Sign in with
Sign up | Sign in
Your question

YAY 1400Mhz+ at LAST!

Last response: in CPUs
Share
October 20, 2001 9:05:49 AM

For you people who have been around for a while, you will remember my Detailed overclocking of an athlon 1200C and a7v133.
154Mhz system bus speed was my limit, as was around 1386Mhz... 1400 just never ran stabily.

now though, with a new mobo (iwill kk266r-plus), same core, same HSF, new PSU (enermax 550W) ive achieved the result ive always wanted!
150x9.5 = 1425Mhz! at 1.95V, 41C idle, 50C full load

interesting that both the a7v133 & kk266r-plus have their max voltage as 1.850V, yet the asus board gives me 1.89V while the iwill gives 1.95V...

am running The Toast! program as reccommended by Raystonn. 1 hour up and still going strong... looking good as previously i could barely run it for 30mins.
full load temp is 50C. mobo 22C, room 19C.

course due to reseating the HSF and the new mobo CPU temps are 7-8C lower than reported on the old asus a7v133. so it would be 58C... bout right for max full load.

hmmm... wonder if 1450 is possible... doubt it...but still...
should i do 9.5x153 or 10x145? 153 is better, but perioulsy close to my 154 hard drive crap out limit :( 


Religious wars are 2 groups of people fighting over who has the best imaginary friend.

More about : yay 1400mhz

October 20, 2001 9:31:33 AM

Heh, that's all well and good...until we have a 40 degree day. :) 

"Ignorance is bliss, but I tend to get screwed over."
October 20, 2001 12:03:44 PM

hmm, could you do me a favour? Try running the Prime95 torture test.

AMD technology + Intel technology = Intel/AMD Pentathlon IV; the <b>ULTIMATE</b> PC processor
Related resources
Can't find your answer ? Ask !
October 20, 2001 12:38:56 PM

thats just a full run of prime 95 right???
nothing special?

Religious wars are 2 groups of people fighting over who has the best imaginary friend.
October 20, 2001 1:14:32 PM

Go to Options, Torture Test

AMD technology + Intel technology = Intel/AMD Pentathlon IV; the <b>ULTIMATE</b> PC processor
Anonymous
a b à CPUs
a b K Overclocking
October 20, 2001 4:49:14 PM

Congrats mate, just wondering, does the increased FSB reflect on your 3dmark2001 score (default setting)?
October 20, 2001 5:06:25 PM

yes..FSB raises clock freq AGP bus ratio and memory bus clock, u tell me would that help?:) 

<font color=green>
*******
*K.I.S.S*
*(k)eep (I)t (S)imple (S)tupid*
*******
</font color=green>
October 20, 2001 5:13:21 PM

lhgpoobaa, please tell me if you can get the Prime95 Torture test to work for at least 10 minutes.


AMD technology + Intel technology = Intel/AMD Pentathlon IV; the <b>ULTIMATE</b> PC processor
Anonymous
a b à CPUs
a b K Overclocking
October 20, 2001 5:43:46 PM

I know, Ops sorry i just noted that it is actually a decrease of FSB earlier his FSB was at 154 and now 150 Ok since FSB raises clock freq AGP bus ratio and memory bus clock like you pointed out would the higher clockrate now make up for the lower FSB in 3dmark2001? Thats how i should have phrased it :) 
October 21, 2001 1:59:51 AM

erm nope. prime torture test craps out after about 5 minutes.
ERROR ROUNDOFF [0.4996490479] > [0.4]
Possible hardware failure, consult readme file.

mostly happens by the 2nd of 3 tests.

find it a bit strange as cure for cancer runs for as long as i want at 1425 without errors, as does Toast! which makes the CPU FAR hotter.

oh well. dont care too much about that.
could try increasing the VIO voltage.

As for 3dMark & an elevated FSB, previously i found 3dmark to be very twitchy with it... anything over 140 or so and it spontaneously goes back to the desktop... then again, that was on the OLD motherboard.
should try 3dmark again.



Religious wars are 2 groups of people fighting over who has the best imaginary friend.
October 21, 2001 2:16:11 AM

What's Prime95, and where do I get it from?

"Ignorance is bliss, but I tend to get screwed over."
October 21, 2001 2:24:14 AM

That's exactly the same thing that happens to me! All the our torture tests work perfectly at 1.4GHz on my computer. Very weird!

AMD technology + Intel technology = Intel/AMD Pentathlon IV; the <b>ULTIMATE</b> PC processor
October 21, 2001 2:30:03 AM

If anyone has a stock 1.4GHz Athlon running at it's default speed, please do me a favor, try running the Prime95 torture test and tell me what happens.

AMD technology + Intel technology = Intel/AMD Pentathlon IV; the <b>ULTIMATE</b> PC processor
October 21, 2001 2:37:16 AM

I can get a friend of mine to do it...if you tell me what Prime95 is, and where I can get it from. :smile:

"Ignorance is bliss, but I tend to get screwed over."
October 21, 2001 10:12:27 PM

Well the problem (if it was one) has been solved. :) 
cranked back the fsb to 133 and the multiplier to 10.5. thus still getting 1400Mhz but with a usual 133 bus speed. bingo. no problem. prime95 torture test pasing with flying colours.

99% chance its my ram. Kingmax is officially rated for 3,3,2 @ 150. See on my old asus board the mem settings wernt as harsh. only allowing 2,2,2, nothing else, AND asus did the sneaky thing by setting the 3.3V line to be PERMINANTLY 3.56!

so when i got my iwill kk266 the 3.3V line WAS 3.3, and i had the added memory option of 2 or 4 way interleave and an extra ras to cas setting, i.e. 6,3,3 or 6,3,2 or 6,2,3 or 6,2,2 or even 5,2,2 (which is what im using)

so when i first tried 150Mhz @ 5,2,2 cas 2, 4way interleave, i had the unusual occurance of "loosing" ram... ie it would boot as 296mb, 320mb, or occasionally if i was lucky 512... sure sign the sticks wernt quite getting the juice they required.
course there is a VIO jumper, allowing me to set the VIO for an additional 5% or 10% extra.
at the moment its set to 5% extra, giving 3.45V... but thats still below the 3.56 that the asus board was set at. looks like i should give 10% jumper setting a go.
gotta be safe as the ram spent 5-6 months on the asus board at 3.56V
easy huh?

oh...P.S. for sheer "lets see if i can fry my processor" nothing ive found so far can beat Toast!. CPU temps for me +4C above that which any other program can achieve! unfortunately it doesnt have error checking like prime95.

Religious wars are 2 groups of people fighting over who has the best imaginary friend.<P ID="edit"><FONT SIZE=-1><EM>Edited by lhgpoobaa on 10/21/01 06:17 PM.</EM></FONT></P>
October 21, 2001 10:25:49 PM

Hmm, interesting, so do you think my problem will go away if I set the VIO to 3.5V? I have it at 3.4V right now because I need 3.4V to overclock my graphics card. Do you think I should raise it to 3.5V?

AMD technology + Intel technology = Intel/AMD Pentathlon IV; the <b>ULTIMATE</b> PC processor
October 21, 2001 10:29:32 PM

*shrugs* maybe... your choice.

as it was default 3.56V on the asus board and nothing got fried it should be allright.
give it a go... tell me what happens.
im gonna do the same shortly.

Religious wars are 2 groups of people fighting over who has the best imaginary friend.
October 21, 2001 11:29:43 PM

Hmm, no, it doesn't work, I think my CPU is the limitation. Anyway, I'm still happy with my SDRAM. I'm currently running at approx. 1350MHz with 142MHz bus and 9.5 multplier. I can probably get it higher but I'll do that later. My memory benchmarks are 585MB/s Int ALU/RAM and 660MB/s Float FPU/RAM. Not bad for SDRAM!!! I'll get it up to 150 later, that should really get me into the mid to upper DDR range. Right now this score is similar to PC2100 Cl2.5 on an AMD760 chipset. At 150MHz, I might be able to get to 700+ for FPU so I'll be in mid DDR range.

AMD technology + Intel technology = Intel/AMD Pentathlon IV; the <b>ULTIMATE</b> PC processor
October 21, 2001 11:56:11 PM

hmm, seems my Micron PC133 CL2 RAM maxes out between 142MHz and 143MHz at 3.5V. I guess the Micron PC133 isn't the best for overclocking. Do you think it's safe to raise the VIO to 3.6V?

AMD technology + Intel technology = Intel/AMD Pentathlon IV; the <b>ULTIMATE</b> PC processor
October 22, 2001 12:13:20 AM

yup, as long as u keep it under 10% over original V specs it should be fine.

<font color=green>
*******
*K.I.S.S*
*(k)eep (I)t (S)imple (S)tupid*
*******
</font color=green>
October 22, 2001 12:23:43 AM

shame.
well remember thats amd 760 ddr, not sis735 or kt266a.

and if you can do 150mhz cas 2 expect to see ALU ~ 630 & FPU ~700.


Religious wars are 2 groups of people fighting over who has the best imaginary friend.
Anonymous
a b à CPUs
a b K Overclocking
October 22, 2001 2:19:06 AM

No. Get it <A HREF="http://www.mersenne.org/freesoft.htm" target="_new">here</A>. This is a newer version that runs hotter and stresses memory more than the previous version.
Anonymous
a b à CPUs
a b K Overclocking
October 22, 2001 7:27:11 AM

I'm surprised you're having such a hard time getting your 1.2 to 1.4.. What HSF are you using ? My brother has a 1.2 T-bird that works flawlessly up to 1.5 Ghz with a less-than-ideal Silverado. He claim its stable, though I have no idea how he tested it.

I have a 1.4 with a swiftech, and it makes it to 1.6. Anything higher gives me disk errors (maybe due to the high fsb ,since I cannot modify multipliers on this ECS board). Anyway, I have temps well below 50° with the swiftech, in spite of the extremely slow (and silent) Pabst fan I mounted on it. iddle: 39° max load: 46°. overclocking it to 1.5 or 1.6 increases the temps by 1 maybe 2° only.

To test stability; I tried Prime95, Seti and 3D Mark2001 for a few days non stop. Ever since, Seti runs in the background, I play WOlfenstein and try to work with Premiere. NO probs so far. I know its outrageously expensive, but if you can afford it, get swiftech.

---- Owner of the only Dell computer with an AMD chip
October 22, 2001 10:51:54 AM

well i believe its my chip.
i never bothered to check the ID code on it.
it DID come factory unlocked...
but its 6 -7 months old now, and i suspect its an earlier AXIA, or maybe even AVIA...
and they just dont have the overclocking potential of the latter AXIA's or AHRYA's.

seems to run at up to 1360 at 1.85V or so... but anything above really requires the voltage to be pumped to MAXX (1.94V on the iwill mobo for 9.5x150 = 1425)
that runs stable enough...
can do seti or cure for cancer indefinately at 1425, but prime95 doesnt like the high fsb on this beast, and i just found in a warm room and running the hot little utility Toast! cauzes a hard lock.

see i only got a Winfop32-1... its allright 4800rpm 60mm fan... but nothing amazing, and definately struggling at 1.94V.
53C full load, add +8 to get to the old asusboard temp... makes it over 60.. not fun.

and YES i am getting a swiftech 462A as soon as the local shop gets them back in stock...
for overclocking AND the downright nasty auzzie summers without AirCon.
32C+ room temps im lookin at here.


Religious wars are 2 groups of people fighting over who has the best imaginary friend.
October 22, 2001 10:58:46 AM

all you need to do is live in london, it never gets into the 30s here :)  very rarely anyway, my 1800+ XP is running, air cooled (glaciator II very quiet) at 44 degrees under load on a 22 degree day @ 150*11.5 1725 mhz aint too shabby im thinking... need to wait for the middle of winter before hardcore overclocking though... this is at stock voltage by the way

if in doubt blame microsoft...
October 22, 2001 11:38:53 AM

I don't think his temps are the problem. I have the same problem on my PC. I have a GlobalWin WBK38 and a lot of case fans. Maybe if I get some Arctic Silver II I might be able to get my CPU to 1.4GHz without problems.

AMD technology + Intel technology = Intel/AMD Pentathlon IV; the <b>ULTIMATE</b> PC processor
October 28, 2001 9:00:31 PM

lhgpoobaa, I think I know know why you can get to 1.4GHz stabily while I can't!!! You're running an Iwill KK266R that gives 1.95V instead of the usual 1.85V. I'm using an Abit KT7A which is only giving me 1.87V. That would make a difference. Well since I don't need DDR and the Iwill board can get to 1.95V and supports the Athlon XP, then I'll get the Iwill KK266Plus-RAID .

AMD technology + Intel technology = Intel/AMD Pentathlon IV; the <b>ULTIMATE</b> PC processor
October 28, 2001 9:08:28 PM

yes. your correct.
my old asus a7v133 had a voltage difference of +0.05V.
i.e. setting the bios to 1.85V (max) gave a real voltage of 1.89-1.90.

the iwill kk266 has a difference even greater.. 0.07 to 0.09 (given monitoring variances)
so setting it to 1.85 gives me 1.93-1.94v...
course that means heat output goes thru the roof... and my fop32-1 cant handle that.

only a week to go till i get the MCX462 :) 


Is that a Northwoody in your pocket or are you just eXPited to see me?
!