how low can vcore get and still be stable (q9550 @ 3.4)
Forum Overclocking : CPUs - how low can vcore get and still be stable (q9550 @ 3.4)
hi comrades,
this is my setup:
ANTEC 12000 case
MB ASUS P5Q Deluxe
Core 2 Quad Q9550
Radeon HD4870x2
Corsair Dominator 1066, (4x2)
PC Power & Cooling 750W
OCZ Vendetta 2
ive clocked my 2.83 quad to 3.4, what im trying to do now is lower the vcore to see how far down it goes while keeping stability
now the original clock was setup like this:
Vcore =1.2000, fsb = 400, multi = 8.5, Vdimm = 2.1, DIMM =1066
everest idle: mobo 39 degrees, cpu =31, core0 =51, core1 =48, core2 =45, core3 =45
speedfan idle: a degree higher than everest
realtemp idle: core0 =46 , core1 = 42, core2 =39, core3 = 39,
VID 1.2375, minimum hottest core 46, maximum 49
CPU-Z: CPU revision C1, Vcore 1.184, ddr frequency 533.3, fsb
ram = 3:4, timings 5-5-5-15-2T
starting prime95 64 bit version multiple cores (4+)(small fft test):
everest mobo 40, cpu 58, core0= 70, core1 = 67, core2 = 73, core3 = 72
coretemp: 0=71, 1=68, 2=72, 3=72
real temp: 66, 63, 68, 68, minimum hottest core 45, maximum 67
everest and core temp show some cores go to 73 degrees, is that too much temp for 24/7 work?
cpu temp during full load 60-61 (everest & speedfan)
after an hour of prime stress test the temp lowered a bit (55-57 cpu, 67-69 cores)
a day later...
lowered Vcore to 1.18125, temps fell to 55 max for cpu, cores maxed at 69 (everest)
real temp says: 62 60 64 64
minimum temps 61 60 61 63
maximum temps 63 60 65 65
how far do u think i can lower Vcore and still keep it stable at this 3.4 clock?
no idea. drop it one increment at a time and prime/orthos it.
| V3NOM wrote : no idea. drop it one increment at a time and prime/orthos it. |
is there any danger involved in droping it at that clock or any other for that matter?
how far do u think this cpu can go safe temp wise for 24/7 use?
update:
dropped vcore to 1.16875, cpu-z is showing 1.152 @ idle, 1.168 @ load
cpu temp is still 55 @ load and why is it still 55? shouldnt the temp drop with dropping vcore, maybe the psu is compensating for lower voltage with more current? shouldnt i be getting bsod at some point, its still a 3.4 clock and 1.16875 doesnt look like enough voltage for that clock, or am i mistaken?
Your temps are fine.
As long as they do not go over 75c you will have no issues.
Keep in mind, when you are stress testing, there is an artificially high load on your CPU.
Under normal usage, you should never strain your CPU that much and, therefor, your temps will be lower.
To get the lowest stable voltage, drop you voltage in one step increments testing with prime95 between each drop. When you get to low, one of your cores will start failing testing. At that point, raise your voltage two settings and call it stable. Lowering your voltage below stock will never harm your CPU. It should allow it to run a bit cooler and draw less power.
thank u outlw, i just came to a wall by dropping vcore, at 1.13125 prime fails and computer restarts, im currently priming with vcore set to 1.13750, so far so good with cpu temp at 53 and real temp cores @ 61 58 60 60, if i can run this cpu at 3.4 gigs @ 1.13750 voltage stable, do i have any chance of running it @ 3.8 gig at stock voltages say 1.2?
According to Comps Core 2 Quad and Duo Temperature Guide a Tjunction of 75C is hot, 70C is warm and 65C is safe. If you approach 70C you would need to do the calibrations for Tcase to Tjunction to ensure that the Tcase max of 71C was not exceeded. I would recommend staying below a Tjunction of 70C and preferably below 65C.
Message edited by Zorg on 11-06-2008 at 11:47:58 AM
If you have your chip stable at 3.4Ghz with only 1.1375v you may just reach 3.8Ghz with only 1.2v. The only way to know for sure it to give it a try. If it is not 100% stable, you still have some room to bump the voltage. 1.3v or a bit higher will still be safe for your chip, assuming you can keep the temps in check. From the looks of it, you have a very good chip.
ok, 3.825 ,fsb = 450 @ 1.2 V gave me bsod so i bumped it up to 1.25 V and now its priming fine with cpu temp @ 68 and cores @ 79 76 85 87 (everest) and max cores 81 72 81 83 (core temp), what about cores being so high? is it ok as long as cpu temp isnt going over 75 or should i lower the fsb and vcore to avoid damaging cpu?
Well, it is an excellent overclock with such low voltage.
Unfortunately, it seems your cooling solution is not quite up to the task.
You need to keep your core temp at or below 75c for every core or you risk damaging your CPU.
For the moment, I would recommend lowering the speed and voltage until you are able to upgrade you cooling. With such a good chip, I would highly recommend putting it on watter and seeing how far you can take it with 1.4v....
what about 1.4 V being outside of intel 0.8-1.3625 specs? my friend managed to push q9550 to 4.2 gigs with water but im not sure if he went that far with voltage
As long as it is kept cool, 1.4v should be safe for your chip.
As you are stable at 3.8Ghz with only 1.25v, I believe you would have no issues running at 4.2Ghz or higher with 1.3625v or less if you can keep the temps in check...
excellent stuff outlw!
im back to 3.4 @ 1.1375 till i get some water, everest is reporting 41 degrees while priming, whats the safe range for mobo temp wise?
According to the Intel Spec Sheet, the P45 Nb is rated to a maximum of 103c.
As the P45 is based off a 65nm process and runs fairly cool, I doubt you will have any issues with it overheating.
hi again, my overclocking soul is restless so ive decided to push the fokker again, as i type this the prime is running for almost 3 hours stable at 3.6 gig fsb 425, vcore 1.18750 and memory @ 1066
cpu temp reported by everest is 53-54, mobo stays at 29
core temps go from 65-70
real temp reports minimum temps at 59 56 57 56
maximum at 63 60 65 65
its incredible even though this chip is C1 revision it manages to keep this overclock stable at only 1.18750 vcore
lucky you, I have a q9550 on 780i, currently overclocked to 3.4 at 1.275. anything lower Intel Burn Test gives me BSOD. My goal is to get to 3.6ghz.
Lemme ask you, did you have to up the FSB volts?
u mean gtl voltages? i was running this cpu at 3.8 the other day with OCZ Vendetta 2 but thew temps of the cores reach and pass 80
do i need some guide for raising or lowering gtl voltages, how can i set those up and not fry anything? i ask coz id like to further lower my vcore
also whats that about PLL voltages, from what i understand theyre tied to sb so wont raising pll result in need to tweak sb voltage too?
Im talking about the voltage setting right under the cpu setting in bios, i think its called the vtt (i have a 780i board).
also this might help give you an idea about ur temps. my temps are:
idle: 29c
load (IBT, OCCT, Prime): 57c max
ya ur cpu temp under load is 57c max but that doesnt mean ur core temps rnt going over 65 or maybe even 70, what r ur core temps?
i havent touched vtt voltage yet, i still have to try gtl voltages first
sorry i meant core temps. they dont go over 57c. thats the highest any one of them has ever gone.
wow thats what im aiming for, at 400 fsb, vcore of 1.14375, and 3.4 gigs my core temps go to 61 tops in everest, 58 in real temp, do u have water cooling or what coz i cant see how can u have such low temps at 3.4 clock
do u guys think id lower my core temps by lowering multi to 8 and raising fsb to 450 and getting that 3.6 clock that way instead ?
I use the tuniq 120 tower. Ill admit the first time I put it on with the thermal stuff my temps were like yours if not higher at stock. Then i took everything off and i spend like 30 mins carefully appling the grease. I did it evenly and carefully and then i mounted the tuniq 120. I loaded up the comp ran tests, at stock core idle was at 25c, 100% load was less than 45c.
im waiting for my xigmatek crossbow or tr s775 bolt-thru kit and then ill apply thermals again
is it true u need to set pcie frequency to 101 otherwise itll be auto raised with fsb and shyt?
i have no idea about pcie, i have read that you need to leave it at 100, but i am not sure at all.
p.s. i just OCed to 3.625 with a vcore of 1.368v and CPU FSB at 1.3v, anything lower on vcore and im not stable at all. and I cant take my q9550 higher than that(i.e. like 3.8) i even tried at 1.4v but its still not stable.
weird, i can get it stable at 3.8 no sweat but the temps r in the 80s
what vcore for your stable 3.8? im scared of going over 1.36v on cpuz since our cpus safe is 1.36 according to intel. Isnt 75c and over too hot?
i was running 3.825 stable at vcore 1.125, 450 fsb, multi 8.5
holy jesus thats crazy good, at 1.125 wow man. thats a special chip you got there. I have no idea why i cant get it stable with a lower vcore. did you up your CPU FSB(vtt?) much?
i never touched any voltages except ram 2.1 and vcore, yeah its a very good chip, im running it at 3.4 stable with only 1.1375 vcore, and 3.8 with only 1.250, im sure i could lower vcore even more if i knew how to set those gtl voltages, by the way the chip is C1 revision, it aint even E0
is there any chance i could run 533 fsb, memory 1066, multi 6.5 for the 3464 clock?
do u think q9550 and corsair dominator can handle such high fsb?
my blend test fails after a few minutes on one of the cores, vcore 1.1475, fsb 400, strap 400, vdimm 2.1, but small fft can go on for hours, i dont see why memory would fail if corsair dominator has been verified to operate at 2.1 , 1066
ok guys, whats ur method of searching for "fsb wall", i lowered the multi to 6, raised the fsb to 533 and kept the memory < or = 1066, it just couldnt boot anymore until bios loaded the default values,
then i took a more gradual approach, i raised it slowly to 450, 470 seeing if itll boot into windows, i managed to boot into windows at fsb 493 but not at 494, even at fsb 493 i managed to stay stable only for a coupla minutes, it froze on me like a bitch, i did all this on vcore 1.25, im not sure if i needed more vcore but i tried setting 1.3 to boot on 494 fsb, still no luck, whats ur recipe of learning what the maximum stable fsb is?
i think your on the right track, mine wasnt stable at 3.6 till i put 1.368 in bios.
lol i aint got the balls to go 1.365 even if i had water
my temps are fine. so im not worried, and i made another thread asking what ppl thought about the high voltage. I guess its okay to leave it as such, i might pull it back down to 1.275 3.4ghz to be safe. I see no difference in games from the 200mhz.
lol i see no difference between 2.83 and 3.4 either
from 2.8 to 3.4 i get like 3-5 more frames playing crysis at ultra high 1650 x 1050. Thats worth it for me, makes the game playable. lol
thats why i went with hd4870x2, aint gonna let no game play below 30fps at 1920x1080, so siree bob
Eh, its not that unusual... I have a q9550 at 3.85 24 hour prime95 stable at 1.25 too
I want some AWESOME CHIP MAN!!!
Love too
Yeah those 5 fps make a huge difference
ROFL !!! HAHAHAHA WHAT ??? 3.85 at 1.25 stable thats INSANE !!!! lol anyone driving that clock at that voltage needs some serious lovin, ill ask my girlfriend if shell come over to ur place and u let ur chip walk over to mine
q9550 love 4 ever
my stupid chip doesnt like anything over 3.4, and when i do make that bitch go over that it wants the max amount of voltage i can give it, lol. you guys are lucky
lol i think u and me both can lower that vcore by playing with gtl voltages, i just gotta figure out how
i havent been stable in blend mode (prime95) since i bought this rig almost 3 months ago
there would always be rounding errors only 2 mins into testing
today i changed these two options from auto to disabled and the last option from auto to lighter and voila its priming without errors for 2 and a half hours now
DRAM Static Read Control: Disabled
DRAM Read Training: Disabled
AI Clock Twister: Lighter
this is the current config:
AI Overclock tuner: Manual
CPU Ratio Setting: 8.5
FSB Strap to North Bridge: 400 MHz
FSB Frequency: 400
PCI-E Frequency: 100
DRAM Frequency: DDR2-1066
DRAM CLK Skew on Channel A1: AUTO
DRAM CLK Skew on Channel A2: AUTO
DRAM CLK Skew on Channel B1: AUTO
DRAM CLK Skew on Channel B2: AUTO
DRAM Timing Control: manual
1st Information : 5-5-5-15-3-52-6-3
CAS# Latency: 5
DRAM RAS# to CAS# Delay: 5
DRAM RAS# Precharge: 5
DRAM RAS# Activate to Precharge: 15
RAS# to RAS# Delay : AUTO
Row Refresh Cycle Time: AUTO
Write Recovery Time: AUTO
Read to Precharge Time: AUTO
2nd Information : 7-3-5-4-5-4-6
READ to WRITE Delay (S/D): AUTO
Write to Read Delay (S): AUTO
WRITE to READ Delay (D): AUTO
READ to READ Delay (S): AUTO
READ to READ Delay (D): AUTO
WRITE to WRITE Delay (S): AUTO
WRITE to WRITE Delay (D): AUTO
3rd Information : 14-5-1-6-6
WRITE to PRE Delay: AUTO
READ to PRE Delay: AUTO
PRE to PRE Delay: AUTO
ALL PRE to ACT Delay: AUTO
ALL PRE to REF Delay: AUTO
DRAM Static Read Control: Disabled
DRAM Read Training: Disabled
MEM. OC Charger: AUTO
AI Clock Twister: Lighter
AI Transaction Booster: Auto
C/P : A1 A2 A3 | B1 B2 B3
LVL : 09 09 09 | 09 09 09
CPU Voltage: 1.14375
CPU GTL Voltage Reference (0/2): AUTO
CPU GTL Voltage Reference (1/3): AUTO
CPU PLL Voltage: AUTO
FSB Termination Voltage: AUTO
DRAM Voltage: 2.10
NB Voltage: AUTO
NB GTL Reference: AUTO
SBridge Voltage: AUTO
PCIE SATA Voltage: AUTO
Load Line Calibration: Enabled
CPU Spread Spectrum: AUTO
PCIE Spread Spectrum: AUTO
CPU Clock Skew : Auto
NB Clock Skew : Auto
CPU Margin Enhancement: Optimized
Advance CPU Settings
CPU Ratio Setting: 8.5
C1E Suppport: Disabled
Max CPUID Value Limit: Disabled
IntelĀ® Virtualization Tech: Disabled
CPU TM Function: Disabled
Execute Disable Bit: Enabled
would any of u guys change anything and why?
ok those 3 options i changed in the last post actually made me fail post today, i tried to turn on my machine and no video signal on the tv, i tried reseating everything and when i finally took out one 2 gig dominator stick the post went thru and it said "overclocking failed"
i immediately knew it was those last changes i made that were the culprit
ok what i did now was lower the voltages one notch above minimum and enabling LLC and disabling CPU and PCIE spread spectrum, the setup looks like this now:
AI Overclock tuner: Manual
CPU Ratio Setting: 8.5
FSB Strap to North Bridge: Auto
FSB Frequency: 400
PCI-E Frequency: 100
DRAM Frequency: DDR2-1066
DRAM CLK Skew on Channel A1: AUTO
DRAM CLK Skew on Channel A2: AUTO
DRAM CLK Skew on Channel B1: AUTO
DRAM CLK Skew on Channel B2: AUTO
DRAM Timing Control: manual
1st Information : 5-5-5-15-3-52-6-3
CAS# Latency: 5
DRAM RAS# to CAS# Delay: 5
DRAM RAS# Precharge: 5
DRAM RAS# Activate to Precharge: 15
RAS# to RAS# Delay : AUTO
Row Refresh Cycle Time: AUTO
Write Recovery Time: AUTO
Read to Precharge Time: AUTO
2nd Information : 7-3-5-4-5-4-6
READ to WRITE Delay (S/D): AUTO
Write to Read Delay (S): AUTO
WRITE to READ Delay (D): AUTO
READ to READ Delay (S): AUTO
READ to READ Delay (D): AUTO
WRITE to WRITE Delay (S): AUTO
WRITE to WRITE Delay (D): AUTO
3rd Information : 14-5-1-6-6
WRITE to PRE Delay: AUTO
READ to PRE Delay: AUTO
PRE to PRE Delay: AUTO
ALL PRE to ACT Delay: AUTO
ALL PRE to REF Delay: AUTO
DRAM Static Read Control: Disabled
DRAM Read Training: Disabled
MEM. OC Charger: AUTO
AI Clock Twister: Lighter
AI Transaction Booster: Auto
C/P : A1 A2 A3 | B1 B2 B3
LVL : 09 09 09 | 09 09 09
CPU Voltage: 1.14375
CPU GTL Voltage Reference (0/2): AUTO
CPU GTL Voltage Reference (1/3): AUTO
CPU PLL Voltage: 1.52
FSB Termination Voltage: 1.12
DRAM Voltage: 2.02
NB Voltage: 1.12
NB GTL Reference: AUTO
SBridge Voltage: AUTO
PCIE SATA Voltage: AUTO
Load Line Calibration: Enabled
CPU Spread Spectrum: Disabled
PCIE Spread Spectrum: Disabled
CPU Clock Skew : Auto
NB Clock Skew : Auto
CPU Margin Enhancement: Optimized
Advance CPU Settings
CPU Ratio Setting: 8.5
C1E Suppport: Disabled
Max CPUID Value Limit: Disabled
IntelĀ® Virtualization Tech: Disabled
CPU TM Function: Disabled
Execute Disable Bit: Enabled
i didnt touch gtl voltages or references coz i presume they wont overvolt above safe values or will they, is there anything out of the ordinary with this setup or is it safe for 24/7 use?
Wow you must have a really good CPU I can't even get my C1 to go to 3.0 on 1.25vcore but then again im using a ep45ud3r and artic freezer 7
Message edited by inragedtwitch on 06-01-2009 at 01:16:17 PM
There are 11 identified and unidentified users. To see the list of identified users, Click here.
