E6420 Overclock Thread

Thanks for the kind words, glad to see you again, Terc.

1. 3820Mhz. Wow.
2. WHAT fancy Thermal paste? Shin Etsu?
3. I would volt mod my board as well because the vDroop sucks, but eventually, it's probably going to be returned.
The Caps or something on the board has a tendency to squeal like a stuck pig under any type of load....even a minimal load.
If I set to 1.525 in BIOS, I can almost be sure I'm only going to end up with about 1.45v-1.465v. :(

On a brighter note - OCZ got me my replacement Reapers and they were waiting for me when I got home! :D
 
Evoic said:
Thanks for the kind words, glad to see you again, Terc.

1. 3820Mhz. Wow.
2. WHAT fancy Thermal paste? Shin Etsu?
3. I would volt mod my board as well because the vDroop sucks, but eventually, it's probably going to be returned.
The Caps or something on the board has a tendency to squeal like a stuck pig under any type of load....even a minimal load.
If I set to 1.525 in BIOS, I can almost be sure I'm only going to end up with about 1.45v-1.465v. :(

On a brighter note - OCZ got me my replacement Reapers and they were waiting for me when I got home! :D
Click to expand...

Sounds like it's time for an RMA possibly. I'm gonna try for a suicide 1M tonight when I get home.
 
ARGH!

So I had determined I could run 3.4GHz stable by leaving all volts at min except Vcore which was set in BIOS to 1.43V (pre-Vdroop). Orthos 24+ hours. But because my last CPU changed its behavior after a while of burn-in (and thermal paste improves its ability to cool as it sets) I figured I'd re-try to see if I could go a bit higher. Got as high as 435x9, bumping up NB volts +1 notch and Vcore to 1.50V in BIOS (again, before Vdroop so the CPU was getting like 1.47V even before load) and suddenly Orthos goes from failing after a few hours to failing after a few seconds even after a reboot where I haven't changed any settings. Confused, I start backing off to known-good settings... and it's still failing right off. :eek: I bring it all the way down to 2GHz, underclocking it... and Orthos still fails in seconds.

I believe my CPU has shat the bed. :( This is my 2nd one too. Man, I have the worst luck... I figured I was save until at least 1.60V Vcore as set in BIOS. Guess not.

This sucks.
 
E64202040203_9GHz20Temps.jpg

E64202040203_9GHz.jpg


http://valid.x86-secret.com/show_oc.php?id=199183

I will post, more to come...
 
User Name: Mr.White
Stability: 11+ hours Orthos

CPU: E6420
OC: 3400Mhz
Stepping: 6 B2
Week: L652
Code: B309
Vcore: 1.5750 (in BIOS)

Cooling: Air
Cooler brand/model: Thermalright Ultra 120

Motherboard: Asus P5K Deluxe WiFi
Bios Version: 0311
Operating System: XP Pro 32

RAM brand: G.Skill (HZ)
RAM rated speed: 6400
RAM actual speed: 1020 Mhz
Latency: 4-4-4-12

http://valid.x86-secret.com/show_oc.php?id=205504
My6420@3400-2.jpg
 
I've got my e6420 @ 3.2ghz and the temps in TAT is 62c. Is this too high for a C2D? The Vcore is 1.45v I believe. Should I back off or am I fine?
 
Josh123 said:
I've got my e6420 @ 3.2ghz and the temps in TAT is 62c. Is this too high for a C2D? The Vcore is 1.45v I believe. Should I back off or am I fine?
Click to expand...

Hi Josh123. If your 62° is @ idle, then it seems really high.
I go to around 57° at load with Orthos at 3.4GHz with 1.57V.

What cooling do you use?
 
I get worried if I see anything over 55c, even under load, but that's just because I'm overly cautious with my CPU's.


SHORTFUSE.......that 3900Mhz speed is insane.

Details please.
What speed are you stable at?
What voltage are you using?
Have you run SuperPi 1M tests yet at 3900Mhz?
What cooling are you using?

Give us details ! ;)
 
Mr.White said:
Hi Josh123. If your 62° is @ idle, then it seems really high.
I go to around 57° at load with Orthos at 3.4GHz with 1.57V.

What cooling do you use?
Click to expand...

Orthos has been going for almost 2 hours now and my temp is 59c full load according to TAT. I'm aircooling with an Artic Freezer 7 pro and right now it is here at the office which is usually cooler than at my house :(.
 
In my opinion, no discussion/comparison of temperatures is worth anything without also stating the ambient temperatures. Since cooling methods aren't able to hit a specific temp as much as they are able to remove heat (delta) towards ambient temp. Especially when people are trying to compare to each other along differences of <10 degrees.

Unless I'm missing something
 
sremick said:
In my opinion, no discussion/comparison of temperatures is worth anything without also stating the ambient temperatures. Since cooling methods aren't able to hit a specific temp as much as they are able to remove heat (delta) towards ambient temp. Especially when people are trying to compare to each other along differences of <10 degrees.

Unless I'm missing something
Click to expand...

No, that's probably a more accurate description, but at the end of the day - total temperature that the CPU is running the only thing that matters.

The only time your way would be 100&#37; correct is if you were trying to compare HSF vs HSF. Then overall temps don't matter......only the difference in Delta between ambient and CPU temps.

In other words, we could hook up a rig outside in Alaska during winter, but if the CPU is under load and showing 80c + for extended periods of time, that's still not a good thing. ;)
 
I agree that there are absolute numbers that should be avoided (60'C is a good limit). But seems a lot of people get like 53'C, see that someone else is getting 47'F with the same CPU/MHz/HSF and wonder what they're doing wrong. My main point was that getting 57'C or whatever doesn't necessarily mean you have a bad CPU, bad heatsink, bad heat-transfer, bad fan, bad case airflow, or bad OC settings in the BIOS or whatever. You might just have higher ambient temps than the other guy. But this doesn't seem to be brought up often enough.

Anyways, this is sort of off-topic. Sorry.
 
Well just ordered a DDMaze4 680i chipset water block and D-Tech Fusion CPU water block. I think my North bridge is being pushed pretty hard and I think this is part of the evga 680I issues. Also was using a stock thermaltake waterblock that came with my Armor LCS case so maybe the D-Tech Fusion will give me a few degree's also.

Once I get it all setup I have an experiment I plan to try using a Peltier cooler to try for a 4+Ghz OC.
 
Evoic said:
I get worried if I see anything over 55c, even under load, but that's just because I'm overly cautious with my CPU's.


SHORTFUSE.......that 3900Mhz speed is insane.

Details please.
What speed are you stable at?
What voltage are you using?
Have you run SuperPi 1M tests yet at 3900Mhz?
What cooling are you using?

Give us details ! ;)
Click to expand...

Stable @ 3.7GHz running prime 8 1/2 hours.
Somewhat Stable @ 3.8GHz running prime for 2+ hours & reboots under full load.
Not Stable @ 3.9GHz Computer will boot & use some programs np .

3.6GHz 1.575v
3.7GHz 1.600v
3.8GHz 1.625v
3.9GHz 1.650v

Setting for 500FSB

VCore Voltage (This can change on different cpu's & mother boards so add ur own & fine tune.
Memory Voltage 2.35000V - 2.45000V (You should use some type of cooling.
NB Voltage 1.55000V (BEST) for 500MHz Bus Speed
SB Voltage 1.60000V - 1.65000V
1.2v HT Voltage 1.45000V MAX
CPU VTT Voltage 1.55000V (Needed to break 500MHz Bus Speed)

Ok this is was I get,
CPU 100% Load 75c - 82c
CPU Core #1 64c - 70c
CPU Core #2 64c - 70c

Cpu Cooling Zalman CNPS9700
NB Cooling Asus NB Cooling Fan
SB Cooling Asus SB Cooling Fan
8 x 120mm CoolerMaster Fans
 
ShortFuse said:
Stable @ 3.7GHz running prime 8 1/2 hours.
Somewhat Stable @ 3.8GHz running prime for 2+ hours & reboots under full load.
Not Stable @ 3.9GHz Computer will boot & use some programs np .

3.6GHz 1.575v
3.7GHz 1.600v
3.8GHz 1.625v
3.9GHz 1.650v

Setting for 500FSB

VCore Voltage (This can change on different cpu's & mother boards so add ur own & fine tune.
Memory Voltage 2.35000V - 2.45000V (You should use some type of cooling.
NB Voltage 1.55000V (BEST) for 500MHz Bus Speed
SB Voltage 1.60000V - 1.65000V
1.2v HT Voltage 1.45000V MAX
CPU VTT Voltage 1.55000V (Needed to break 500MHz Bus Speed)

Ok this is was I get,
CPU 100&#37; Load 75c - 82c
CPU Core #1 64c - 70c
CPU Core #2 64c - 70c

Cpu Cooling Zalman CNPS9700
NB Cooling Asus NB Cooling Fan
SB Cooling Asus SB Cooling Fan
8 x 120mm CoolerMaster Fans
Click to expand...

That Asus Striker board is unreal. Wow.

I can't set MCP or SB volts higher than 1.4v.
CPU cannot be set higher than 1.6v......even if I set it that high, I'd probably end up with 1.55v or lower because of vDroop.
Your load temps are pretty high, I don't think my chip has ever been north of 55c if memory serves me correctly.
Also, My memory can be set as high as 2.5v, but anything over 2.2v-2.3v on the eVGA 680i motherboards creates problems and can kill your RAM.
I had a set of OCZ Reapers at 2.3v die in less than a month.

Your Mobo is amazing.
I may have to hit eBay and throw some money at one of those.

My board, although extremely good IMO, tops out at under 450Mhz FSB.
 


Volts : 1.3v (lower than stock)

3.2ghz fully stable with ram 1:1 400mhz 4-4-4 2.0v :D

You can see the orthos loaded temps in that pic^^.

Here's TAT temps:

Idle:


Load:


Very good cpu imo :D
 
Still moving up.... back to the Vdroop modded p5b with the new block and fans...

3.720-1.5.jpg
 
OV3RCLOCK3R said:


Volts : 1.3v (lower than stock)

3.2ghz fully stable with ram 1:1 400mhz 4-4-4 2.0v :D

You can see the orthos loaded temps in that pic^^.

Here's TAT temps:

Idle:


Load:


Very good cpu imo :D
Click to expand...

You mention that your vcore is lower than stock (1.3v) however, CPU-z is known to report incorrect voltages for the E6420, can you show your voltage with speedfan or everest?
 
I was poking around over at xtreme and saw a post from a guy who just picked up 6420 from the egg with a batch# 3717A922. Other than its coming out of the Costa Rica fab, I can't find much else about this batch.

Anyone else have this batch or seen any results of how it OC's?
 
Here is another update with my 6420....Im currently priming 3760, this was last nights run though.

3735-1.5375v.jpg
 
hrmmm seems stable enough.. on water cooling 650sli-ds4. very large gap between idle and load, any idea why? i have an e6300 that only varied 10 degrees @3150. i am running at 1.55 v shows 1.53 in.
6420 sla4t malay
2.13ghz/4m/1066/06
l632f557
water cooling
just set this up and went for it... but orthos load is way hot anyone else having this issue or seen that stepping before ordered from newegg pack date 5/28/07 not complaining but never seen a pcu vary 35 degrees under load
 
Hi everyone, I'm new here (both to this board and to C2D), previous system was a 939 x2.

Here's my E6420 L649G593 pack date 04/03/07 using a Asus P5B-Deluxe Rev.1.10g, Scythe Mine B, and Crucial Ballistix ddr2-800 c4 with Ultra Xfinity 600W PSU.

Managed to get it 6 hours Orthos stable (ran a bit longer than the pic below before I stopped Orthos). Using 1.5375 cpu volts (1.5 idle/1.48 load), 1.45 on the NB and auto for rest. Temps are very good, core temps under TAT are 31 idle and 54 max under orthos load.

3.5ghz is rock solid stable for me, but I wonder if anyone here can advise whether I should back off for daily use, I'd like preserve the chip's lifespan as much as possible while maintaining the maximum overclock. Temps are very good but are these voltages too high?

I'm pretty sure I can get 3.6 stable but it'd involve cpu volts of of 1.55 to 1.6; temps should be around 60 (but too close to the danger zone so I probably won't pursue that). Can you tell me if degradation of the cpu is purely a heat related issue or does the voltage play a role as well? For example what's the lifespan of those water cooled chips running on high voltages? Thanks!

I love this chip, btw, a really sweet overclocker compared to my old AMD system.

35ghzoverclockorthosstajg0.jpg
 
Voltages of 1.5v and up are probably dangerous for this chip. I posted some wicked overclocks earlier using as high, but never more than 1.6v I now experience some cold boot issues with my chip.

I would recommend 1.45 as an absolute maximum daily voltage, but 1.4 might be more safe. I am not an expert on these chips, just speaking from my own experience. 3.2Ghz is a nice little sweet spot if you ask me.

High voltages will break down the gates within the chip. Do more reading on this if you want to know more.
 
i'm runnin 3666 @ 1.55 volts in bios which shows up as 1.5 in windows. the question is wether its the voltage or the heat that hurts the chip mine idles @ 35 and loads at 36. which board are you runnin? i have a ga-650sli-ds4 now that replace my old ds3 i have absolutly no cold boot issues now.
 
Thanks for the reply, Terc. So voltage by itself does have a negative effect? 3.5 is really sweet for me; will be sad to lose it since the temps are still so benign, but if your experience pans out I think I may head back to 3.4, I can get that for just a little over 1.47 cpu volts in bios.

Yeah, if others with more experience with the C2Ds could add their views on this I'd really appreciate it. This chip is amazing compared to my old AMD x2, it makes you want to keep overclocking since it's so easy, maybe that's Intel's evil plan, get you to overclock too much until you burn out the damn thing and has to get a replacement :D
 
To my knowledge, regardless of voltage, I have yet to hear a single story of someone killing a C2D.

Excessively high temps are obviously a concern for all parts of the hardware, and of course high voltage for extended periods of time can lead to silicon and gate degradation.....

Having said that - I don't think that 1.5v will do anything at ALL to hurt the lifespan of these chips, other than knock some time off the end.

I would say that you're looking at the life expectancy of your CPU dropping from 12-15 years......down to maybe 7-9 years.

These are numbers that I'm completely pulling out of my ass, mind you.
But don't be surprised if they're close, and even if they're not, you get the basic idea.

Will your CPU explode after 6 months of running like that, no.
Will it possibly / probably diminish the life expctancy, yes.

More than voltage, the primary thing that you, me, or any user should be concerned with is heat.
If your system & CPU stays cool at 1.6v (35c or under idle / 55c or lower under load) then I wouldn't worry about it at all.

By the time that chip finally kicks the bucket, you will have purchased his great-great-great-great-great-great-great grandson chip.

The Intel Ocho-Quad Duo at 10000Mhz.

8x4 cores x's 2 Chips mated together in Intel propietary packaging with each core operating at 10Ghz.

How much voltage do you think your CPU will take then? :D
 
I have mine running comfortably @ 3ghz with vcore @ 1.425. I know for a fact I can go higher but I have a mediocre Coolermaster Hyper48 heatsink so I don't want to run the temps too high. Gonna get a high-end heatsink soon though and see what that brings me (not that i'm complaining about 3ghz I mean c'mon).
 
I found the :) happy spot for my E6400 to be 3600 MHz which is Orthos stable with a bios setting of 1.525 volts which is 1.500 volts at idle and 1.480 volts while running Orthos. I did a slight pencil mod with my P5B deluxe which cut the voltage droop a hair but nothing major.

I agree with Evoic. The amount you harm your processor by running it like this is minimal. I have an old Pentium III fully overclocked and it has been running for years like that without any problems. My niece probably wishes it would die so she could get something decent! :D

As long as your cpu is running within spec you don't have to worry. A core voltage of 1.50 volts is still within the Intel specifications for long time use.

Most people don't use their computer to run Orthos 24/7 so when you're gaming or running whatever apps you run, you'll never heat up your C2D anything like Orthos does.

Run it hard and don't worry about it. They have a 3 year warranty if you ever have a problem.

VanH: I'm using 1.55 volts for the northbridge at 3600 MHz. Try that and play around with the memory voltage before using any more core voltage when you try for Orthos stable at 3600 MHz. Keep the core temp under 60C as well for long term reliability at this speed.

The Core 2 Duo is one of the most bullet proof processors ever built. It's full of technology to prevent it from becoming the next Chernobyl. I've run my E6400 without a heatsink on it while bringing it up about 12C above the supposed maximum junction temperature of 85C without a crash. I ran Orthos at 83C for half an hour while it was bouncing off the thermal throttling rev limiter and have generally abused it as much as anyone, anywhere. Result: The thing continues to run like a bear. I found out yesterday while reading the Intel Core 2 Duo DataSheet for the mobile processors that they don't shut down until the core temperature hits 125C. ( see THERMTRIP# on page 73 ) The desktop C2D processors are designed to shut down when the core temperature gets up to approximately 105C.

Piles of head room so no worries.
 
Thanks for the feedback unclewebb. I am sure you're right, it just seem to be reaching a point of diminishing return, though. I can't really tell the difference between 3.4 and 3.5 for daily use and stepping back from 1.5 volts does make me feel more comfortable.

I think I'd be willing to run this at 3.5 24/7 if I could get rid of the vdroop and can get the lower cpu volts at idle, but too chicken to try the pencil mod (I'm sure that droop is there for a good reason).

So for daily use I've reverted to 3.4ghz with 1.475 bios volts to the CPU (about 1.45 real volts at idle), it takes 5 degrees off of the cores at load so now I'm at 48-50 load, nice comfy margin for error. Funny thing is, I managed to get 3.4 almost the first time I overclocked this, guess this is the sweet spot for me.
 
Do you think that setting the northbridge to 1.55 could make a difference in terms of 3.5GHz being stable? My few attempts at 3.5 have been with an inordinate amount of Vcore which of heated things up a bit. Even then, it wasn't fully stable.

Update: The p6n sli-fi Vnb only goes to 1.5. Tried the following: 425 x 8, ram at 800, Vcore at 1.525 (1.54 in Win), Vnb @ 1.5, Vsb @ 1.6, Vtt @ 16%. Idle temps with Tuniq fan on mid are 35/36, super pi temps 50/52 (it passed 32m), orthos temps of course go crazy up to 68C or so. Vcore drops to 1.50 in Orthos, 1.52 in super pi. So far so good...gonna play a little bf2142 and see if it crashes. This time i decided not to set the fsb/ram to 1:1, so the ram isn't oc'd. Plus, loosened up the timings to cas5 cr1.
 
Those orthos core temps are a bit warm? I think 60 should be max under orthos load. BTW have you tried overclocking with the side panel off? If you get considerably cooler temps it might well be lack of ventilation that's limiting your OC.
 
yeah for some reason orthos really cooks the cpu - and i'm using a Tuniq Tower. It's been fine so far though...played a few hours of battlefield and did some downloading. Ventilation probably is an issue...though I have a window fan on the side panel. I could probably turn the speed up on it (it's a zalman with a high power/low power connector). But like I said, i don't really see the orthos type temps for regular usage.

Ok well I'm gonna take the side panel off and see how it goes. 33C idle with the panel on at the moment.
 
oooo....this thread is beginning to make me happy about my recent E6420 purchase. I can't wait for it to arrive. I coupled it with a EVGA nForce 680i motherboard and 2gb of mushkin memory.
 
prasvt, guess it's not the ventilation then, those are pretty good idle temps (I switched out a sealed plexiglass side panel for one with a mesh and my temps dropped 10C (both idle and orthos), unbelievable, and this was with 3 120mm case fans and a couple more in the psu). The only other thing is to check that you've got that monster of a cooler fully seated (I really detest the push pin mechanism Intel used, when I first installed my scythe mine I thought I had it fully seated, but when I went back to change out the fan, turned out a couple of the pins were lose, had to work at them for half an hour just to get them fully in (confirmed by visual inspection of the back of the MB, really lousy design imo). Might want to try it with the cooler fan set to max.

Hey Dolphin, I think it is a really sweet chip as well for the price (you made a good choice imo), just skip the 350-400 fsb strap and head directly to 400+; I managed to get 3.4 almost immediately (intel totally blew me away here, speaking as a former amd fan as well).
 
VanH said:
prasvt, guess it's not the ventilation then, those are pretty good idle temps (I switched out a sealed plexiglass side panel for one with a mesh and my temps dropped 10C (both idle and orthos), unbelievable, and this was with 3 120mm case fans and a couple more in the psu). The only other thing is to check that you've got that monster of a cooler fully seated (I really detest the push pin mechanism Intel used, when I first installed my scythe mine I thought I had it fully seated, but when I went back to change out the fan, turned out a couple of the pins were lose, had to work at them for half an hour just to get them fully in (confirmed by visual inspection of the back of the MB, really lousy design imo). Might want to try it with the cooler fan set to max.

Hey Dolphin, I think it is a really sweet chip as well for the price (you made a good choice imo), just skip the 350-400 fsb strap and head directly to 400+; I managed to get 3.4 almost immediately (intel totally blew me away here, speaking as a former amd fan as well).
Click to expand...

yea, i decided to ditch AMD as the s939 dual core processor @ 3GHz got quite hot (around 70c load) and doubled as a heater during the winter months. Yea, i will try a 400+ fsb once the system is built and bootable.
 
Vanh: Yeah, I actually reseated the Tuniq Tower 2 days ago. Thankfully, this cooler doesn't use the intel push-pin system. It uses springloaded screws that connect to the backplate. I didn't see much of a temp difference (if any) with the window off, so I just put it back on. I've got 2 120mm case fans, 1 90mm window fan, the 120mm psu fan, and the 120mm hs fan, so I think I can't do a whole lot more :) But yeah, temps are down to about 29C at the moment. This is probably the longest the machine has been perfect at above 3.3ghz. Maybe all I had to do was set the Vnb to 1.5 or not OC the ram for the sake of 1:1 timings. Ah well, not arguing :)
 
Dolphin - You'll love the Intel setup. I literally built a brand new AMD system back in August of last year (right around the time the Core 2s came out, but were hard to find).
x2 4600+
2gb mushkin ddr500
dfi ultra-d
x1800xt
creative x-fi
etc.

And slowly I could feel my unhappiness get worse lol. Replaced the mobo twice, upgraded the video card twice (x1950xtx, 8800gts 320mb xxx), sound card once (x-fi extrememusic --> fatality pro gamer), and never got the cpu up above 2.7Ghz on air. Finally I bit the bullet, sold the cpu / ram , picked up the e6420 and some g.skill and life's been great. 3.4Ghz on air with idle temps in the low 30s? Frickin I love it! lol. 94fps @ 1680 x 1050 with everything except AA on max in BF2142. About time.

needless to say, i made full use of my financial aid and college loans ;-)

OH btw, I stole the northbridge heatsink/fan from my Ultra-D and put it in my MSI P6N SLI-FI - just a thought in case the mobo you got doesn't have an actively cooled northbridge. Seems to be working better with the replacement cooler.
 
Ive had mine at 3.6 ghz. However, I am currious about temps for these chips. I see alot of people sitting idle w/ 50+ degrees c core temps.... isnt that alot for being idle?

I idle arround 38-40 degrees when at 3.6 ghz. Under load it jumps up to 65 degrees (dual prime 95's) 68 degrees (when using Intel TAT tool)

each core itself is 57-60 degrees, the CPU temp is the 65-68.

Is this safe? I dont want to fry my CPU.

I am using 1.475v.

Also, outside of

FSB
Voltages
Memory speeds/timings

is there anything else you should change in the bios to abtain a high overclock?
 
You must log in or register to reply here.
Back
Top