If you are using the motherboard in your signature, its going to be a tough ride to overclock. Can you list your settings as the spreadsheet can't be accessed.
Printable View
If you are using the motherboard in your signature, its going to be a tough ride to overclock. Can you list your settings as the spreadsheet can't be accessed.
VTT=1.3
NB=1.45
SB=1.55
HT=1.3
DRAM=1.85
Vcc=1.35 BIOS, 1.32 real
With these settings core 2 failed after 20 minutes.
Try these links: http://spreadsheets.google.com/pub?k...j6-dWLOJRG-9Bg
http://spreadsheets.google.com/pub?k...j6-7U-LxAFs1HA
I don't think I published it before
much better...ty
I have a Q9450 @ 3.2ghz 1.225 stable. But I would like 3.6ghz, what settings would i need voltage wise etc on a Abit Quad GT X38?
been playing with q9450 p45-ds4p today and have managed 500fsb 4ghz
http://i3.photobucket.com/albums/y83/dasa09/spi4ghz.jpg
http://i3.photobucket.com/albums/y83...9/P1100012.jpg
some of the v settings may not need to be that high it will take some time to work thew them all and to find a fully stable speed
just a shame i killed my 8800gt that was 850mhz core 24\7 stable while installing the new gear would have been some fun to bench it
Prime stable 450 fsb under 1.20 VCore ;)
My System:
Intel Q9450
Asus Triton 79 Amazing heatsink
Asus P5Q Deluxe
2x2GB G.Skill DDR2-1066 PC2-8500 rated for 2.1V
Asus EAH3870 X2 1GB vid card
(This photo shows prime run for about 30 min with large FFTs. My longest prime run was 5 hours blend with no problems)
http://img237.imageshack.us/img237/2628/oc2ko1.jpg
Notice the giant difference between core3 and core0. I have a feeling that the temp sensors on this one are off. It is hard to believe that there is a swing of 15+ degrees C between the cores. I have tried resetting but little effect. I need to lap the heatsink as there were lots of factory scratches when I installed. I am not very impressed by the Triton 79. Once I have my heatsink lapped and temps fall back into acceptable ranges, I will really try to push this baby. I was able to boot into windows at 460 fsb with these settings but it is not prime stable.
Bios Settings:
JumperFree Configuration Settings
AI Overclock tuner: MANUAL
CPU Ratio Setting: 08.0
FSB Strap to North Bridge: 333MHz
FSB Frequency: 450
PCI-E Frequency: 100
DRAM Frequency: DDR2-1081MHz
DRAM Timing Control: auto
DRAM Static Read Control: disabled
DRAM Read Training: disabled
MEM. OC Charger: auto
AI Clock Twister: auto
AI Transaction Booster: auto
CPU Voltage: 1.18750
CPU GTL Voltage Reference (0/2): 0.655
CPU GTL Voltage Reference (1/3): 0.675
CPU PLL Voltage: 1.54
FSB Termination Voltage: 1.18
DRAM Voltage: 2.02 (G.SKILL DDR2 8500 rated for 2.1V)
NB Voltage: 1.26
NB GTL Reference: auto
SBridge Voltage: 1.2
PCIE SATA Voltage: auto
Load Line Calibration: Enabled (This helps a lot)
CPU Spread Spectrum: Disabled
PCIE Spread Spectrum: Disabled
CPU Clock Skew : auto
NB Clock Skew : auto
At idle it will draw 1.184V. Load Line Calibration set to Enabled helped me a lot. Again I will play around with these a lot more once I have lapped my heatsink and I see more acceptable temps.
Processor Info:
Core Temp reports VID of 1.1625V
Pack 06/01/2008
BATCH #: L818A685
Version #: E34934-001
holy jesus man ... thats insane.
given the massive difference and the fact that the heat would naturally be drawn to the cooler core, thus making the cooler core hotter.
i would say its one of 2 things
1)
temp probe on the core is bad
2)
Your system was under such heavy load that the temp PROGRAM was unable to update the temps correctly thus making it appear as if there is such a huge difference.
If the temp diodes are indeed bad on core0, I don't think I am going to RMA it. Getting to 450 fsb on less than 1.20 VCore is quite a feet that almost no one is able to replicate.
I get 1.125v for 8x450 fsb here.Quote:
Originally Posted by zeusthemoose
I can't get my P5K large FTT stable in prime95 at 465fsb. One core always fails.
Very nice ;)
prime stable for 5 hours at 3.9ghz 488fsb 1.525v
it could only prime for ~30min at 1.5v
http://i3.photobucket.com/albums/y83...prime39ghz.jpg
3dm06 8533 4ghz x1900xtx 783c 891m
http://service.futuremark.com/compare?3dm06=7450825
11.453 spi at 4080mhz 510fsb
http://i3.photobucket.com/albums/y83/dasa09/pi11453.jpg
it seems to become very unstable and have trouble posting when i take the ddr2 over 1000mhz or turn the v up to its ratted setting of 2.2v
I would up the vtt (cpu termination (no more than 1.4v)) that way you can lower vcore a bit.
thanks for the tip but unfortunately putting it up to 1.36v didnt allow me to lower the vcoreQuote:
Originally Posted by Shadowthor
its possible taking it up to 1.4v may be enough to run the cpu at 1.51875v but that dosent even seem worth testing
hey dasa can you post your specs please ?
here's mine Q9450 on ep45-ds3r F7 bios
http://img208.imageshack.us/img208/2413/testid1.jpg
http://img208.imageshack.us/img208/2...4e166ae9ff.jpg
I'm getting random reboots with this mobo, I can't figure out what's causing it. :(
Hy, i have some problems with asus maximus extreme,
Q9450(thermalight ifx-14), asus maximus extreme (bios 1102), corsair 2x2giga 1600mhz (CM3X2g1600c9DHX).
1 i whanted to test how much can u OC without any voltage , so i manualy put all voltages on default , you can see may settings on bios :
-Ai overclock tune = manual
-OC mem lev up= auto
-cpu ratio set= 8.0
-FSB frecv= 333
-pci frecv= 100
-ddram frecv = ddr1333mhz
-ddram command rate = 2t
-ddram timing control= manual
-cash# latency = 9
-ras# to cas# latency =9
-ras# precharge = 9
-ras# active time=24
-ras# to ras# delay = auto
-row refresh cycle time = auto
-write ricovery time = auto
-read to precharge time = auto
-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
-ddram static read control = disable
-dram dynamic write control= disable
-Ai clock twister = strong
-ai clock skew for chanel A = auto
-ai clock skew for chanel B = auto
-Ai transaction booster = enable
-Boost level=2
-Cpu voltage 1.24375v (temperatures at full load 68*c with core temp 0.99)
-cpu ppl voltage = 1.50v
-north bridge voltage= 1.25v
-ddram voltage = 1.80v
-Fsb termination voltage= 1.20v
-south bridge = 1.050v
-Loadline calibration = enable
-Cpu gtl reference= x63
-north bridge gtl reference = x67
-ddr3 chanel A+B REF voltage = auto
-ddr3 controller REF voltage = ddr3_REF
-sb 1.5 voltage = auto
CPU ADVANCED SETTINGS
-Cpu ratio setting = [8.0]
-C1E support = disable
-Cpu tm function = disable
-vanderpool technology = disable
-execute disable BIT = disable
-MAC CPUID value limit = disable
North Bridge chipset conf.
-memory remap value feature=enable
-initiate graphic adapter= [PEG/PCO]
-PEG Port control = auto
-peg port force x1 = disable
PCIPnp
-plug and play O/S = [NO]
POWER
-suspend mode = [auto] (u can set to S1 (po)only or S3 only)
-repost video on S3 resume = disable
-ACPI 2.0 support = enable
-ACPI APIC support = enable
SATA conf.
-sata config = [enchanted] (u can set ro disable or compatible)
-configure sata os = IDE (raid or AHCI)
-Hard disk write protect disable
-sata detect time out (sec) = [0]
With this default settings i try to rise fsb and surprised sistem boot @400fsb and even @450fsb (but not stable some programs like nvidia driver , bitdefender 2008 , logitech driver worked bad)
I try test more at 425 , the probrams worked fine at start-up pc , but in 3dmark2006 textures were wrong
So go lower @400fsb (here all worked fine but after 1 hour maibe 2 i get blue screens , i even get blue screens @ 333fsb but rare)
So i continued test @400fsb , i use higher voltages :
-Cpu voltage 1.24375v ( with default and i got blue screen after 1 hour temp@71*c , and then 1.35v but again cpu freezed after about 15 min , temperatures rised to 77*c)
-cpu ppl voltage = 1.60v
-north bridge voltage= 1.40v
-ddram voltage = 1.84v
-Fsb termination voltage= 1.40v
-south bridge = 1.100v
I tought maibe is a ddram problem so i lower DDram from 1333 to 1.066 and again rise fsb to 400-410 , same result blue screen
So now im stuck ,any hlp plzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz, sugestions etc.
I just whant my pc to be stable arount 400fsb .
Ps: sry for my bad english.
rusty i may have trouble with rebooting if i take the v for the ram over 2.08v when the ram is rated for 2.2v so it may be worth trying to lower your ram v and speed i have the divider set to 2.00B i think it is and id be interested to hear if this helps you
with all v stock at 400fsb programs will sometimes freeze and prime wont error but it will just disappear
i was able to get up to 460fsb without increasing anything but vcore and by that i mean it could run prime for a ~10min before i moved on to try higher speeds
specs...
q9450
p45-ds4p bios f7
2x2g 800 444-12 patriot 2.2v
x1900xtx
510w pcp&c
Mine idles at around 51 and hits 63 under load, are you using liquid cooling to be in the 40s? If not, wtf is going on with my chip? I've reinstalled the heatsink twice.Quote:
Originally Posted by rusty
no sir I'm using air cooling Thermalright 120x(panaflo fan) with mx-2 thermal paste.Quote:
Originally Posted by DarkDisciple
Quote:
Originally Posted by dasa
please check here for my specs and story :(
http://www.clunk.org.uk/forums/overc...html#post25386
I'm using an thermalright ultra 120 with AS5, and idling at 50-53. Bad chip?Quote:
Originally Posted by rusty
how's it setup ?Quote:
Originally Posted by DarkDisciple
mine is almost push pull, almost cause I have 2 fans only.
on is connected to the case grill the other one is to the heatsink pulling the air out. Try to reseat the heatsink make sure it sits tight.
Side fan blows in general direction of memory / behind heatsink fan. Fan on heatsink blows towards back of case and standard exhaust fan there. PSU fan also pulls air out from above the heatsink.
I have reinstalled it twice now with varying ammounts of thermal compound, and temps remain ~ the same. I am using the metal bracket behind the mobo to secure it, as opposed to the push pins. It feels like it is on there securely, and looks to be lined up correctly.
It does seem possible that their is something wrong with heatsink install, i just can't figure what. Even when the cpu is at 60 degrees, the top of the heatsink is cool / room temperature to the touch.
GTL Refs are very high. Any way you can raise the FSB Term a little and drop the Refs?Quote:
Originally Posted by dasa
post some pictures of your setup maybe we missed something.Quote:
Originally Posted by DarkDisciple
something is wrong if yer temps are that high & the cooler is cool ...Quote:
Originally Posted by DarkDisciple
vcore is the only one that is a bit high from what i can tell and if cpu-z is accurate it has a fair bit of vdrop but thankfully idle vs load difference is smallQuote:
Originally Posted by Cob
as for everything else it was set to stock settings (normal not auto) except cpu pll, dram v, mch core, and vtt which i tried to raise so i could lower vcore
so cpu ref is at what this mb sets with no overclock
thanks
edit:
ah it is a tad high i have been having a look around the forums
and this is probably the most interesting thing i found although its a e8500 not a q9450
http://www.clunk.org.uk/forums/24996-post36.html
rusty
it looks like you have already been helped by people who know more about the mb than i do as it only arrived a few days ago and before that i was using dfi rd600 e6600
the only other thing i can think of is try some different ram or try one stick and have you used memtest86
Hi all. Needed a new system and ended up going with the 9450 and P5Q. My new gear is below. I'd like to offer a special thanks to those who've gone before me here... cob, luci5r...
JumperFree Configuration Settings
AI Overclock tuner: MANUAL
CPU Ratio Setting: 8
FSB Frequency: 475
PCI-E Frequency: 101
FSB Strap to North Bridge: AUTO
DRAM Frequency: 951
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
CAS# Latency: 5
DRAM RAS# to CAS# Delay: 5
DRAM RAS# Precharge: 5
DRAM RAS# Activate to Precharge: 18
RAS# to RAS# Delay : AUTO
Row Refresh Cycle Time: 55
Write Recovery Time: AUTO
Read to Precharge Time: AUTO
2nd Information
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
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: AUTO
MEM. OC Charger: AUTO
AI Clock Twister: AUTO
AI Transaction Booster: AUTO
Common Performance Level: AUTO
Pull-In of CHA PH1:
Pull-In of CHA PH2:
Pull-In of CHA PH3:
Pull-In of CHA PH4:
Pull-In of CHB PH1:
Pull-In of CHB PH2:
Pull-In of CHB PH3:
Pull-In of CHB PH4:
CPU Voltage: 1.33125
CPU GTL Voltage Reference (0/2): .635
CPU GTL Voltage Reference (1/3): .675
CPU PLL Voltage: 1.54
FSB Termination Voltage: 1.34
DRAM Voltage: 2.06
NB Voltage: 1.30
NB GTL Reference: AUTO
SBridge Voltage: 1.2
PCIE SATA Voltage: AUTO
Load Line Calibration: ENABLED
CPU Spread Spectrum: DISABLED
PCIE Spread Spectrum: DISABLED
CPU Clock Skew : AUTO
NB Clock Skew : AUTO
Advance CPU Settings:
CPU Ratio Setting:8.0
C1E Suppport: DISABLED
Max CPUID Value Limit: DISABLED
Intel® Virtualization Tech: DISABLED
CPU TM Function: DISABLED
Execute Disable Bit: DISABLED
3.8 Ghz
Prime stable at least 8 hours... but
even at only 1.33v, I'm hitting mid 70s on all cores with my TRUE...
Does this sound like a remount is needed or a hot chip?
What are you using to measure temps? Is there good air circulation in the system?
Good show there mate! Looks awesome.
Made a couple of suggestions in your BIOS Settings, in case you'd like to try.
As far as temperatures - during Stress Testing (100% on all 4 cores), most on air rigs are hitting 70 ... mine does too. I don't think that's a big problem.
Quote:
Originally Posted by tdenton1138
Q9450 L813A994 3.6GHz @ 1.424V 8 hours prime stable on 780i P05P
temps are 15C diff between bios and windows?!
http://img356.imageshack.us/img356/296/36001424ps7.jpg
http://img155.imageshack.us/img155/1...stabri8.th.jpg
BIOS Settings:
CPU Core - 1.50000V
CPU FSB - 1.30V
Memroy - 2.10V
SPP - 1.450V
MCP - 1.650V
SPP <-> MCP - 1.40V
Lane 0-3 - Auto (Lane 1 - +0.25mV)
Can you please tell us the vtt(vfsb) needed for that?Thanks.
Hello guys
Been following this thread for a few weeks now and just applied so I can post. Below are the numbers I am working with and the equipment I have.
CPU Intel Q9450
Motherboard EVGA 750i FTW
Memory Corsair 4-4-4-12 2.1 volts DDR2 800MHz (2 x 1 GB and 2 x 2 GB)
Nvidia SLI Certified case with 6 120mm fans (5 fans are 1800 RPM and 1 fan is 3000 RPM)
S1283 CPU cooler
600 Watt Power Supply with a built in 140MM fan that sucks air in and pushes it out the back.
(2) MSI GeForce 9600 GT 512mb video cards
I currently have this overclocked to 3.7 GHz and it runs stable playing EverQuest 2 for over 6 hours no problem but when I run Prime 95 it seems to be russian rullet. Sometimes it goes 4+ hours fine and sometimes it dies in 2 hours.
Settings
CPU Voltage 1.4volts (vdroop 1.322 - 1.360)
FSB Voltage 1.4 volts
Memory Voltage 2.2 volts (Memory set to 5-5-5-18 1014 MHz)
North Bridge 1.45 volts
South Bridge 1.40 volts
PCI-E was at 102MHz but downclocked it to 100MHz
C1E is Disabled
GTL (0) +0.65mv
GTL (1) +1.00mv (This is the core that always has the error)
GTL (2) +0.65mv
GTL (3) +0.65mv
Temps, well I need to move my fans around a bit more. When I turn my case sideways so it can suck in more air from the side fans I never exceed 62C (on full Prime95 load) when I put it into the position I want it to be in then it goes up to 66C (on full Prime95 load). I have 4 fans for intake and 3 fans (including power supply) for out take. My outtakes are the back of the case and top of the case. The intakes are the side of the case and lower front of the case. When I do a benchmark on 3DMark06 I am getting 13058 on 1 GPU and when I do SLI I am getting 17787. I wanted to get 3.7 GHz completely stable and if temps are resonable I was going to try 3.8 GHz on air cooled.
Any pointers or ideas is most appreciative.
Q9450
FPO/BATCH # L816A683
vid 1.25000
P5Q-D
vcore - 1.1800
CPU PLL - 1.52
FSB Term Volt - 1.12
Load Line Calibration - Dissabled
All other volts 1 notch above lowest value.
8 x 400fsb for 3.2Ghz = OCCT stable....vcore under load 1.14v
Woo Hoo so far
Haven't applied Luci5r's suggestions yet, but did drop the vcore down a bit 1.325. Still stable @ 3.8.
http://img172.imageshack.us/img172/6...uretestph1.jpg
I accidentally stopped the priming before taking the screenshot. CPU-z read 1.320 volts under load.
are all L813A's made the same? that's a great chip you got. I can't get mine up to 3.8 without a lot of voltage. I'm wondering if I should just sell my 780i and upgrade to a P5Q Deluxe. Seems like most everyone is having awesome oc'ing luck with these things.
the good news is, i was able to shave 0.025V off my 3.6 oc and bring my voltage to 1.408V and the temps down 5C.
I'm at 8 x 450fsb - 3.6GHz now with 1.24vcore under load. OCCT stable.
CPU PLL - 1.54
FSB Term Volts - 1.28
CPU GTL (0/2) - .630
CPU GTL (1/3) - 670
NB volts - can't remember offhand but still on low limit.
NB GTL - Auto or standard value
Load Line Calibration - disabled. I have found if I enable this then OCCT graphs show spikes up to 1.6volts on vcore under load so I feel it best to disable this function and just allow for vdroop rather than have the massive voltage spikes.
Next to try for 475fsb. Hopefully I should be able to hit 3.8GHz with under 1.3vcore considering my results so far.
Looks like I got 3.7 stable on the EVGA 750i motherboard but I can't take it higher cause I reached hte limits of the motheboard.
CPU Voltage 1.4 (vdropp 1.33 - 1.37)
FSB Voltage 1.4
Northbridge Votlage 1.55
South Bridge voltage 1.35
Memory Voltage 2.1
Memory settings 5-4-4-12 Sync and Linked at 925Mhz
Rated FSB 1851Mhz
FSB 463 MHz
C1E Disabled
PCIE at 102MHz
GTL 0,2,3 = +0.65mv
GTL1 = +0.90mv
Seems to be completely stable and no more lockups. I can get 3.8GHz to boot into Vist and I can do things but it never passes Prime95. My guess is I need more NorthBridge voltage but 1.55 is max I can do. Temps are real low as you can see and everything is Aircooled.
I have also Q9450, batch L814B464..on P5E3 premium, Air case Thermaltake kandalf.
On the box is writen that max. 1.25V..
At 3,6Ghz vcore on auto is 1.32..I have raised only FSB and multiplier to 8, other on AUTO.Is that not to high?
Also can somebody who's knowlegde is better than my, say an opinion about this batch of CPU?
10x guys..
Rampage Formula
4x2GB G. Skill
8x475
Vista Ultimate x64
http://www.hostthenpost.org/uploads/...3d1d8edc0b.jpg
that's where i'm at with my 780i, although it takes me more vcore - 1.440V :(Quote:
Originally Posted by Lothos
I seem to only be able to squeeze 3.2ghz out of mine. I believe the 8gb memory is really limiting me. As soon as I push the FSB over 400 I run into immediate problems -- no matter how much I increase mch and fsb voltages. Even as little as 410 -- it struggles to boot into windows and fails prime95 in the first 30 seconds of running. Perhaps I'll try to take out the extra 4gb memory and see if that makes a difference.
What is the timing on the memory? Can you tell me who makes the memory, what are the specs for it and what is the voltage recomended by the company who makes them?Quote:
Originally Posted by Wolken007
Q9450 // Gigabyte EP45-DS3R // Mushkin Ascent PC2-8500 // Seasonic M12 PSU // 8800GT // Noctua NH-U12P
3.92 ghz = 8x 490 fsb
I am running on too high vcore right now, 1.48 BIOS -- :eek: -- results in 1.44 idle / 1.36-1.39 load (CPU-Z reading). I may be able to drop it down a bit, but this isn't an every day OC, just wanted to see what I could get stable without going too far out of spec. Temps are a bit high but still acceptable at 490: 44-46C idle and 63-67C load (core 0/1 are hot, core 2/3 are almost always 5C cooler).
http://nofxwiki.net/f/fsb-490-%203.92ghz-stable.jpg
http://valid.x86-secret.com/show_oc.php?id=390599
My settings for 490, which aren't minimums just what I used to be certain it was stable:
vcore: 1.48125
vtt: 1.32
pll: 1.61
mch core: 1.34
vdimm: 2.06
all other voltages Auto. Tweaking GTLs/reference voltages doesn't seem to be as essential on this board as it is on some Asus boards. However, tweaking the clock skews was vital to me getting these high clocks, and allows me to get 3.5-3.6 with nearly stock voltages (vcore still needs some love). Memory is 1:1, as I couldn't run these high speeds unlinked. 333 latch/strap, 2.00B. CL5-5-5-15.
I can't wait to see what my E8400 can make this board push.
http://www.newegg.com/Product/Produc...82E16820227195
This is the memory I have. I have tried lowering the fsb and running the ram at 375x2.40 at 5-5-5-15 for 900mhz on the memory and it ran just fine. I have tried pumping up to 2.1v through it without any luck. Do you think the NB is just too heavily stressed with all dimms filled at anything over 400fsb? I have even tried raising the mch voltage +0.3 (board max) with little luck.
The timing on that memory is actually not that good. It is 5-6-6-18, the higher the numbers the slower the memory but sometimes that can be compenstated with higher voltage. If you need the memory to run faster try setting in the bios the memory to 6-6-6-24 and set the voltage to 2.0 volts which is the max it will take according to the specs on NewEgg. If I had to guess that will get you to a max around 1000 MHz on that memory.Quote:
Originally Posted by Wolken007
THis is the memory I have
2x2 Corsair
2x1 Corsair
I don't think the speed/timings of the memory is the problem. I know the memory runs faster than it is rated. I have had success running it at up to 5-5-5-15 @ 1000mhz at only 1.9v. What I was talking about is the stress that having 8gb memory puts on the northbridge. As I suspected, when taking out a 4gb pair, I was able to push the fsb more without having to make crazy adjustments to the voltage. Now I just need to decide if I want 3.2ghz w/ 8gb memory or 3.6ghz with 4gb memory.
I am honestly not sure.
I'd go the 3.6Ghz option :D
I did some benchmarking today in some of the games that I often play and there was no difference between the two. The only game that had any difference was Crysis, and it was only in the minimum fps department due to the 8gb memory. I guess for now I'll leave the 8gb in place since I use this PC only for gaming. I am going to be ordering a P45 chipset board next week. Any suggestions on which I should get? I was thinking the P5Q-E.
how high has anyone taken the q9450 to?
im having problem getting to 3.8ghz its prime stable but the motherboards bios is kinda flaky it will sometimes boot sometimes not at 3.8 because im stuck at mem ratio of 1:20 cant select 1:1 for some reason dont know why used the latest beta bios still no go so currently using 450fsb @ 3.6ghz
hi guys, tomorrow i'm going to receive a Q9650 E0, ES... i want to try for a daily OC, i'm still into 65nm but i've heard that with these 45nm i have to pay attention to the CPU PLL (i've read 1.54 in bios at max), CPU GTL (67x on 65nm, 63x max on 45nm), and FSBT (less than 1.30 for 45nm). I have a Rampage Formula motherboard, is this all correct?
thanks in advance. :)
Q9450 on sale at www.microcenter.com for $279.99.
Go here:
http://microcenter.com/single_produc...uct_id=0287628
Can one please tell me the maximum 24/7 safe voltages?
Vcore:1.4
Vtt:1.4
NB:1.6?Is this correct?
man, that's such a shame, $280 that's round about 178 euros, but the cheapest q9450 here is around 236 euros...Quote:
Originally Posted by Telo
i hate it :<
Quote:
Originally Posted by crash2k
Jupiler made a post about Vcore thought it was 1.375 coming from Intel direct. Most common : CPU PLL below 1.6 and FSB Term voltage certainly below 1.4 for 24/7 use... Like I have told before, go easy on it at first, then start to clock it up... latest CPU's I had didn't degrade one bit, first ones went straight up to 4-4.5Ghz and they degraded after a while... but that's my own 5 centsQuote:
Originally Posted by SuporterPoli
i'm running the 9450 on a x48 chipset with a tuniq tower and *extensive* air cooling . what's the likelihood of my bein able to push it to 3.8-4.0, and what voltage settings might i need?
i get it to 3.6 stable at stock voltages, but at 3.7 it sometimes fails to POST.
Q9450 // gigabyte x48 mobo // 4gb mushkin DDR3 1066 RAM // ATI HD4850
(Tuniq | Antec 900 | plenty of fans)
If you can do 3.6ghz on stock voltages then you've got a pretty damn good CPU. Shouldn't take much more voltage to get to 3.8ghz.
damnit. i just RMAd the thing for faulty temp diodes! although i guess that was necessary.
i didnt reaalize 3.6 was that great. there are at least a few people with 3.6 over at EOCF too nuub to voltmod. owell, hopefully the chip i get back will be just as good.
new to this forums, how do i subscribe this thread?
Okey..... first i though that i had my chip rock solid stable,but i was wrong.I't can prime smallFFT +22h without no problem.Two days ago i decidet to run some large FFT (prime torture test) and it failed instantly.
My prime smallFFT settings:
Cpu 1.2750
Vtt 1.280
nb 1.26
cpu pll 1.54
I had to pump my vcore ---> 1.3v , vtt 1.320 and pll 1.56 to get it prime large stable.
http://www.kolumbus.fi/otukka/Q9450/...rime_large_14h
I was surprised when I saw people using small FFT fro stability test. I always start with Large FFT to test a clock, then blend to test compatability with mem.
Large FFT tests the stability of the max heat and power the cpu will have to dissipate, so it seems like a natural stability test...
why start with small FFT?
those volt settings sound about right (if not a little low) for 3.8ghz.
Thx i hope i can lower my vtt a bit.I don't feel comfortable running over 1.3v.
out of curiosity, what are you cooling with? 3.8 is pretty good for a 24/7 clock for this chip.
True+noctua S12 in antec 900 case.On load, my temps are 67,65,61,65.
sweet deal. I've got the similar Tuniq, same case. can't wait till my RMA comes back =].
3 cores on my other chip read about the same temps at at load, but core 0 would be at 80+. Did the movement test and found out they were all faulty, so it's...in the shop.
Quote:
Originally Posted by Pontus
Very nice ;) What's your CPU VID?
Been a while XS! Vacations are too short.
Well water couldnt net me my 4Ghz on this board so I said FAK it and sold the water to a friend and bought a TRUE. New settings in my signature!
Anyone know what I should expect from a chip with batch L813B225? Tried searching for it, but couldn't find anyone with the same batch. Just got it and I am in the process of putting my system back together, still have a few parts coming. Anyway, just wondering what I should expect. I get my Q6600 to 3.725Ghz 24/7 and I will benching the two against each other at 3.6Ghz. Debating on ditching quad core for an E8500 since I only do video conversions with Divx 6.8 and the rest is gaming. I know the E8500 can do 4Ghz and then some.
Any thoughts?
Thanks,
Utnorris
yeah, but you get a free gas card with your purchase at newegg :rofl:Quote:
Originally Posted by Telo
Lapped TRUE and 3200 @ stock volts gives funky temps :( Looks like a part of my IHS is seriously off :shakes: Checked the cooler and there's an imprint on one of the sides so a bad IHS. RMA it to Holland (again 15€ down the drain) or just lap the sucker?
I had a problem of a similar ilk, albeit much more severe.
My core 0 was reading temps consistently 20C higher than the other cores, and the other cores were reading suspiciously high as well. Did that nifty movement test and they were all out of range, and core 0 was a 26, far out from the others. Had to RMA it to newegg (at no cost to me though =] ).
I've been hearing a lot of stories like this, so i deduce that yorkfield quad cores (or maybe just q9450's) often have faulty diodes. But unless yours is getting hugely offbase temps at load, I'd save the money. Although, you might not want to lap it, just in case.
Try clocking it higher, say, 3.6ghz, and see how it does there. But, I guess, this all depends on where you plan on running it 24/7
Well I had to downclock my overclock from 3.7 GHz to 3.6 GHz. I think I figured out the main issue. It is the motherboard above 3.6 GHz it gets lost and needs allot of power to run. When I downgraded to 3.6 GHz the difference in voltages to get it running are simply ridiculous which is why I suspect its the MB and needs a BIOS upgrade (already have latest). Here are my spects to get 3.6 GHz stable.
CPU Voltage 1.35 (vdropp 1.23 - 1.28)
FSB Voltage 1.25
Northbridge Voltage 1.3
SouthBridge Voltage 1.2
Memory Voltage 2.1 volts (5-4-4-12)
GTL (0) +0.25mv
GTL (1) +0.50mv
GTL (2,3) Auto (stays at 0)
When I had it stable at 3.7 GHz I used the following voltages.
CPU Voltage 1.40 (vdroop 1.30 - 1.36)
FSB Voltage 1.45
Northbridge Voltage 1.55
SouthBridge Voltage 1.35
Memory Voltage 2.1 volts (5-4-4-12)
GTL (0,2,3) +0.65mv
GTL (1) +0.90mv
I just felt the extra 100MHz was simply not worth that much of a voltage difference and degrading the equipment. This leads me to believe it needs a BIOS update. I am currently using the EVGA 750i.
Well it just gets hotter I guess with core0 always 10 degrees hotter :shrug:Quote:
Originally Posted by Slovnaft
But it can't be a sensor that gets jammed that doesn't explain the seriously off temps at core0 :( TRUE is lapped with a 1200 grit going to use my 2000 maybe something went wrong whilst lapping? But that would explain 2 cores hotter but not 1 I mean they share a die ffs..
Loaded @ 3.2Ghz Prime 4 cores max heat setting
honestly, try the movement test again, it looks to me like your core 0 diode is faulty. that's the only thing i can think of and it seems to be happening a lot to quad cores.
it's been such a long day, i don't even remember making that other post =].
EDIT:
Oh, and you may just have a lesser chip.
not the end of the world though...load temps aren't that bad. you could probably go to 3.4 without much worry.
well i just got my q9450 today and dropped it into my old setup batch L818A757 haven't done anything major default bios vcor is 1.115 runs a 3.2ghz stock and on my old water cooling (storm+120mm rad) idles in teh high 40s on a bad mount coretemp says VID of 1.0375 that accurate at all?
coretemp isnt rated for 45m quads.
for temps use RealTemp.
anf CPUZ is fine for volts far as i know.
Quote:
Originally Posted by chance
Coretemp 0.99 shows 1.2250
My Vid is 1.125
will 1.5volts - 1.6vcore kill the q9450? even with proper cooling? whats the max safe temps on this?
It will cause the chip to degrade at those high volts, not immediately kill it though.
I'm thinking it is your memory. What memory ratio are you running or speed. I'm 1 to 1 ratio at 475MHz on good memory (running below rated speed).Quote:
Originally Posted by jadeiceman
nForce SPP 1.45v
nForce MCP 1.50v
FSB 1.30v
HT nForce SPP<->MCP 1.20v
Aux 1.50v
Memory 2.075v (rated at 2.1volt)
CPU 1.375v
Upping cpu voltage can strengthen the quality of the signal to the northbridge and probably clean up the signaling to the memory. I've seen a little more memory voltage can make up for a lot of cpu voltage. personally i wouldn't go beyond the voltage rating on the memory.
Okay... my chip is 22h smallFFt,14H largeFFT and 10h Blend stable.Now i can consider this chip is stable.
http://www.kolumbus.fi/otukka/Q9450/blend.jpg
According to the website version .99.1 does support 45nm chips. I do believe Realtemp is a better program, but it has to be calibrated otherwise it will be off too, by as much as 10 degrees.Quote:
Originally Posted by Slovnaft
Utnorris
so real temp need to be calibrated?
can you teach me how to calibrate it or some links please thanks
This is correct. The newest version of coretemp however has been changed to 95C tjmax just like it is on real temp.Quote:
Originally Posted by Utnorris
shows a tjmak of 105*C for mine in 0.99.1
is 1.15 a low VID or that normal on these chips (my first 45nm chip)
Hmm.. maybe you have to set it? I dont remember doing that though.
ah yeah forgot to look in settings, there is an offset so is 95* the "standard" to use for the q9450s?Quote:
Originally Posted by jas420221
P5Q Deluxe
Q9450
TRUE 120 Black+Scythe 2x120 SlipStream
Mushkin 996599 (2x2GB)
FPS EPSILON 700W
CM 690
http://img528.imageshack.us/img528/8...4031qe0.th.jpg
1.296v - 3928 MHz:
http://img520.imageshack.us/img520/2938/q8fa4.th.jpg
1.288v - 3928 MHz:
http://img520.imageshack.us/img520/4694/q11kh8.th.jpg
OCCT:
http://img520.imageshack.us/img520/8620/q10lm1.th.jpg
1.280v - 3936 MHz
http://img520.imageshack.us/img520/5230/q13dh9.th.jpg
1.37v - 4000 MHz
Cinebench - 10
http://img520.imageshack.us/img520/3781/q6xm9.th.jpg
Nuclearus Multi Core
http://img520.imageshack.us/img520/6633/q5fr4.th.jpg
Lightsmark 2007
http://img520.imageshack.us/img520/9342/q1ot0.th.jpg
wPrime
http://img520.imageshack.us/img520/5051/q4uv9.th.jpg
3DmarkVantage
http://img520.imageshack.us/img520/4931/q2xl0.th.jpg
Please ppl put also your vtt(fsb) voltage along with your vcore. These quads require more vtt than vcore. Vtt is more interesting.
3928 - 3936 MHzQuote:
Originally Posted by SuporterPoli
CPU Voltage: 1.281
CPU PLL Voltage: 1.48 -50
FSB Termination Voltage: 1.48-50
And how about the general warning that going over 1.4 FSB Term voltage or CPU VTT 24/7 could degrade and even kill ya 45Nm CPU...
So 1.4 and below we are good for VTT and PLL?
Well my comment is still based on the anand review where they smoked one of their Qx CPU's. They got a reply from Intel that it was the high VTT that killed it... as they ran it at 1.45 and advised the readers to stay below 1.4 (if possible) , all depends ofcourse on ya cooling, mobo setup, etc... some users punish their 45nm CPU's at 1.45 volt daily and have zero degredation while others run below 1.4 and have experienced a little degredation...
Just my 5 cents, we are at extremesystems, but not everyone swaps their CPU's at 6 months... so for those that want to keep it a long time I would back down and enjoy it...
Well, Im on water (CPU and Vid card). NB is stock Asus cooler...I want 4Ghz stable, and I would like to keep this CPU for at least a year.
Quote:
Originally Posted by Pontus
what are your volt settings for this clock?
are you keeping this as your 24/7 clock?
Temps?
Max recommended vtt is 1.4 and vcore 1.375. Also a lot of people say that 1.35-1.4 vtt is already cpu killing teritory. My personal maximum is 1.35 for the vtt voltage.PLL can be 1.5-1.6 but not over 1.6.Quote:
Originally Posted by jas420221
I have posted my settings already in this thread,but here they are again:Quote:
Originally Posted by Slovnaft
cpu vcore 1.3v set in bios (load line calibration enabled) idle 1.288 and load 1.298 according cpu-z
cpu pll 1.56v
cpu vtt 1.32v
nb 1.26v
ram 1.98v
and i'm running with these settings 24/7
my load temps are: 67,64,62,64 (OCCT smallFFT) idle 39,32,31,39 according Real Temp 2.60