then overall 32M might be faster than B2 but with -100C monster of yours I'm sure you'll be able to max out any cpu LOL

Any cpu EXCEPT a quad-core

[massman]

massman please post a thread explaining exactly what you did

i will test that against your claims with a nlite-d XP and installation at 3.8GHz to see if it makes a difference

Owkey, I installed a few server03 OS'es with my 3000+ venice at:

7x200=1400
8x200=1600
9x200=1800
9x333=3000

Same timings, mem speed ... Only the multiplier changed the first three installations. The last one was done with 133 divider.

The 3000 > 1800 > 1600 > 1400 installation when benching at the same speeds.

[k|ngp|n]

Owkey, I installed a few server03 OS'es with my 3000+ venice at:

7x200=1400
8x200=1600
9x200=1800
9x333=3000

Same timings, mem speed ... Only the multiplier changed the first three installations. The last one was done with 133 divider.

The 3000 > 1800 > 1600 > 1400 installation when benching at the same speeds.

Are you speaking of installation times or actual speed of the benchmarks when you refer to one being faster than the other? Got any screenshots for bench results?

BTW-- my comment about installing the OS overclocked was joking ,that is definitely not one of my tweaks lol
In fact the last time I installed 2k3 I did it at 800mhz more than previous time because I was too lazy to go back into bios to lower settings, and the times were indentical on all of my pre-ln2 run checks.

[massman]

Are you speaking of installation times or actual speed of the benchmarks when you refer to one being faster than the other? Got any screenshots for bench results?

Actual speed of benchmarks. Results were on the crashed OCX forum and I haven't saved the screens.

[dinos22]

#1 or #2? BTW, nice work on that 32m
Maybe you - or anybody else for that matter - can help me figure out what the hell is going on with this Intel rig I've got here. I thought it'd be fun to try to go sub-13 min @ 3600, but it turns out I spent most of the night trying to go sub-14 min @ 3600 . I applied almost every tweak I know of - keep in mind that I only know AMD tweaks - and didn't even break into the 13's. WTF IS WRONG????? This is on an EVGA 680i and the memory's @ 2.5V. For the record, these stix are a pair of gems as they'll boot @ 4-4-3-4, DDR2 1250 at 2.5Vmem....too bad they don't run worth a crap @ that speed though. Time to do the vmem mod....

S7 there are two things not helping you there

the chip is B1 stepping
you are on 680i mobo

move to P965/975 and find a B2 chip and you will start hitting low 13M

[dinos22]

Owkey, I installed a few server03 OS'es with my 3000+ venice at:

7x200=1400
8x200=1600
9x200=1800
9x333=3000

Same timings, mem speed ... Only the multiplier changed the first three installations. The last one was done with 133 divider.

The 3000 > 1800 > 1600 > 1400 installation when benching at the same speeds.

i just find this comical

can you tell us what the difference in times was between 1M and 32M at same CPU speeds after every one of those installs

[Zeus]

Managed to shave off more than 30MHz for sub 13minutes:





Still miles away from KP though.

[Gautam]

Probably the fastest for i975X yet though.

[dinos22]

Zeus have you tried 3-3-3-x timings on 975X.....you should be able to do better with tighter timings from what i've seen on all W.R. attempts with 975X ...none of them ran CAS4

[Gautam]

Hm, getting the CAS 3 power to do it would be difficult. I should have tried when I had the chance. My guess is 460x8-470x8 3-3-3 with the right subs would pull it. But yes, 450 3-3-3 is about 5-10 seconds faster than 500 4-4-3 on i975x, though it takes a hell of a lot more voltage, and much better clocking silicon for both the ram and northbridge.

I did 13:21 myself at 450x8, never posted it out of shame against these super 965 times.

[HermS]

Owkey, I installed a few server03 OS'es with my 3000+ venice at:

7x200=1400
8x200=1600
9x200=1800
9x333=3000

Same timings, mem speed ... Only the multiplier changed the first three installations. The last one was done with 133 divider.

The 3000 > 1800 > 1600 > 1400 installation when benching at the same speeds.

I was really intrigued by this so decided to give it a little test for myself and here is what I found:-

Installed at 9 x 266 (2.4GHz) - Bench run at 9 x 400 (3.6GHz)


Then I installed @ 9 x 500 (4.5GHz) - Bench run at 9 x 400 (3.6GHz)



I figured that the vast difference in clocks at install should show any inconsistency if there was any, but basically they turned out pretty much identical, Slightly interesting that the second run started out marginally quicker then slowed down until in the end it was a couple of seconds slower, but nothing to get excited about.


Oh well was worth a try, nothing ventured nothing gained as they say


Maff

[dinos22]

nice one Herms (Maff)

so much for that theory heh

[alpha0ne]

Farrrk me Maff I dont think I have ever seen anyone load an OS @ 9x500

[Zeus]

Zeus have you tried 3-3-3-x timings on 975X.....you should be able to do better with tighter timings from what i've seen on all W.R. attempts with 975X ...none of them ran CAS4

I tried but something on my rig has a big dislike for CAS 3, don't know wheter it's the board or the RAM, tried to boot at 350 with 2.4V but it doesn't make it all the way into windows.

Somewhere in the end when it's loading the USB driver for my wireless dongle it's hanging.

I think the mem should be capable of doing CAS 3 somewhere up to 400MHz at least (Gskill 6400HZ) with reasonable voltage.

Also tried 2:3 divider but that's a sort of same story as CAS 3, the board probably hates it.

I will try some more but i don't want to fubar my OS as it's my everyday install.

[alpha0ne]

I tried 4-3-3-7 with my HZ's in sig and no matter the Vdimm or sub timings they just wont play stable, thats using both 975 and 965 boards

[hipro5]

Heh, heh, heh........We are all of theories and nothing else as for gain....

My theory is to install Windows at 166MHz fsb with the lower multi x 6 ......Thas will give you EXACLTY 6.780sec less in 32M......

[SLi_dog]

Heh, heh, heh........We are all of theories and nothing else as for gain....

My theory is to install Windows at 166MHz fsb with the lower multi x 6 ......Thas will give you EXACLTY 6.780sec less in 32M......

LOL, try installing XP with this then

From here --> http://www.winhistory.de/more/386/xpmini_eng.htm

[hipro5]

LOL, try installing XP with this then

Dumn!......I bet that you'll get the BEST time even.....

[Viss]

My first result with new mem kit. Stil a lot of work to be done Need a little more vdimm for 600 4-4-4.

570 4-4-4:

[massman]

i just find this comical

can you tell us what the difference in times was between 1M and 32M at same CPU speeds after every one of those installs

I wish I'd remember, but I tested it a long time ago. If you want ... I can't give it another try this weekend

[massman]

I was really intrigued by this so decided to give it a little test for myself and here is what I found:-

Installed at 9 x 266 (2.4GHz) - Bench run at 9 x 400 (3.6GHz)
http://i135.photobucket.com/albums/q...ultinstall.png

Then I installed @ 9 x 500 (4.5GHz) - Bench run at 9 x 400 (3.6GHz)
http://i135.photobucket.com/albums/q148/JAFFA22/4.png


I figured that the vast difference in clocks at install should show any inconsistency if there was any, but basically they turned out pretty much identical, Slightly interesting that the second run started out marginally quicker then slowed down until in the end it was a couple of seconds slower, but nothing to get excited about.


Oh well was worth a try, nothing ventured nothing gained as they say


Maff

Windows XP?

[HermS]

yes massman.

[massman]

yes massman.

I used Server03 to do the tests + A64.

I just checked my other hdd, but haven't found the screens . If there is serious interest, I'll do the tests again this weekend. If you all think this is worth nothing, please tell me, so I won't waste time.

[s7e9h3n]

try 600mhz 4-3-3-4 to see

Finally making some progress....Check out the Vmem

[Zeus]

Very nice memclocks and timings but the calculationtime is still bad.

Congrats on breaking 14 minutes, must have been a relief.