The San Diego OC DISCUSSION Thread

Printable View

Show 100 post(s) from this thread on one page

05-16-2005, 04:37 AM
agenda2005

Quote:

Originally Posted by RAMMIE

Is that GO software the same as mersenne.org Prime95?

Yes! They are exact same thing. The "GO" or normally called Stress Prime is the window interface of mersenne.org Prime95 that allow you to see the clock speed at which you prime and time(how long has elapsed). I will trust this than the usuall Prime95 interface that doesn't provide any such information. You could prime at stock speed and then use clockgen to OC and claim it's stable at the OC speed with the usuall Prime95 interface. The Stress Prime will not even show your clockgen speed, except you set your OC in the BIOS.
05-16-2005, 04:51 AM
eviLRuLeZ

Hi guys,
do you think that 2870 at 1.65 vcore is good for a 3700+ San Diego? It's doing Super pi 32M (Does It imply the cpu to be Stable? I don't know, I don't have cpu in my hands to test it.. :( )
Watercooling I precise. Not box cooling.
05-16-2005, 05:40 AM
Eldonko

Quote:

Originally Posted by eviLRuLeZ

Hi guys,
do you think that 2870 at 1.65 vcore is good for a 3700+ San Diego? It's doing Super pi 32M (Does It imply the cpu to be Stable? I don't know, I don't have cpu in my hands to test it.. :( )
Watercooling I precise. Not box cooling.

Yup that’s pretty good for h20 cooling. Mine does 2800 at 1.65v. SuperPI is a start for a stability test however I wouldn't trust a run or two of 32 to be stable.
05-16-2005, 07:44 AM
Unrealcpu

I just ordered a 3700 SD , I have a few questions

1. Cold bug? using vapochill LS
2. I see people getting 3.1 gighertz or even 3.3 gighertz on this CPU with some GSKILL.
What is the average OC with my setup
thanks
05-16-2005, 09:26 AM
joe2004

I am getting 2.94 GHz on air with 1.75V.
So it is possible to fly high, it all depends on cooling. With average cooling you should expect 2.8 GHz with 1.5-1.6 GHz. Anything better is with a grace of overclocking god.
Stability is a different issue.
05-16-2005, 09:37 AM
stealth17

Quote:

Originally Posted by joe2004

I am getting 2.94 GHz on air with 1.75V.
So it is possible to fly high, it all depends on cooling. With average cooling you should expect 2.8 GHz with 1.5-1.6 GHz. Anything better is with a grace of overclocking god.
Stability is a different issue.

heh yeh i bet you arent 24 hour prime stable at 2.94ghz lol

just amazes me sometimes...hell my max screen is like 3.2 or 3.3...but i dont post that as my overclock lol...

its gatta be atleast pi 1m stable imo to be important unless you are showing and label it as "max screen"
05-16-2005, 10:17 AM
[timko]

Has anyone got the TPAW version rather than the SPMW ones yet?

I've not yet bought an SD-3700 so would like to see if there's much OC difference between the two (although I'm not holding my breath for staggering difference :) )

I've contacted some stores that have confirmed they have the TPAW ones so sourcing one won't be an issue for me.
05-16-2005, 10:35 AM
soul

Even with 1.7 volts, I can't get above 2710mhz prime'able. I MIGHT be able to get some more out of it with more volts, but it's at the point of diminishing returns already. A little disappointing, but not at all surprising.
05-16-2005, 11:18 AM
ben805

Quote:

Originally Posted by soul

Even with 1.7 volts, I can't get above 2710mhz prime'able. I MIGHT be able to get some more out of it with more volts, but it's at the point of diminishing returns already. A little disappointing, but not at all surprising.

More than likely heat is preventing you from getting it higher, check your cpu temperature and if you're running on air, I wouldn't put more than 1.66v for 24/7 use. Dropping the vcore might actually help you stabilize 2.7Ghz, try somewhere around 1.50~1.62v for 2.7Ghz.
05-16-2005, 11:32 AM
agenda2005

2 Attachment(s)

Quote:

Originally Posted by soul

Even with 1.7 volts, I can't get above 2710mhz prime'able.

That's not NICE. My chip will boot at 2700MHz with just 1.4V and run 32M super pi without any hitch. It fail prime at that voltage after sometime, but with 1.5V prime run all day, just have to terminate it after 15Hours.
05-16-2005, 11:35 AM
ewitte

Mine starts needing more vcore at 2.7Ghz (about 1.5v). Its fairly stable at 2.81Ghz but with 1.67v. It ran prime for 1hr18min before I stopped it. I'm running some burn in programs and will try 1.65v prime for hopefully a longer time. If that is successful I'll try 1.625v. Currently prime around 20 minutes at 1.65v.

Eric
05-16-2005, 12:14 PM
RAMMIE

Thanks agenda2005.
Downloaded and running.
05-16-2005, 12:34 PM
soul

Unfortunately, I can't really dismiss user error at this point. Last night I was trying to find the lowest voltage that I need to run ~2700mhz with a 10x and 11x multiplier, and I came back to 1.6v.

Just for the hell of it though, I decided to give 8x337 and 9x300 a try. At 1.5v, both error in prime almost instantly. Being the impatient person that I am, I bumped the voltage up to 1.55v, and tried 8x337 again. It's running happily so far, although it's only been running for under 30 minutes. If it runs for a couple hours, I'll try for a bit more speed at the same voltage, and see what happens.

Temps shouldn't be an issue, since I haven't seen anything over 45C so far. 43C seems to be about a normal load, spiking up to 45C for a second or two every so often.

Edit: 8x337 at 1.55v failed prime after 2 hours 41 minutes.
05-16-2005, 05:29 PM
bubblethumper

Quote:

Originally Posted by Unrealcpu

I just ordered a 3700 SD , I have a few questions

1. Cold bug? using vapochill LS
2. I see people getting 3.1 gighertz or even 3.3 gighertz on this CPU with some GSKILL.
What is the average OC with my setup
thanks

1. Sorta. I've heard stories and I have an SD3700 with cold bug. There is a good chance my Sandy is generally defective though. So my cold bug experience might be due to a defective processor.
05-16-2005, 07:41 PM
cpulloverclock

Quote:

Originally Posted by formyfaith

Been playing around with my sandy,

and I think I can push this baby a bit more. :D

Cooling : XP-90C with 92mm 38T Vantec Tornado at 5100RPM.

45°C at 1.71v on air?

I havent that at 1.4v stock :D

2.8GHz 1.4v = 50-51°C
05-16-2005, 07:54 PM
Eldonko

GJ Dread, you don't see many posting 25 hour Prime stable OCs.
05-16-2005, 07:57 PM
stelleg151

Quote:

Originally Posted by formyfaith

Been playing around with my sandy,

and I think I can push this baby a bit more. :D

Cooling : XP-90C with 92mm 38T Vantec Tornado at 5100RPM.

Youre my hero.
05-17-2005, 02:13 AM
Mrki

Quote:

Originally Posted by formyfaith

Been playing around with my sandy,

and I think I can push this baby a bit more. :D

Cooling : XP-90C with 92mm 38T Vantec Tornado at 5100RPM.

awesome!!! :slobber:

stepping?
05-17-2005, 05:21 AM
joe2004

Memory is the main limitation, I mean memory controller.
If I run my Winbond BH-5 at 220-230 2-2-2-5-1 then I get to 2.92-4 GHz with this SD 3700. But ... if I run it at 260 at 2-2-2-5-1 then I get max to about 2.85-86 GHz 32 MB SuperPi stable. This is all on air.
As if there was some magic ceiling for the bandwidth. But of course this is much better than Winchester chips.
05-17-2005, 05:25 AM
Rukee

I`m at about the same place with my SD3700 and OCZ PC3200 EL Gold kit.
257MHz CAS2-2-2-6 1T @ 2.6v DFI NF4 Ultra SuperPi and MemTest86 stabile.
05-17-2005, 06:05 AM
jikdoc

Quote:

Originally Posted by joe2004

Memory is the main limitation, I mean memory controller.
If I run my Winbond BH-5 at 220-230 2-2-2-5-1 then I get to 2.92-4 GHz with this SD 3700. But ... if I run it at 260 at 2-2-2-5-1 then I get max to about 2.85-86 GHz 32 MB SuperPi stable. This is all on air.
As if there was some magic ceiling for the bandwidth. But of course this is much better than Winchester chips.

i've noticed the same thing with utt. i've also noticed that the rig gets unstable when htt is ~300. as to whether it's cpu, mobo or memory, i'm pretty sure it's the dfi. here's why:

SD3700, twinmos utt 5:6 divider (runs at 228 when fsb is 290). this memory is stable 2226 timings @ 250 3.2v. when i put on the divider and run the memory slower, i crash when i near 300.

SD3700, gskill tccd, 1:1 divider. completely and abruptly unstable at 300htt.

i have tried changing volts on ldt, chipset without luck, just can't seem to get >300ht stable. and yes, ldt is set to 3x.
05-17-2005, 06:44 AM
formyfaith

Quote:

Originally Posted by Mrki

awesome!!! :slobber:

stepping?

It's a 0516 TPAW. ;)
05-17-2005, 07:40 AM
agenda2005

Quote:

Originally Posted by cpulloverclock

45°C at 1.71v on air?

I havent that at 1.4v stock :D

2.8GHz 1.4v = 50-51°C

You might wanna factor in the ambient temps, high speed rev fan 5100 RPM WOW! or could be a bad MB sensor altogether.
05-17-2005, 07:42 AM
agenda2005

DOES anyone here living in Atlanta knows whether Monach allows you to hand pick a CPU if you walk into the store?
05-17-2005, 07:46 AM
charlie

Quote:

Originally Posted by agenda2005

DOES anyone here living in Atlanta knows whether Monach allows you to hand pick a CPU if you walk into the store?

Most stores will, call Monarch and ask. The OEM CPU's come to them in these plastic trays, I wouldn't see why they wouldn't let you pick from the tray. That's what my local store does.

C
05-17-2005, 07:56 AM
mrlobber

Quote:

Originally Posted by Rukee

257MHz CAS2-2-2-6 1T @ 2.6v DFI NF4 Ultra SuperPi and MemTest86 stabile.

257 Mhz 2-2-2 and 2.6v? :slobber: :eek:
05-17-2005, 08:52 AM
stealth17

how good are the memcontrollers on these SD cores? i should beable to run my TCCD at 295 or 300 for 300*9.5, 300*9, or so correct?
05-17-2005, 08:54 AM
easypanic

I'm fully stable @ 300*9 with my TCCD sticks ;).
Make sure your memory can handle it...

Don't try the *9.5 thingie, x.5 multipliers are not done with A64 :).
05-17-2005, 09:10 AM
ewitte

Quote:

Originally Posted by charlie

Most stores will, call Monarch and ask. The OEM CPU's come to them in these plastic trays, I wouldn't see why they wouldn't let you pick from the tray. That's what my local store does.

C

Can you bring in a system and try each one first :p:
05-17-2005, 09:10 AM
s7e9h3n

Quote:

Originally Posted by easypanic

I'm fully stable @ 300*9 with my TCCD sticks ;).
Make sure your memory can handle it...

Don't try the *9.5 thingie, x.5 multipliers are not done with A64 :).

Actually, they're done, just not correctly ;)
05-17-2005, 09:18 AM
Soulburner

My 3700+ will do 2500 at 1.45v but for the same stability at 2600 it needs 1.6v??

Does that seem right?

Yes my cooling is bad, its temporary for a while until it goes on water, but still...
05-17-2005, 09:38 AM
stealth17

Quote:

Originally Posted by easypanic

I'm fully stable @ 300*9 with my TCCD sticks ;).
Make sure your memory can handle it...

Don't try the *9.5 thingie, x.5 multipliers are not done with A64 :).

so for ocz rev 2 platinum should i run orange or yellow slots? im flashing to -2 bios now as that supposed to be best for my case, but which slots have been working best?
05-17-2005, 09:57 AM
agenda2005

Quote:

Originally Posted by charlie

Most stores will, call Monarch and ask. The OEM CPU's come to them in these plastic trays, I wouldn't see why they wouldn't let you pick from the tray. That's what my local store does.

C

Thanks. I will be in Atlanta in couple of days and just want to try my luck with week 16 SD CPU's they seems to OC better than others.
05-17-2005, 10:00 AM
wickedld9

Quote:

Originally Posted by jikdoc

i've noticed the same thing with utt. i've also noticed that the rig gets unstable when htt is ~300. as to whether it's cpu, mobo or memory, i'm pretty sure it's the dfi. here's why:

SD3700, twinmos utt 5:6 divider (runs at 228 when fsb is 290). this memory is stable 2226 timings @ 250 3.2v. when i put on the divider and run the memory slower, i crash when i near 300.

SD3700, gskill tccd, 1:1 divider. completely and abruptly unstable at 300htt.

i have tried changing volts on ldt, chipset without luck, just can't seem to get >300ht stable. and yes, ldt is set to 3x.

I've noticed something similar but it looks like my CPU is a little more forgiving.....or I'm tweaking it just enough. ;) With Mushkin TCCD I can run it at 311*9 2.5-4-3-7 with a 166 divider (my tccd's give up at ~275 regardless of timings). I've noticed the looser the timings the less vcore required to run a given CPU speed. I can get her to boot 311*9 at 1.45 but it fails pretty much everything at that voltage.
With my UTT's it's kind of the same story, the bigger the divider/looser timings the better the cpu scales.
05-17-2005, 11:22 AM
s7e9h3n

Quote:

Originally Posted by wickedld9

I've noticed something similar but it looks like my CPU is a little more forgiving.....or I'm tweaking it just enough. ;) With Mushkin TCCD I can run it at 311*9 2.5-4-3-7 with a 166 divider (my tccd's give up at ~275 regardless of timings). I've noticed the looser the timings the less vcore required to run a given CPU speed. I can get her to boot 311*9 at 1.45 but it fails pretty much everything at that voltage.
With my UTT's it's kind of the same story, the bigger the divider/looser timings the better the cpu scales.

It's cuz your cpu is generating less heat due to the onboard memory controller ;)
05-17-2005, 04:25 PM
pershoot

Quote:

Originally Posted by jikdoc

i've noticed the same thing with utt. i've also noticed that the rig gets unstable when htt is ~300. as to whether it's cpu, mobo or memory, i'm pretty sure it's the dfi. here's why:

SD3700, twinmos utt 5:6 divider (runs at 228 when fsb is 290). this memory is stable 2226 timings @ 250 3.2v. when i put on the divider and run the memory slower, i crash when i near 300.

SD3700, gskill tccd, 1:1 divider. completely and abruptly unstable at 300htt.

i have tried changing volts on ldt, chipset without luck, just can't seem to get >300ht stable. and yes, ldt is set to 3x.

try Max Async to 9x (10x will crash you out round 300-305 and 8x lower then that) and Read Preamble to 6.5/5.5 & LDT: 2.5x/2x >300-305HTT.
05-17-2005, 04:54 PM
jikdoc

Quote:

Originally Posted by pershoot

try Max Async to 9x (10x will crash you out round 300-305 and 8x lower then that) and Read Preamble to 6.5/5.5 & LDT: 2.5x/2x >300-305HTT.

thanks! i'll try this! i've been using 3x on the assumption that anything under 1000HTT was fine. this would explain alot, including why i always thought i had more in my fx55.
05-17-2005, 07:57 PM
uwackme

hmmm

Im having no troubles with 4x 512M sticks of UTT/VX at 250Mhz x11, 3.4Vdimm 1.58Vcore.... but this is 2T. 1T wont boot with 4 sticks of 512M.

$100/1G dirt cheap fantastic performance.

3700+SD and X850XT from Monarch.... my new fav place.

I lost about 500 pts in 3D2K1...down from 30,400 to 29,900 with the switch to 2T.

I think I'll live :toast: :woot: :clap:
05-17-2005, 10:00 PM
Soulburner

Wow...cooling??
05-17-2005, 11:49 PM
t-max

If you have stability problem with your tccd chips try to increase the Trtw from 2 to 3 that helped me a lot from getting tousind of erros in test 5 in memtest 86 at 320 mhz to run 10 pass with no erros at all.
05-18-2005, 07:29 AM
Unrealcpu

I noticed everyone has a 0515 VPMW 3700+ , i just received a 0512 XPMW and will not be testing until tomorrow on my vapochill LS. What is the difference anyone .
Which stepping is the best
05-18-2005, 07:59 AM
Soulburner

Not sure, I have a CABGE 0516WPAW...
05-18-2005, 08:04 AM
Unrealcpu

whatever it is im sure ill be around 3.2 gigherz haaha
05-18-2005, 08:10 AM
uwackme

hmmm

Soul.... I have a 80mm fan temp blowing on the DIMMs. Still mod'ing a second case to rig up 120mm fans for motherboard cooling. So the door is off with a 80mm fan on the ram, and stock retail AMD HSF on cpu. Though I used AS3 on CPU and replaced the NB chip goop with AS3 as well. At 2750 under load it max's in mid 50C's with 1.58Vcore, 3.4Vdimm (using 5V jumper position).

X850XT is bone stock...520/540, I need to pull it apart and replace goop with AS3/AS Ceramique before going down the OC road with the video. I'd have to say the stock HSF on X850 is a bit on the loud ide when it cranks. Im using ATI tool to keep the temps peak at 60C.

Have to mention, RADCLOCKER doesnt know what the FK an X850 is and wont run. Its awesome on my X800 but the AGP and PCIe X850's jut confuse it. So Im using ATItool instead.

I've been screen snapping all the tests, Ill photoshop up a montage...all of it wouldnt fit readable on the desktop. The SATA raid0 with 2x Raptors is a SCREAM too!!!

The OCZ Powerstream is working great, super stable voltages, much more stable than my Infinity system. The voltage regulators on the DFI NF4 are truly excellent designs, power is rock stable.
05-18-2005, 08:48 AM
XTCgr

Quote:

Originally Posted by Unrealcpu

I noticed everyone has a 0515 VPMW 3700+ , i just received a 0512 XPMW and will not be testing until tomorrow on my vapochill LS. What is the difference anyone .
Which stepping is the best

not everyone, i got one CABGE 0512 XPMW but not satisfied. Just 2.82ghz on water s&m stable and with 1.65vcore set in bios (1.5*110%)
05-18-2005, 09:45 AM
dexth77

Hi,
Does anybody know if the vcore more than 1.45 works? I have winchester bug which doesnt set vcore >1.45v properly, ie. if I set 1.5v or more, bios/mbm/cpuz will read ~1.4v. I can set 1.45v and add % to get higher but seems like my SD 3700+ is also having this issue. Is this a board problem then? I have K8N neo2 pt.

sincerely,
dexth77
05-18-2005, 10:27 AM
Eldonko

Sounds like a board/BIOS prob to me.
05-18-2005, 11:37 AM
dexth77

Quote:

Originally Posted by Eldonko

Sounds like a board/BIOS prob to me.

Quite some folks with later ccbid winchester had this issue from what i read. It's a board issue and not the processor then?
05-18-2005, 02:11 PM
sideeffect

Quote:

Originally Posted by dexth77

Quite some folks with later ccbid winchester had this issue from what i read. It's a board issue and not the processor then?

It does the same for me with the neo 2 platinum. I think its because the older athlon 64's ran with 0.10 v more voltage and 1.55 was the max so now 1.45 becomes the max setting in bios. If you open corecenter in windows then it reads vcore as 0.1v more than it is set at with the san diegos.

I dont think it matters though as these chips seem to hate voltage, anything above 1.7 volts makes my san diego less stable, thats at -20 degrees so its not heat related :)

Also I have a cold bug glitch with the k8n neo 2 i think. When i go to look in bios at temperatures it reads cpu at 265 degrees and freezes. But if i dont go into that menu everything else works fine and it doesnt effect anything else. Also if i let the vapochill get to cold the cpu diode reading that i monitor with speedfan stops working. Doesnt crash or anything though. Keeping vapochill at -20 works best for me.

3700+ (2866mhz) @ 1.5v
2x 512 ballistix dual channel 1t 260 fsb 2.5-3-3-10 (work in progress)
msi neo 2 platinum 1.8 bios
vapochill PE
05-18-2005, 03:38 PM
dexth77

Quote:

Originally Posted by sideeffect

It does the same for me with the neo 2 platinum. I think its because the older athlon 64's ran with 0.10 v more voltage and 1.55 was the max so now 1.45 becomes the max setting in bios. If you open corecenter in windows then it reads vcore as 0.1v more than it is set at with the san diegos.

Well, my voltage reading is kind of off. Here's what I got
1.45+10%=1.595 (mbm reads 1.61-1.63)
1.45+13%=1.6385 (mbm reads 1.64-1.66)

Thanks for your input though. :toast:
05-18-2005, 03:54 PM
s7e9h3n

Quote:

Originally Posted by dexth77

Hi,
Does anybody know if the vcore more than 1.45 works? I have winchester bug which doesnt set vcore >1.45v properly, ie. if I set 1.5v or more, bios/mbm/cpuz will read ~1.4v. I can set 1.45v and add % to get higher but seems like my SD 3700+ is also having this issue. Is this a board problem then? I have K8N neo2 pt.

sincerely,
dexth77

No, its not a problem your board - all neo2 plats are like that. The highest "real" vcore you'll acheive through bios is 1.45 (plus and additional %). The only way you can go higher than the base 1.45Vcore is to increase voltage through CoreCenter.
05-18-2005, 04:07 PM
stealth17

Quote:

Originally Posted by s7e9h3n

No, its not a problem your board - all neo2 plats are like that. The highest "real" vcore you'll acheive through bios is 1.45 (plus and additional %). The only way you can go higher than the base 1.45Vcore is to increase voltage through CoreCenter.

wow that would piss me off! you should get a DFI nF4 :p:
05-18-2005, 04:55 PM
dexth77

Quote:

Originally Posted by s7e9h3n

No, its not a problem your board - all neo2 plats are like that. The highest "real" vcore you'll acheive through bios is 1.45 (plus and additional %). The only way you can go higher than the base 1.45Vcore is to increase voltage through CoreCenter.

Why do they include the 1.5, 1.55, etc selection then? :confused:
05-18-2005, 05:11 PM
stealth17

anyone have a pic of their naked SD?

adn maybe a before and after compare?

thanks
05-18-2005, 05:21 PM
s7e9h3n

Quote:

Originally Posted by dexth77

Why do they include the 1.5, 1.55, etc selection then? :confused:

Have no idea, but trust me, they don't work..... ;)
05-18-2005, 07:40 PM
Boogotop

just got a 3700+ San Diego: ADA3700DAA5BN
CABGE 0517UPAW

i'll try some aircooled action and after i'll try with my Mach 2 GT
05-18-2005, 11:05 PM
Soulburner

Quote:

Originally Posted by s7e9h3n

Have no idea, but trust me, they don't work..... ;)

They work for me?

CPU-Z shows the voltage correctly does it not?

They are a little lower than you would expect but they "work"...just use the +% to make it even out...
05-19-2005, 02:42 AM
Unrealcpu

Its funny i just ordered from monarch and i get a 0512 XPMW and everyone else has a 0515 or newer, I think i got screwed with old warez
05-19-2005, 02:46 AM
Jessfm

Quote:

Originally Posted by stealth17

anyone have a pic of their naked SD?

adn maybe a before and after compare?

thanks

http://www.ricco.karoo.net/images/cpu/3700sandy.jpg

Sat at 2825 with IHS on, stuck at 2950 with it off (watercooled)
05-19-2005, 06:54 AM
Unrealcpu

Does anyone have a XPMW 3700 SD?
0512

I feel left out because i dont have a 0515 lol
05-19-2005, 07:20 AM
agenda2005

Quote:

Originally Posted by Unrealcpu

Does anyone have a XPMW 3700 SD?
0512

I feel left out because i dont have a 0515 lol

Post your result and let us see.
05-19-2005, 07:28 AM
XTCgr

3700+ CABGE 0512 XPMW here.
2.82Ghz max s&m stable 15*110% vcore in bios
05-19-2005, 08:46 AM
mrlobber

Hmm, not much 0516 around, and that's exactly what I have from Monarch coming... interesting.
05-19-2005, 09:00 AM
s7e9h3n

Quote:

Originally Posted by Soulburner

They work for me?

CPU-Z shows the voltage correctly does it not?

They are a little lower than you would expect but they "work"...just use the +% to make it even out...

Maybe things have changed with newer revisions of the board, but for me the options for over 1.45 are available, but they did nothing in terms of pushing the Vcore over 1.45. Try this: Set your vcore to 1.45 and don't add any % and see what cpuz says. Then try higher and see if there's any difference in what's reported.
05-19-2005, 09:17 AM
Unrealcpu

i have the worst cpu 3700+ 0512 XPMW SD .

Cant do past 2.55 gighertz on my DFI
using Redline 4000+ and GSKILL PC4400 LE

VAPOCHILL LS

cough cough back in the box in a few weeks, putting back fx55
05-19-2005, 10:28 AM
uwackme

hmmm

New SuperSystem: Courtesy of Monarch

3700+ SD cabbage, 4x 512M sticks Twinmos TM400II UTT "AA4T", DFI Ultra converted to SLI-D, 5V memory jumper, OCZ 520 Powerstream, 2x 36G raptor raid0, X850XT, mod'd AMS CF1009 case, NEC 3520A 16xDVD burner.

Stock HSF on SD, AS3 used for all. X850XT at stock clocks. SD at 250Mhz x11, 250Mhz HTT (x3), 1.250 + 126% (1.50)Vcore, Vdimm @ 3.4V, 1.4Vldt, 1.9Vchipset.

2Gigs @ 250Mhz 1:1 2T 8,2,2,2.0, bankinterleave ON.

PCpitstop... 2233 (my Infinity is 1806)

32M Superpi mod 1.4 26m54.156sec... temp peaks 49C stock HSF

3D2K3 13075 (cpu 1356)
3D2K5 5931 (cpu 5611)... GPU Mhz dependent, Infinity sys gets 6300+
3D2K1 30159

HDTach 218MB/s burst, 116MB/s avg, 8.0ms access, 10% CPU utilization
Atto 0.5 10180write 27469read
Atto 1024 127522write 107440read

SSoft CPU 12755/5638
SSoftMEM 6004/6014.... 8000 theoretical max, 2T hurts here 75% rating
Everest 2.0 6635 read, 2300 write, 41.9ns lat

The new memory controller is working like a charm on 4 sticks. Strikes me X850XT's HSF should do better job, maybe I need to lap GPU mating surface, just AS3 still hits 72C GPU peak.

I am seeing a coldboot quirkyness. Warm boot its fine, rock solid, but unplug power, come back in a while, boot takes some trying. Maybe this needs tweaking of the startup Vcore? Any ideas?
05-19-2005, 10:56 AM
dexth77

Quote:

Originally Posted by s7e9h3n

Maybe things have changed with newer revisions of the board, but for me the options for over 1.45 are available, but they did nothing in terms of pushing the Vcore over 1.45. Try this: Set your vcore to 1.45 and don't add any % and see what cpuz says. Then try higher and see if there's any difference in what's reported.

That's what i'm saying, works for some, doesnt work for some :confused:
I'm guessing it's board issue...since 2 diff cpus doesnt work on mine.
05-19-2005, 12:13 PM
agenda2005

Quote:

Originally Posted by Unrealcpu

i have the worst cpu 3700+ 0512 XPMW SD .

Cant do past 2.55 gighertz on my DFI
using Redline 4000+ and GSKILL PC4400 LE

VAPOCHILL LS

cough cough back in the box in a few weeks, putting back fx55

Relax, check your bios setting again. Somthing is probably holding you back. PSU? RAM? BIOS Setting? 2.6GHz should be doable.
05-19-2005, 12:15 PM
Unrealcpu

Quote:

Originally Posted by agenda2005

Relax, check your bios setting again. Somthing is probably holding you back. PSU? RAM? BIOS Setting? 2.6GHz should be doable.

2.6 is not doable.

I had my fx55 at 3300mhz easy. Some reason its not likeing voltage or is speed binned. The 11x multi does not work at all past 250 fsb

i can barely get 9x270

Ok seems to be ok at 2.65 gighertz

this processor is lame on VAPO
05-19-2005, 02:29 PM
stealth17

Quote:

Originally Posted by Unrealcpu

2.6 is not doable.

I had my fx55 at 3300mhz easy. Some reason its not likeing voltage or is speed binned. The 11x multi does not work at all past 250 fsb

i can barely get 9x270

Ok seems to be ok at 2.65 gighertz

this processor is lame on VAPO

sell it or rma it and get a week 16...they seem to be doing well. never seen any 12s :confused:
05-19-2005, 02:30 PM
s7e9h3n

Quote:

Originally Posted by stealth17

sell it or rma it and get a week 16...they seem to be doing well. never seen any 12s :confused:

Don't RMA it if it does stock speeds.
05-19-2005, 04:32 PM
Unrealcpu

Thanks Monarcrap
05-20-2005, 08:52 PM
chefnr1

Quote:

Originally Posted by BMORIN

How do you like this for a 24/7 stable superpi time?

very nice :toast: my mach II krasch yesterday so i have to run with water untill i'll have it fixed :mad:
05-22-2005, 12:46 AM
Der_KHAN

what cooling maxxxpct ?
05-22-2005, 01:44 AM
El Snorro

Nice to see San diego's dont have cold bug :)
I'm gonna buy myself a prometiea mach1 @r507 and see what my cpu (0516 tpaw) can do :toast:
05-22-2005, 02:07 AM
maxxxpct

ours it has problems under -45°c
05-22-2005, 07:29 AM
BMORIN

Peter Lu, that's one of the best I've seen. 32mb stable @ 1.5v. You have a gem there. I'd like to see that under some freezing cold.
05-22-2005, 11:40 AM
Nedjo

ADA3700DAA5BN
CABHE0505FPMW
2.8 GHz (10x280) @ 1.31V
DFI LP nF4 Ultra (515-3 BIOS)
Corsair PC4400C25 (2x512MB)
Artic Cooling Freezer64@ 2200rpm
Prime stable /w temp of 40 C
05-22-2005, 01:17 PM
Mari0-Br0s

Quote:

Originally Posted by Saftlad

Stable

http://www.theqbs.com/swampy/prime.jpg

Jesus Christ!!! It's run at 3.0GHz at stock voltage, how the hell did he managed to get this!!??
05-22-2005, 01:22 PM
eva2000

Quote:

Originally Posted by Mari0-Br0s

Jesus Christ!!! It's run at 3.0GHz at stock voltage, how the hell did he managed to get this!!??

easy he's at -16C on cpu heh

sandy clock for clock seems really fast - i broke 26min at 32M Superpi at only 2772mhz :)

http://www.fileshosts.com/DFI/NF4_SL...rpi-32m_tn.jpg
05-22-2005, 01:29 PM
Mari0-Br0s

yeah but it's not suppost to become hotter if you do not change the voltage, water cooled or not, it should still be possible, am i right?
05-22-2005, 02:22 PM
fagin

1 Attachment(s)

Suicide shot!

On stock HSF, 1.664volts.
Stepping CABGE 0515SPMW

Validation
05-22-2005, 02:29 PM
Rukee

I`m having great luck with mine. It`ll run all the 3D benchs at 291fsb for 3.2ghz but won`t pass superpi.
Vapo LS
DFI NF4 Ultra-D
OCZ PC-5000
OCZ Powerstream 520
DC 6600 GT
http://img.photobucket.com/albums/v675/Rukee/oc.jpg
05-22-2005, 04:21 PM
Soulburner

Mine acts similarly where there are settings where it will game at stabily but fail Super Pi 1M.
05-22-2005, 08:54 PM
fareastgq

anyone getting their retail san deigo from newegg? I can't decide whether to get it from newegg or monarch. I want to get it from which ever will be the best chance, heh.
05-22-2005, 09:22 PM
Gorod

How does 0516 or newer SanDiego's clock compared to older ones ? I think i got a crapy one (CABGE 0515SPMW) , it will do only 3122-3150 Max. stable with my Vapo LS at highest voltage my Neo2 can give (~ 1.76V ) , same as my Venice 3000+ did :slap: . Do u guys think i shoud try a newer stepping ? I was hoping to get at least 3200MHz stable and now i am liltle dissapointed :(
05-23-2005, 02:50 AM
Unrealcpu

Quote:

Originally Posted by Gorod

How does 0516 or newer SanDiego's clock compared to older ones ? I think i got a crapy one (CABGE 0515SPMW) , it will do only 3122-3150 Max. stable with my Vapo LS at highest voltage my Neo2 can give (~ 1.76V ) , same as my Venice 3000+ did :slap: . Do u guys think i shoud try a newer stepping ? I was hoping to get at least 3200MHz stable and now i am liltle dissapointed :(

i cant even get past 2700mhz with my 3700+SD

and im on vapo

its a 0512 xpmw
05-23-2005, 05:04 AM
Major

Sorry guys I didn't read through the whole thread to see if this has been discussed yet.

I switched over to some UTT with my SD 3700+ this weekend, was using G.Skill LE. The max prime stable OC with TCCD was 2808mhz 312x9, with the UTT I'm able to do 2937mhz 267x11. Also the 10x multi was flaky with the TCCD, with the UTT no issues with it. I know this isn't exactly a scientific conclusion and this all could boil down to BIOS related issues, but I have a feeling that the new cores like LL memory better.
Just a hunch.

Has anyone else experienced the same thing ?
05-23-2005, 05:38 AM
SVTSnake

Quote:

Originally Posted by Unrealcpu

i cant even get past 2700mhz with my 3700+SD

and im on vapo

its a 0512 xpmw

Don't feel bad Unreal! I got the same stepping and week 3700+ as you did from Monarch and it was crap as well :(

Sold it off on eBay over the weekend. Mine on water would barely do 2600.

Think I may try another one from Newegg this time. Seems everytime I get a processor from Monarch, its either old stock or crap...
05-23-2005, 06:05 AM
ewitte

Quote:

Originally Posted by Unrealcpu

i cant even get past 2700mhz with my 3700+SD

and im on vapo

its a 0512 xpmw

:eek: I'm only on water but it was a PITA to get it stable at 2.81Ghz. It has to be at least 1.67v.

Eric
05-23-2005, 06:14 AM
soul

I'm in the same boat with my 0512XPMW. 2710mhz seems to be my limit for anything even remotely stable.
05-23-2005, 06:17 AM
chefnr1

Quote:

Originally Posted by Unrealcpu

i cant even get past 2700mhz with my 3700+SD

and im on vapo

its a 0512 xpmw

i have the same stepping this link is where i use the boxed cooler, cant use my mach II because it krashed in the komp...
on water i have it up in 2988mhz but i just wait till my new prommy will arrive...

http://www.xtremesystems.org/forums/...9&postcount=83
05-23-2005, 06:43 AM
pershoot

all luck of the draw. def. seeing a trend where difference of boards and also memories factor in, and not just singled out on cpu.

for example:
i hit just shy of 2.8 (2.77-2.79 (water)), with ~1.64, while i know someone else with my same exact stepping is doing that with ~1.52-55. i know someone else who has a retail, same week as mine, who can do ~2.9+ (water).

and also, this thing is crazy sensitive about timings. if one is off, the whole thing wont even have a chance. and its such a very long process to get anywhere remotely stable (specially when you use prime! haha).
05-23-2005, 06:43 AM
mongoled

Quote:

Originally Posted by Major Slaughter

Sorry guys I didn't read through the whole thread to see if this has been discussed yet.

I switched over to some UTT with my SD 3700+ this weekend, was using G.Skill LE. The max prime stable OC with TCCD was 2808mhz 312x9, with the UTT I'm able to do 2937mhz 267x11. Also the 10x multi was flaky with the TCCD, with the UTT no issues with it. I know this isn't exactly a scientific conclusion and this all could boil down to BIOS related issues, but I have a feeling that the new cores like LL memory better.
Just a hunch.

Has anyone else experienced the same thing ?

Im not having exactly the same problems as yourself, nor do I have a SD (i have a Venice), but i am suspecting tht there is some sort of divider/htt problem with these new CPU's. Unfortunately my efforts in trying to gain some information from various online communities has not resulted in any positive results

:(

Hopefully a bios will come soon which will help resolve these problems, im not holding my breath, seeing as the 0510 bios's just released have done nothing to help eleviate the problems a few peeps such as yourself are currently having
05-23-2005, 06:49 AM
zxlr8

I am getting 2800 on stock volts and stock heatsink. CABGE0515spmw. I am running 560 on the x800xt and I have only gotten 30405 in 3d01 so far...
05-23-2005, 06:57 AM
pershoot

Quote:

Originally Posted by zxlr8

I am getting 2800 on stock volts and stock heatsink. CABGE0515spmw. I am running 560 on the x800xt and I have only gotten 30405 in 3d01 so far...

we have the same exact stepping. i can do ~2.55-2.6 on stock volts
05-23-2005, 07:18 AM
BMORIN

Quote:

Originally Posted by pershoot

and also, this thing is crazy sensitive about timings. if one is off, the whole thing wont even have a chance. and its such a very long process to get anywhere remotely stable (specially when you use prime! haha).

Listen to that advice. It is VERY true. You all need to make sure that you find the max for your memory BEFORE you go looking to find the max for your cpu. As you find the max for your memory, you need to really mess with a64 tweaker. As pershoot said, one timing can be off and throw the whole thing unstable. And make sure you use the same timings whether you are 1:1 or using a divider because dropping the divider will change memory timings in many cases.
05-23-2005, 07:32 AM
zxlr8

I am running my memory pretty conservatively. I think that is some of my biggest help. My problem right now is getting raid 0 to work on my mobo. (MSI k8n Neo 2)
05-23-2005, 07:43 AM
eva2000

Quote:

Originally Posted by Major Slaughter

Sorry guys I didn't read through the whole thread to see if this has been discussed yet.

I switched over to some UTT with my SD 3700+ this weekend, was using G.Skill LE. The max prime stable OC with TCCD was 2808mhz 312x9, with the UTT I'm able to do 2937mhz 267x11. Also the 10x multi was flaky with the TCCD, with the UTT no issues with it. I know this isn't exactly a scientific conclusion and this all could boil down to BIOS related issues, but I have a feeling that the new cores like LL memory better.
Just a hunch.

Has anyone else experienced the same thing ?

similar thing but with 414-3 bios my TCCD allows my cpu to clock higher than with UTT/BH-5 ram. Some folks says 510-2 is better for UTT/BH-5 so i'm flashing to 510-2 tomorrow to see what's up.

still sandies love UTT BH-5 - 2x 512MB OCZ PC3200 Gold

http://www.fileshosts.com/DFI/NF4_SL...rpi-32m_tn.jpg

@268mhz 2-2-2-6
http://www.fileshosts.com/DFI/NF4_SL...rpi-32m_tn.jpg
05-23-2005, 08:08 AM
pershoot

Quote:

Originally Posted by eva2000

similar thing but with 414-3 bios my TCCD allows my cpu to clock higher than with UTT/BH-5 ram. Some folks says 510-2 is better for UTT/BH-5 so i'm flashing to 510-2 tomorrow to see what's up.

still sandies love UTT BH-5 - 2x 512MB OCZ PC3200 Gold

@268mhz 2-2-2-6

i will be picking up some UTT BH to test this theory out.

can you all give me ram brands (and model #'s if possible), which contain UTT BH?

Show 100 post(s) from this thread on one page

All times are GMT -8. The time now is 10:17 PM.

XtremeSystems