If this is your first visit, be sure to check out the FAQ by clicking the link above. You may have to register before you can post: click the register link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below. |
|
|
|
Thread Tools | Display Modes |
#1
|
|||
|
|||
P4 2.8C OCing Results??
Hi Troops,
I've got my 2.8C up to 3360MHz but it's not quite stable - reboots after some intense flying in my sim. Intel 2.8C, SL6WT, Zalman 7000A, ABIT IC7-Max3, ATI 9800 Pro w/Zalman VGA cooler + fan, 1G OCZ Deluxe Gold PC4000. Lian-Li full tower case (PC-75) with 6 fans (2-front intake, 2- lower rear exhaust, 2-upper rear exhaust), PCP&C 510 P/S. Settings: NB Strap - By CPU Dram Ratio - 1:1 AGP Ratio - Fixed Fixed AGP/PCI Freq - 66/33 CPU Core V - 1.725v (Too high? Temp = 48° idle, 54-56° load) DDR SDRAM V - 2.7v AGP V - 1.55v DRAM Timings - Manual 3,8,4,4 AGP - 8x 3DMark2003 = 5873 (good/bad/mediocre?) I checked Overclockers.com and my CPU voltage is already higher than 99% of the DB responders. Think I should raise my DDR V to 2.8? 3.4GHz is OK if it's stable but I was hoping for 3.5 :-( -- Tally Ho! Ed Forsythe Maryland, USA |
#2
|
|||
|
|||
"Ed Forsythe" wrote Think I should raise my DDR V to 2.8? 3.4GHz is OK if it's stable but I was hoping for 3.5 :-( Hi Ed, good to hear you got your rig together. I have been reading that you should first increase vDimm *before* you increase vCore. -- Wayne ][ |
#3
|
|||
|
|||
OK Wayne, I'll back off to 2.8GHz then increase DDR SDRAM (vDiimm?) to 2.8.
Boot and fiddle then increase vCore to 1.7 shoot for 3.4 again and see what happens. Still the troops on Overclockers.com are getting excellent results with voltages (vCore) in the 1.6 range. Or maybe I'll just be satisfied at 3.2GHz. Frankly I can't see any real difference between 3.4 and 3.2. When I was using a sim at 3.4 MBM indicated my CPU temp was 61°C - that's a bit too high for me. Thanks for the tip -- Tally Ho! Ed Forsythe Maryland, USA "Wayne Youngman" wrote in message ... "Ed Forsythe" wrote Think I should raise my DDR V to 2.8? 3.4GHz is OK if it's stable but I was hoping for 3.5 :-( Hi Ed, good to hear you got your rig together. I have been reading that you should first increase vDimm *before* you increase vCore. -- Wayne ][ |
#4
|
|||
|
|||
Ed Forsythe wrote:
OK Wayne, I'll back off to 2.8GHz then increase DDR SDRAM (vDiimm?) to 2.8. Boot and fiddle then increase vCore to 1.7 shoot for 3.4 again and see what happens. Still the troops on Overclockers.com are getting excellent results with voltages (vCore) in the 1.6 range. Or maybe I'll just be satisfied at 3.2GHz. Frankly I can't see any real difference between 3.4 and 3.2. When I was using a sim at 3.4 MBM indicated my CPU temp was 61°C - that's a bit too high for me. Thanks for the tip How'd it go, Ed? Inquiring minds want to know... they also want to go FASTER! -Pham |
#5
|
|||
|
|||
Hi Pham,
I'm still at 2.8 contemplating another try at 3.2 or 3.4. Maybe I'm just a candy-a** but the high temps really bothered me. When I have time/guts for another go I'll post -- Tally Ho! Ed Forsythe Maryland, USA "Pham" wrote in message s.com... Ed Forsythe wrote: OK Wayne, I'll back off to 2.8GHz then increase DDR SDRAM (vDiimm?) to 2.8. Boot and fiddle then increase vCore to 1.7 shoot for 3.4 again and see what happens. Still the troops on Overclockers.com are getting excellent results with voltages (vCore) in the 1.6 range. Or maybe I'll just be satisfied at 3.2GHz. Frankly I can't see any real difference between 3.4 and 3.2. When I was using a sim at 3.4 MBM indicated my CPU temp was 61°C - that's a bit too high for me. Thanks for the tip How'd it go, Ed? Inquiring minds want to know... they also want to go FASTER! -Pham |
#6
|
|||
|
|||
"Ed Forsythe" wrote in message ... Hi Troops, I've got my 2.8C up to 3360MHz but it's not quite stable - reboots after some intense flying in my sim. Intel 2.8C, SL6WT, Zalman 7000A, ABIT IC7-Max3, ATI 9800 Pro w/Zalman VGA cooler + fan, 1G OCZ Deluxe Gold PC4000. Lian-Li full tower case (PC-75) with 6 fans (2-front intake, 2- lower rear exhaust, 2-upper rear exhaust), PCP&C 510 P/S. Settings: NB Strap - By CPU Dram Ratio - 1:1 AGP Ratio - Fixed Fixed AGP/PCI Freq - 66/33 CPU Core V - 1.725v (Too high? Temp = 48° idle, 54-56° load) DDR SDRAM V - 2.7v AGP V - 1.55v DRAM Timings - Manual 3,8,4,4 AGP - 8x 3DMark2003 = 5873 (good/bad/mediocre?) I checked Overclockers.com and my CPU voltage is already higher than 99% of the DB responders. Think I should raise my DDR V to 2.8? 3.4GHz is OK if it's stable but I was hoping for 3.5 :-( -- Tally Ho! Ed Forsythe Maryland, USA Yes you vcore is to high and your AGPV is to low. Put both your mem and core volts to specs. Check your idle and load temps and write them down. It's important to do this before you o/c because the max three reads temps high. Start at 2.8 and work your way up. Use prime and memtest as you slowly increase your speed. Increase vcore by the smallest increments and only when needed. Mem voltage should not be increased until your o/c becomes totally obscene. Closely monitor your temps. If you get to a point where the small vcore increase does not help then instead of moving the vcore even higher try raising the AGP.At a certain pointt AGP voltage will cause problems. Usually it happens higher than 3.3 but I personally run mine at 1.65. Here are some bios settings you might try. Set nb strap to 800 and disable the last two gat settings. Disable cache bios and the other cache one next to it hmm? I forgot what it was called. Don't use auto settings if you know what the setting should be. ie. divider 1:1 not auto. P.S. Taking into account that the fan on your power supply is an exhaust fan this mean that you have 5 exhaust fans and 2 intake fans. This is not good. The total cfm of the two should be more closely matched. |
#7
|
|||
|
|||
Hi WD,
Thanks for the tips. I'll use them when I OC again. I'm not sure I agree with your cooling adviice. I'm inclined to go for a negative pressure in the case so that cool air is drawn in from every possible ingress point. A positive case pressure would, IMHO, cause hot air to pile up inside the case disrupting the laminar flow in the theoretical non-turbulent plenum. I suppose a 1:1 system would be perfect theoretically but you'd need a sealed case with no turbulence producing obstructions to insure that it would work at max efficiency. Tell you what - next time I'm in the case I'll disconnect the two top exhaust fans and check my temps Thanks again - -- Tally Ho! Ed Forsythe Maryland, USA "wasdiscovered" wrote in message ... "Ed Forsythe" wrote in message ... Hi Troops, I've got my 2.8C up to 3360MHz but it's not quite stable - reboots after some intense flying in my sim. Intel 2.8C, SL6WT, Zalman 7000A, ABIT IC7-Max3, ATI 9800 Pro w/Zalman VGA cooler + fan, 1G OCZ Deluxe Gold PC4000. Lian-Li full tower case (PC-75) with 6 fans (2-front intake, 2- lower rear exhaust, 2-upper rear exhaust), PCP&C 510 P/S. Settings: NB Strap - By CPU Dram Ratio - 1:1 AGP Ratio - Fixed Fixed AGP/PCI Freq - 66/33 CPU Core V - 1.725v (Too high? Temp = 48° idle, 54-56° load) DDR SDRAM V - 2.7v AGP V - 1.55v DRAM Timings - Manual 3,8,4,4 AGP - 8x 3DMark2003 = 5873 (good/bad/mediocre?) I checked Overclockers.com and my CPU voltage is already higher than 99% of the DB responders. Think I should raise my DDR V to 2.8? 3.4GHz is OK if it's stable but I was hoping for 3.5 :-( -- Tally Ho! Ed Forsythe Maryland, USA Yes you vcore is to high and your AGPV is to low. Put both your mem and core volts to specs. Check your idle and load temps and write them down. It's important to do this before you o/c because the max three reads temps high. Start at 2.8 and work your way up. Use prime and memtest as you slowly increase your speed. Increase vcore by the smallest increments and only when needed. Mem voltage should not be increased until your o/c becomes totally obscene. Closely monitor your temps. If you get to a point where the small vcore increase does not help then instead of moving the vcore even higher try raising the AGP.At a certain pointt AGP voltage will cause problems. Usually it happens higher than 3.3 but I personally run mine at 1.65. Here are some bios settings you might try. Set nb strap to 800 and disable the last two gat settings. Disable cache bios and the other cache one next to it hmm? I forgot what it was called. Don't use auto settings if you know what the setting should be. ie. divider 1:1 not auto. P.S. Taking into account that the fan on your power supply is an exhaust fan this mean that you have 5 exhaust fans and 2 intake fans. This is not good. The total cfm of the two should be more closely matched. |
#8
|
|||
|
|||
The laws of physics can only be applied to a certain point. You must
take all the wire & cables in the system into consideration plus interference from other cards ie video, nic etc.. Your lower fans will suck out air pushed in from the front fans leaving the upper fans sucking vacuum (path of least resistance) Your upper fans will attempt (not necc do) suck in air from the other orifices/or cracks in the case (couldn't resist that one;-). Heat rises it does not drop to the bottom of the case. Is your CPU in line with the lower fans or is it higher? If higher the incoming air (front fans) will not be available to cool it, passing by at a lower level. Solution disconnect at least 1 lower fan to direct air flow past the CPU or open a decent hole in the upper front of the case so that negative pressure has something to work with other than vacuum. I do not find your load temps that high, Mine is 48C. Idle temps are 27C however. Locust On Tue, 11 May 2004 17:40:47 -0400, "Ed Forsythe" wrote: Hi WD, Thanks for the tips. I'll use them when I OC again. I'm not sure I agree with your cooling adviice. I'm inclined to go for a negative pressure in the case so that cool air is drawn in from every possible ingress point. A positive case pressure would, IMHO, cause hot air to pile up inside the case disrupting the laminar flow in the theoretical non-turbulent plenum. I suppose a 1:1 system would be perfect theoretically but you'd need a sealed case with no turbulence producing obstructions to insure that it would work at max efficiency. Tell you what - next time I'm in the case I'll disconnect the two top exhaust fans and check my temps Thanks again - -- Tally Ho! Ed Forsythe Maryland, USA Yes you vcore is to high and your AGPV is to low. Put both your mem and core volts to specs. Check your idle and load temps and write them down. It's important to do this before you o/c because the max three reads temps high. Start at 2.8 and work your way up. Use prime and memtest as you slowly increase your speed. Increase vcore by the smallest increments and only when needed. Mem voltage should not be increased until your o/c becomes totally obscene. Closely monitor your temps. If you get to a point where the small vcore increase does not help then instead of moving the vcore even higher try raising the AGP.At a certain pointt AGP voltage will cause problems. Usually it happens higher than 3.3 but I personally run mine at 1.65. Here are some bios settings you might try. Set nb strap to 800 and disable the last two gat settings. Disable cache bios and the other cache one next to it hmm? I forgot what it was called. Don't use auto settings if you know what the setting should be. ie. divider 1:1 not auto. P.S. Taking into account that the fan on your power supply is an exhaust fan this mean that you have 5 exhaust fans and 2 intake fans. This is not good. The total cfm of the two should be more closely matched. |
#9
|
|||
|
|||
"Ed Forsythe" wrote in message ... Hi WD, Thanks for the tips. I'll use them when I OC again. I'm not sure I agree with your cooling adviice. I'm inclined to go for a negative pressure in the case so that cool air is drawn in from every possible ingress point. A positive case pressure would, IMHO, cause hot air to pile up inside the case disrupting the laminar flow in the theoretical non-turbulent plenum. I suppose a 1:1 system would be perfect theoretically but you'd need a sealed case with no turbulence producing obstructions to insure that it would work at max efficiency. Tell you what - next time I'm in the case I'll disconnect the two top exhaust fans and check my temps Thanks again - -- Tally Ho! Ed Forsythe Maryland, USA "wasdiscovered" wrote in message ... "Ed Forsythe" wrote in message ... Hi Troops, I've got my 2.8C up to 3360MHz but it's not quite stable - reboots after some intense flying in my sim. Intel 2.8C, SL6WT, Zalman 7000A, ABIT IC7-Max3, ATI 9800 Pro w/Zalman VGA cooler + fan, 1G OCZ Deluxe Gold PC4000. Lian-Li full tower case (PC-75) with 6 fans (2-front intake, 2- lower rear exhaust, 2-upper rear exhaust), PCP&C 510 P/S. Settings: NB Strap - By CPU Dram Ratio - 1:1 AGP Ratio - Fixed Fixed AGP/PCI Freq - 66/33 CPU Core V - 1.725v (Too high? Temp = 48° idle, 54-56° load) DDR SDRAM V - 2.7v AGP V - 1.55v DRAM Timings - Manual 3,8,4,4 AGP - 8x 3DMark2003 = 5873 (good/bad/mediocre?) I checked Overclockers.com and my CPU voltage is already higher than 99% of the DB responders. Think I should raise my DDR V to 2.8? 3.4GHz is OK if it's stable but I was hoping for 3.5 :-( -- Tally Ho! Ed Forsythe Maryland, USA Yes you vcore is to high and your AGPV is to low. Put both your mem and core volts to specs. Check your idle and load temps and write them down. It's important to do this before you o/c because the max three reads temps high. Start at 2.8 and work your way up. Use prime and memtest as you slowly increase your speed. Increase vcore by the smallest increments and only when needed. Mem voltage should not be increased until your o/c becomes totally obscene. Closely monitor your temps. If you get to a point where the small vcore increase does not help then instead of moving the vcore even higher try raising the AGP.At a certain pointt AGP voltage will cause problems. Usually it happens higher than 3.3 but I personally run mine at 1.65. Here are some bios settings you might try. Set nb strap to 800 and disable the last two gat settings. Disable cache bios and the other cache one next to it hmm? I forgot what it was called. Don't use auto settings if you know what the setting should be. ie. divider 1:1 not auto. P.S. Taking into account that the fan on your power supply is an exhaust fan this mean that you have 5 exhaust fans and 2 intake fans. This is not good. The total cfm of the two should be more closely matched. Operating a case with significant negative air pressure means it will draw air (and dust !!) inefficiently through every slit and pinhole in the case/case cover. IMO that is not a positive situation. I prefer balanced or slightly positive air pressure, to keep things inside clean (no dust accumulation). All air should be drawned in via filters. My advice to you situation is try to modify the exhaust fan connection from 12V to 7V (you can do that with a modded 4pin molex or cut & solder connectors. this will a) still utilize your existing hardware, b) reduce noise levels, c) reduce air draw to more balanced level, d) reduce power draw (and heat) from PSU. Start with lower exhaust fans - they draw relatively cool air from front (lower) intake fans out of the case before it removes much heat. let the air flow across your system upwards and backwards, to the PSU+upper exhaust fans. Optionaly disable one or both lower exhaust fans (to reduce noise even more) or put them both at 5V for extremely low RPM, noise and power draw... and put your 2 upper exhaust fans at 17V (or one at 7V, one at 12V to be on the safe side) - if you are comfortable with this kind of modification (crossing the wires of the 5v+, 0, 0, 12v+ molex connectors). Internal air flow pattern is invariant to external differences. Its turbulency will only be governed by geometry, path, and the gradient of intake/exhaust ports, whether it is higher or lower than outside pressure is irrelevant. |
#10
|
|||
|
|||
You're not being a candy-a, Ed. If I had an idle temp of 48C I'd be a
little concerned too. Have you got a good CPU heatsink/fan setup? Also, what par of Maryland you from? Todd Ellicott City, MD "Ed Forsythe" wrote in message ... Hi Pham, I'm still at 2.8 contemplating another try at 3.2 or 3.4. Maybe I'm just a candy-a** but the high temps really bothered me. When I have time/guts for another go I'll post -- Tally Ho! Ed Forsythe Maryland, USA "Pham" wrote in message s.com... Ed Forsythe wrote: OK Wayne, I'll back off to 2.8GHz then increase DDR SDRAM (vDiimm?) to 2.8. Boot and fiddle then increase vCore to 1.7 shoot for 3.4 again and see what happens. Still the troops on Overclockers.com are getting excellent results with voltages (vCore) in the 1.6 range. Or maybe I'll just be satisfied at 3.2GHz. Frankly I can't see any real difference between 3.4 and 3.2. When I was using a sim at 3.4 MBM indicated my CPU temp was 61°C - that's a bit too high for me. Thanks for the tip How'd it go, Ed? Inquiring minds want to know... they also want to go FASTER! -Pham |
|
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
OC'ing and memory and misc results | Destroy | Overclocking | 4 | February 28th 04 08:03 PM |
Best MB for OC'ing P4 | JM | Overclocking | 12 | January 2nd 04 01:58 PM |
My 3DMark 2003 results | crapbottom | Overclocking | 14 | November 13th 03 08:00 AM |
First Time Overclocking Results | uncooked meat prior to state vector collapse | Overclocking | 3 | August 30th 03 02:36 PM |
P4 OCing references | CS | Overclocking | 0 | August 29th 03 07:54 AM |