|
The Official OC (OverClocking) Thread!
|
|
|
NO Fanboy comments needed
|
|
Red_Maw
Senior Member
|
23. June 2007 @ 13:37 |
Link to this message
|
|
What are the good free system stability tests other than Prime95?
Thanks.
|
|
Advertisement
|
 |
|
|
Senior Member
|
23. June 2007 @ 14:04 |
Link to this message
|
|
AfterDawn Addict
|
23. June 2007 @ 14:18 |
Link to this message
|
sammorris,
Quote:
After playing a game, Everest (and the BIOS reads the same as everest) reported 44ºC and the heatsink was 80ºC.
I'm assuming the 44C is the CPU and the 80C is the Northbridge/MB. If that's the case somethings not right! I used to get squeamish when my MB would hit 47C on the D-940. Gina's P5N-E never got close to 40C as I think the highest it ever got was 37C, and I didn't even mod the heatsink as it wasn't going to get any cooler than room temp anyway!
Play with your fan speeds a bit as it's not going to last very long at that temp if it is the MB! 80C for the heat sink if it is the Northbridge is firmly in the danger zone!
Clock On,
theone :}
GigaByte 990FXA-UD5 - AMD FX-8320 @4.0GHz @1.312v - Corsair H-60 liquid CPU Cooler - 4x4 GB GSkill RipJaws DDR3/1866 Cas8, 8-9-9-24 - Corsair 400-R Case - OCZ FATAL1TY 550 watt Modular PSU - Intel 330 120GB SATA III SSD - WD Black 500GB SATA III - WD black 1 TB Sata III - WD Black 500GB SATA II - 2 Asus DRW-24B1ST DVD-Burner - Sony 420W 5.1 PL-II Suround Sound - GigaByte GTX550/1GB 970 Mhz Video - Asus VE247H 23.6" HDMI 1080p Monitor
 |
|
Red_Maw
Senior Member
|
23. June 2007 @ 14:41 |
Link to this message
|
|
I was running Prime95 and my computer started to make this middle-high pitch sound. I heard that it could be coming from the motherboard or the cpu, either way is it bad or just something I have do endure while running prime95?
Thanks.
|
AfterDawn Addict
4 product reviews
|
23. June 2007 @ 15:18 |
Link to this message
|
|
Perhaps. It can happen with normally functioning components, a lot of X1900s do it (mine does) during games, particularly openGL ones.
Russ, 44C is what everest reports, and 80C is how hot the heatsink actually is.
|
AfterDawn Addict
|
23. June 2007 @ 17:48 |
Link to this message
|
Sam,
Quote:
Russ, 44C is what everest reports, and 80C is how hot the heatsink actually is.
Are you talking about the CPU, Motherboard or Video card here? How did you arrive at the 80C for which heat sink??? I know video tolerates higher temps, as mine idles around 48-50C! About the same as it was before I overclocked it!
Clock On,
Russ :}
GigaByte 990FXA-UD5 - AMD FX-8320 @4.0GHz @1.312v - Corsair H-60 liquid CPU Cooler - 4x4 GB GSkill RipJaws DDR3/1866 Cas8, 8-9-9-24 - Corsair 400-R Case - OCZ FATAL1TY 550 watt Modular PSU - Intel 330 120GB SATA III SSD - WD Black 500GB SATA III - WD black 1 TB Sata III - WD Black 500GB SATA II - 2 Asus DRW-24B1ST DVD-Burner - Sony 420W 5.1 PL-II Suround Sound - GigaByte GTX550/1GB 970 Mhz Video - Asus VE247H 23.6" HDMI 1080p Monitor
|
AfterDawn Addict
4 product reviews
|
23. June 2007 @ 17:58 |
Link to this message
|
|
The chipset. I measured it with a thermal probe on my multimeter, and touching it I think it's right (I can't without burning myself).
|
Senior Member
|
23. June 2007 @ 18:14 |
Link to this message
|
|
a hot heat sink means it doing its job. but that is a bit too hot for my liking, my nb hs is quite cool to touch but it has a 60mm fan spinning at 4.5k rpm on top of it.
i plan to get a heat pipe solution for it when i get a new case. i will also set my 7 pro so it sucks air thru it before blowing it out the top fan. the only thing that worries me about having it setup this way is how well the power regs for the cpu will be cooled. if push comes to shove i could always get some little heatsinks that go on the ram of a g card if needed.
|
AfterDawn Addict
4 product reviews
|
23. June 2007 @ 18:28 |
Link to this message
|
|
Indeed. Arctic Silver V would surely only make the heatsink hotter still, but keep the chipset cooler. If it's already at 80C, I can't possibly believe that 44C is the correct temp for the chipset.
Note, when gaming I can touch my graphics card heatsink (it's fanless and yet uses well over 100W) and I can touch my freezer 7 Pro heatsink even when the fan's low, but I have to withdraw my finger from my chipset heatsink within half a second or it burns.
hehe, like you my GPU temps are fine except the Vregs, before i switched my side fan to exhaust they got over 100C in games. While the GPU may be able to stand that, I'm not sure what the VRMs can take so I played it safe. They now max at around 86.
This message has been edited since posting. Last time this message was edited on 23. June 2007 @ 18:30
|
AfterDawn Addict
|
23. June 2007 @ 19:10 |
Link to this message
|
sammorris,
Quote:
Note, when gaming I can touch my graphics card heatsink (it's fanless and yet uses well over 100W) and I can touch my freezer 7 Pro heatsink even when the fan's low, but I have to withdraw my finger from my chipset heatsink within half a second or it burns.
You might consider doing a little experiment and set it to 1.8GHz and see how hot the chipset gets then. Something is not right, that's for sure! That's 176F and way too hot. None of the 3 P5N-Es I had here were uncomfortable to touch. Could be anything from a cold solder joint to a faulty component, but something for sure isn't right!
Clock On,
theone :}
GigaByte 990FXA-UD5 - AMD FX-8320 @4.0GHz @1.312v - Corsair H-60 liquid CPU Cooler - 4x4 GB GSkill RipJaws DDR3/1866 Cas8, 8-9-9-24 - Corsair 400-R Case - OCZ FATAL1TY 550 watt Modular PSU - Intel 330 120GB SATA III SSD - WD Black 500GB SATA III - WD black 1 TB Sata III - WD Black 500GB SATA II - 2 Asus DRW-24B1ST DVD-Burner - Sony 420W 5.1 PL-II Suround Sound - GigaByte GTX550/1GB 970 Mhz Video - Asus VE247H 23.6" HDMI 1080p Monitor
|
AfterDawn Addict
4 product reviews
|
23. June 2007 @ 19:22 |
Link to this message
|
|
Hmm, interesting, because I'm told the same was true of a friend's A8N-SLi premium, the heatsink seemed much hotter than the software dictated.
|
|
docTY
Member
|
23. June 2007 @ 21:09 |
Link to this message
|
fasfrank
Quote:
Thanks again Doc and everyone, I'm learning a lot and having a good time applying it. I'm going to start OCing the new E6700. I'll stay below 1.45 volts indicated and below 50°C and see what that will get me.
thanks for that extra info on the temp readings and monitoring software, i guess there really IS nothing wrong w/ your first E6700 :)
like i mentioned before, i USUALLY trust Everest w/ my temps and such, but since i use so many beta versions (i'm actually one of a group of 100 of Lavalys's current beta testers for Everest since i have access to various cpus, mobos, chipsets, platforms, etc.); but that's not the point...
good luck w/ your overclocking, you honestly MAY be somewhat disappointed in what you can achieve keeping your vcore @ 1.45v and below, (going based on what your vcore was before to hit the higher OC's) BUT you will have the satisfaction of knowing that your system isn't going to fail or melt-down anytime soon w/ those voltages~ :)
it's a compromise of sorts, i always take into consideration the longevity of MY OWN components when i max them out, if i see any indication that i could be GREATLY shortening the lifespan of any particular component, i take the volts down a few notches and just settle with a "less than" ideal overclocking expectation~ :) (ie. high end memory, VERY few companies will warranty voltages on DDR2 sticks over 2.4v, so that is where i "max" out as the limit during testing, also for cpu's...when i reach a temp. that is uncomfortable to ME based on rated specifications by Intel or AMD, i look for more powerful cooling solutions OR take it down a few notches to ensure i don't fry my chip anytime soon) :)
p.s. but i HAVE been known to pump out 1.65v+ on vcore, albeit on water cooling, just to stabilize my highest OC'in attempts :)
but NOT for prolonged periods of time naturally. i hit 4.0ghz on my own E6600 with a rather low vcore (considering i have an earlier stepping/binned chip) of 1.5v to POST and boot into Windows at that speed, but it was not stable enough to run my usual barrage of "tests"...testing-wise i completed just enough to qualify and satisfy the requirements for entry into the 4ghz club on a C2D, but anything more than that it was a no-go :(
docTY
|
AfterDawn Addict
|
23. June 2007 @ 23:55 |
Link to this message
|
sammorris,
Quote:
Hmm, interesting, because I'm told the same was true of a friend's A8N-SLi premium, the heatsink seemed much hotter than the software dictated.
If I remember rightly, someone from the other forum had that problem with the same MB, I'm just not sure whether it was the Premium or the Deluxe. I know I offered to send them an Asus chipset cooler for it that was sent to me by mistake by Asus as theirs didn't come with one! I know it has a rep as a hotter running chipset. I also know the 6100 chipset runs a lot cooler as I have one in my 4000+ A64! 36C is max in a 29C room under load running OCCT!
Clock On,
theone :}
GigaByte 990FXA-UD5 - AMD FX-8320 @4.0GHz @1.312v - Corsair H-60 liquid CPU Cooler - 4x4 GB GSkill RipJaws DDR3/1866 Cas8, 8-9-9-24 - Corsair 400-R Case - OCZ FATAL1TY 550 watt Modular PSU - Intel 330 120GB SATA III SSD - WD Black 500GB SATA III - WD black 1 TB Sata III - WD Black 500GB SATA II - 2 Asus DRW-24B1ST DVD-Burner - Sony 420W 5.1 PL-II Suround Sound - GigaByte GTX550/1GB 970 Mhz Video - Asus VE247H 23.6" HDMI 1080p Monitor
|
Senior Member
|
24. June 2007 @ 05:58 |
Link to this message
|
Originally posted by sammorris:
Indeed. If it's already at 80C, I can't possibly believe that 44C is the correct temp for the chipset.
i have been told that the mb temp gets its readings from a sensor just below the first pci-ex slot, so this might be the reason for the diffference in temps.
|
AfterDawn Addict
4 product reviews
|
24. June 2007 @ 07:00 |
Link to this message
|
|
That's interesting, and pointless. Why would they do it?
|
Senior Member
|
24. June 2007 @ 08:01 |
Link to this message
|
|
tbh im not sure, i guess it may be down to space (or lack of?) around the nb/cpu area, maybe to save a few coppers?
i do agree tho it is somewhat pointless.
|
AfterDawn Addict
|
24. June 2007 @ 08:23 |
Link to this message
|
Doc,
All my friends and relatives get a strange glassy-eyed look when I talk about computers and overclocking. They're just glad I'm in to fast PCs and not racing motorcycles. So I've come here to babble on about it, to the only folks I know will pay any attention and actually have something constructive to say.... Hopefully :)
So....
I compared the raw hex core temperature data from CrystalCPUID to Everest, SpeedFan and CoreTemp and they all were identical for the E6700 as well as the E6600. I did end up adding offsets in SpeedFan for my QX6700.
You are right about the CPU voltage limit restricting the OC. I used 1.45 indicated at full load using Everest as my limit. With the 680i board I get about a .4 volt difference between what I set and what is indicated. Vdroop with the new PSU is .2 volts. So my current setting is 1.5125 volts.
This gets me 3.5 GHz at 47°C max, Ambient is 22°C.
The temp is the maximum reached during a two hour Orthos test. It only goes that high during a specific part of the test the rest of the time it stays about 7°C lower.
Typical:
After the Orthos test I ran 3DMark06:
This score is mainly a function of overclocked 8800GTS cards in SLI. (660 and 962).
So I have a few conclusions....
1. I have a fast computer with a moderate OC.
2.E6600's are easier to OC than E6700's. Less voltage is needed for the same speed, however temps are a bit higher.
3. If a person owns a 680i board and wants strange temperature readings, simply disable C1E as well as Virtualization Technology(Vanderpool Technology). Otherwise, leave them enabled. Regardless of their intended functions, they do make the temperature readings more reliable at idle as well as at load.
4. OCing my quad core is impossible if I tried to use the same temperature limits as the E6700. I'd have to add 15°C to my imposed 50°C limit. The thing idles at 39°C and full load is about 57°C at stock settings.
Cheers,
Frank
EVGA 680i board
C2D E6700
2 Gb Corsair C5D RAM
2 WD 150 SATA Raptor HDDs, not in RAID.
2 ASUS 8800GTS 640 Mb Video cards
Enermax Galaxy 850 Watt PSU
Thermalright 120 Extreme Heatsink with Scythe 120 mm fan.
Arctic Silver Ceramique
Gigabyte Aurora3D 570 Case
Lite-On SATA DVD burner
Asus SATA DVD-ROM
A floppy drive and other various bits.
|
AfterDawn Addict
|
24. June 2007 @ 17:26 |
Link to this message
|
|
fasfrank
i will have to agree with your observations & conclusions~ :)
1. you sure do and it's all within "safer" settings now, congrats!
2. i've never owned a E6700 personally so i can't compare but i see your point regarding the E6600s, i've built and OC'ed (3) three E6600s...my own, my father's, and a friends...they vary in OC levels: 4.1ghz testing speed on water b4- 3.9ghz 24/7 usage now on AIR, 3.4ghz 24/7 usage, 3.2ghz 24/7 usage (respectively starting with my own, then my dad's, then my friend's) :)
the higher temps associated with the E6600s is naturally due to starting with a LOWER cpu speed and overclocking to same level as a E6700, even with the same vcore, the E6600 is overclocked "higher" percentage wise than the E6700, resulting in higher temps.
3. all of my BIOS settings regarding OC'in have been disabled from the get-go~ i didn't want to limit myself in any way, shape, or form so i said screw it even if it meant inaccurate temp readings :)
4. when OC'in quad-cores, you are in a totally different ball-park regarding temps...even on my own Q6600 on water, the gap in temps compared to my E6600 is still approximately 10C-12C higher! (idle + load) even though it IS overclocked, well, both of them are, the limit for temperatures NEEDS to be raised when dealing with the quads...it wouldn't surprise me to see FULL LOAD temps on quads in excess of 65C (OC'ed of course) but i would become uncomfortable if it ever broke the 70C mark...is your QX6700 stock at this point? or have you OC'ed it a little?
all in all, you should be very proud of your build there...it's much less dangerous "going fast" on a PC than racing your motorcycles~ i never really got into bikes, but have always admired those who race in circuits similar to the ones you've participated in. :)
a good friend of mine raced almost every weekend w/ his old 2006 model GSX1300R, he's thinking bout selling it and getting the 2007 Hayabusa instead...do you think that would be a wise decision? or just keep what he has now since it's been good to him over the past year or so...? he's much like me in always "keeping his tech. up to date" even though we're dealing with 2 completely different techs. here :) i know nothing about sportbikes and he's much the same way about computers, so i figure it's an even trade to educate each other~ LOL
have a good one
docTY
Recommended Media:
Taiyo Yuden 4x dvd-r TYGO1/ 8x dvd-r TYGO2/ 8x dvd+r YUDEN000T02/ 16x dvd+r YUDEN000T03
Verbatim 8x dvd+r MCC003
Verbatim dvd+r DL (MKM001)= flawless no compression backups
"Do Yourself A Favor, Use The Good Stuff
TY & Verbs 4 Life~ :)" ~docTY~
"Its better to be quiet and appear stupid, then to open your mouth and remove all doubt."
I am always prepared to recognize that there can be two points of view - mine and one that is probably wrong - John Gorton
|
AfterDawn Addict
|
24. June 2007 @ 18:19 |
Link to this message
|
fasfrank,
Quote:
3. If a person owns a 680i board and wants strange temperature readings, simply disable C1E as well as Virtualization Technology(Vanderpool Technology). Otherwise, leave them enabled. Regardless of their intended functions, they do make the temperature readings more reliable at idle as well as at load.
I use both C1E and the CPU Smart Fan Control and between the two of them, they keep my temps lower than if I just ran the CPU fan at full speed! I was just told that on some motherboards, in order to raise the fsb, you have to disable C1E! Bummer!! It doesn't affect the speed at all! Idles at 2.2GHz and jumps to 3.276 when needed!
Mine idles at about 27-28C and my MB temp idles at 35-37C, with a fairly hefty OC of 82%! My CPU fan runs at about 800-1000 rpm at idle and most times never exceeds 1600 rpm! I set the Silverstone 120mm case fan to 1600 rpm and it all works out fine!
Glad you are having fun!
Clock On,
theone :}
GigaByte 990FXA-UD5 - AMD FX-8320 @4.0GHz @1.312v - Corsair H-60 liquid CPU Cooler - 4x4 GB GSkill RipJaws DDR3/1866 Cas8, 8-9-9-24 - Corsair 400-R Case - OCZ FATAL1TY 550 watt Modular PSU - Intel 330 120GB SATA III SSD - WD Black 500GB SATA III - WD black 1 TB Sata III - WD Black 500GB SATA II - 2 Asus DRW-24B1ST DVD-Burner - Sony 420W 5.1 PL-II Suround Sound - GigaByte GTX550/1GB 970 Mhz Video - Asus VE247H 23.6" HDMI 1080p Monitor
This message has been edited since posting. Last time this message was edited on 24. June 2007 @ 18:21
|
AfterDawn Addict
4 product reviews
|
24. June 2007 @ 18:30 |
Link to this message
|
|
70C is a little hot for a CPU, but for me the temp is academic. 56C is the limit, and above that I start to get stability issues.
|
AfterDawn Addict
|
24. June 2007 @ 21:08 |
Link to this message
|
|
theone,
I set the bios back to default. Installed the new E6700. Booted and ran Everest to check temps. I took notes as a enabled and disabled these two features, checking temps at idle as well at 100% load with Orthos.:
Default settings:
CPU: 31
Core 1: 34
Core 2: 36
Not inverted, Stable.
Disable C1E Enhanced halt state:
CPU: 30
Core 1: 33
Core 2: 35
Idle temps fluctuating between cores, no inversion.
Enable C1E and disable Virtualization Technology:
CPU: 30
Core 1: 27
Core 2: 30
Core/CPU temps invert occasionally.
Disable both C1E and VT:
CPU: 28
Core 1: 35
Core 2: 34
Nothing wrong here.
Run Orthos:
CPU: 38
Core 1: 36
Core 2: 35
Inverted CPU/core temps, rather low temps but only one minute into Orthos.
Enable VT:
At Idle:
CPU: 24
Core 1: 27
Core 2: 28
Orthos 1 minute:
CPU: 37
Core 1: 40
Core 2: 36
Average Core temp above CPU temp, no inversion. Reasonable numbers.
Enable C1E:
(Forgot to write down idle temps but were OK)
Orthos 1 minute:
CPU: 41
Core 1: 42
Core 2: 40
Almost inverted but at this point I'm OK with these numbers.
I concluded that by leaving C1E and V.T. on I get reasonable numbers at idle and under load.
C1E causes idle temp fluctuations that get worse when overclocked.
VT disabled causes Core/CPU temp inversions.
As you mentioned before CPU temp should be lower than core temp. Apparently the tolerance on this is 1-15 degrees. I get about a 1 degree lower case temp most of the time so I'm OK with that.
I'll have to look at the fan settings. I haven't changed the default settings in bios. They seem OK though.
I wish someone would tell me the correct way to connect a 3 pin case fan to a 4 pin CPU fan header on the board. It works but never changes speed. I have it plugged into the left 3 pins of the header. Thats where the key on the connector plug allows it to fit. Is that right?
Cheers,
Frank
This message has been edited since posting. Last time this message was edited on 24. June 2007 @ 21:20
|
AfterDawn Addict
|
24. June 2007 @ 23:01 |
Link to this message
|
fasfrank,
Quote:
I wish someone would tell me the correct way to connect a 3 pin case fan to a 4 pin CPU fan header on the board. It works but never changes speed.
Usually the plug has a slot in it to match which pins it's connected to on the MB. I'm not sure about the speeds though! My MB only controls the CPU fan through PWM to match the fan. The other fan plug on the MB is not affected by it at all! There are a couple of other settings for fan control as well, but the PWM is the best by far when you have a PWM fan Like I have on my Arctic Freezer 7 Pro. It gives very linear response to the fans and the change is always on a sliding scale rather than just a sudden burst of speed. I have the speed controller that came with the Silverstone mounted in one of the 5.25" bay blanks. Basically I've set it at 1600 RPM and have never had to change it other than to test different speeds initialy. It all just seems to work the best this way!
The Silverstone can go as high as 2400 and 110 CFM. Mine is blowing about 70 CFM and is extremely quiet doing it. As Sam pointed out it's not too good in an aluminum case as he gets noises I've never heard, that are very annoying. They're on the expensive side at $16 (USD) but on the plus side they come with a controller and are double ball bearing. At 1600 RPM it moves about twice as much air as the stock Cooler Master fan that came with the case did, at about half the noise. At 2400 RPM it would be difficult to even stay in the same room, but if your CPU fan ever goes out, it will keep the CPU cool! Great Fan! Very highly recommended!
Clock On,
theone :}
GigaByte 990FXA-UD5 - AMD FX-8320 @4.0GHz @1.312v - Corsair H-60 liquid CPU Cooler - 4x4 GB GSkill RipJaws DDR3/1866 Cas8, 8-9-9-24 - Corsair 400-R Case - OCZ FATAL1TY 550 watt Modular PSU - Intel 330 120GB SATA III SSD - WD Black 500GB SATA III - WD black 1 TB Sata III - WD Black 500GB SATA II - 2 Asus DRW-24B1ST DVD-Burner - Sony 420W 5.1 PL-II Suround Sound - GigaByte GTX550/1GB 970 Mhz Video - Asus VE247H 23.6" HDMI 1080p Monitor
|
Senior Member
3 product reviews
|
24. June 2007 @ 23:14 |
Link to this message
|
fasfrank,
Quote:
I wish someone would tell me the correct way to connect a 3 pin case fan to a 4 pin CPU fan header on the board. It works but never changes speed.
They install just as described above , using pins 1,2,&3 of mobo socket. However, PWM (Power management control from CPU) will not control 3-wire fans. So you need to if given the option, set CPU Fan in BIOS to DC mode (all ASUS w/ 4-pin PWM headers) to allow DC control of 3rd wire from mobo BIOS. Some fan controllers can except the PWM signal and act on 3-wire fan but w/ not so accurate control. On ASUS mobos w/ good ampage on 4th pin, you can series 2 more PWM fasn to except PWM control from CPU load (ie. 2 case fans).
And that's what's Clockin' like it
Hope that helps
Sony PSP/PS3,
ASUS RAMPAGE II EXTREME(X58) w/ i7 930 DO @ 4.305GHz (205x21 @ 1.323v) 2:8 DDR1680 @ 6-6-6-18
ASUS RAMPAGE EXTREME (X48) w/ Q6600 @ 3.81GHz, 422x9 @ DDR1680 6-7-6-20 @ 1.71v
ASUS CROSSHAIR w/ x2 6400+
This message has been edited since posting. Last time this message was edited on 24. June 2007 @ 23:23
|
AfterDawn Addict
4 product reviews
|
25. June 2007 @ 04:27 |
Link to this message
|
|
The 4th pin varies pulses of current to the fan, 3 pin systems only work by changing the voltage, which of course isn't needed in a 4 pin environment.
|
|
Advertisement
|
|
|
|
Lii
Junior Member
|
26. June 2007 @ 00:42 |
Link to this message
|
|
Overclocking kills the CPU faster ;)
History Of The PSP Firmware(s):
1.5>2.0>2.6(eLoader) Stolen
2.5>3.0>3.03>1.5>3.10OE-A'>3.30OE-A>Interdementional Tuning Data Modification Tool
|
