6420 Disappointment

[Diogenes2]

FWIW ? I picked up a 6420 to put together my 2nd C2D rig .

Could not get this 6420 stable at 2.8, even with volts up to 1.5 ..

MOBO is GA-965P-DS3 3.3
RAM is OCZ2P8002GK which Is stable to 1000mhz, however I was
leaving below 800 to find stable CPU OC ?

I may pop it in my sig rig later, just to make
sure it is not a non CPU issue , but it doesn?t seem
likely at this point..

I've noticed there doesn't seem to be a lot of people praising
the OC ability of this latest batch of chips..

 

[ghitz]

It's all about the CPU temperature.

 

[FireChicken]

My friend just got his up to 3.44ghz and hit a hard wall. 430 FSB was fine and othos stable for 8hrs, but wouldnt post at 431 and we tried everything. Wierd thing is that he could get it up to 3.2 at default voltage. He wants to hit 3.5-3.6 think that is possibe with the 6420?

these cpus are suppoed to be the same at the 6400 but with the extra cache. But the 6400 are monster OCers. Seems like all the 4mb cache cpus dont overclock as well as the ones with the 2mb cache. I wonder how the extra cache affects C2D OC ability?

 

[HopJokey]

Originally posted by: FireChicken
My friend just got his up to 3.44ghz and hit a hard wall. 430 FSB was fine and othos stable for 8hrs, but wouldnt post at 431 and we tried everything. Wierd thing is that he could get it up to 3.2 at default voltage. He wants to hit 3.5-3.6 think that is possibe with the 6420?

these cpus are suppoed to be the same at the 6400 but with the extra cache. But the 6400 are monster OCers. Seems like all the 4mb cache cpus dont overclock as well as the ones with the 2mb cache. I wonder how the extra cache affects C2D OC ability?

I don't believe the cache affets the OC ability as the 6400 physically has 4MB of cache (2MB is disabled). It probably has to do with the manufacturing; the 6420's come from different wafers than the 6400's.

 

[aka1nas]

The cache disabled 6400s will probably run a lower Die temp as the die has the same surface area as a 4MB cache chip but the disabled cache is "dead" and not producing heat.

This is sometimes the case when cache-disabled chips sometimes overclock better/ run cooler than the natively lower cache models that replace them. The ClawCastle/Newcastle and the cache-disabled Toledo/Manchester were both examples of this.

 

[The-Noid]

Originally posted by: HopJokey

Originally posted by: FireChicken
My friend just got his up to 3.44ghz and hit a hard wall. 430 FSB was fine and othos stable for 8hrs, but wouldnt post at 431 and we tried everything. Wierd thing is that he could get it up to 3.2 at default voltage. He wants to hit 3.5-3.6 think that is possibe with the 6420?

these cpus are suppoed to be the same at the 6400 but with the extra cache. But the 6400 are monster OCers. Seems like all the 4mb cache cpus dont overclock as well as the ones with the 2mb cache. I wonder how the extra cache affects C2D OC ability?

I don't believe the cache affets the OC ability as the 6400 physically has 4MB of cache (2MB is disabled). It probably has to do with the manufacturing; the 6420's come from different wafers than the 6400's.

e6400 = 2mb of bad cache. e6420 = 4mb without error. They come from the same wafers.

Are you thinking of E4300?

 

[PuffDaMagicSamus]

Sounds like it's something other than your chip affecting your OC. First off I'd def. get a better cooler. Considering I have my chip at 3.2 on 1.32v and got up to 3.6 on 1.5 you shouldn't be having a problem getting to 3.0Ghz. Post your setting in your BIOS. List what options you have enabled/disabled and what all your voltages are. Also what are your RAM timings?

 

[aigomorla]

Originally posted by: Diogenes2
FWIW ? I picked up a 6420 to put together my 2nd C2D rig .

Could not get this 6420 stable at 2.8, even with volts up to 1.5 ..

MOBO is GA-965P-DS3 3.3
RAM is OCZ2P8002GK which Is stable to 1000mhz, however I was
leaving below 800 to find stable CPU OC ?

I may pop it in my sig rig later, just to make
sure it is not a non CPU issue , but it doesn?t seem
likely at this point..

I've noticed there doesn't seem to be a lot of people praising
the OC ability of this latest batch of chips..

See, thats where you went wrong.

1 chip never = another chip. This means, because on chip A i got X overclock, Chip B will give me the same.

THATS TOTAL BS. Chip A and B can hvae the same week/bin. But i promise you the max chip A to B will be off by serval mhz.

NO CHIPS ARE EVER THE SAME.


Also, what is your PSU? What is your Cooling? Whats your BATCH? Your missing a lot of variables which is leading me to believe your an overclocking noobie. <i dont mean this is a downgrading sesne, just mean you must be new if you take offense to this statement i appologize>

Cooling/RAM/CPU/BOARD/PSU
ALL of these are required to overclock. IF one of them is WACKED, it will throw your overclocking off WACK as well.

Lastly, NB needs voltage when you overclock. That will inturn make it HOT. Read the complaints about people with gigabyte DS3 boards and NB cooling. Im guessing thats where 1 problem for you lies.


Also u failed to mention any temps, or programs used to overclock.

Go back to google, and read a overclocking guide.

OVERCLOCKING IS NOT PLUG AND PLAY.

I dont understand why people think it is. Learn to do it properly, or dont do it at all. A very small % of computer owners overclock. If you dont know how to, DONT ATTEMPT IT.

BAD OVERCLOCK =
Loss of major amounts of time
Loss of major amounts of data
Possible damage of hardware
Possible hours of headache because of overcooked hardware leading to errors.


READ up on it.

 

[f4phantom2500]

Originally posted by: aigomorla
1 chip never = another chip. This means, because on chip A i got X overclock, Chip B will give me the same.

THATS TOTAL BS. Chip A and B can hvae the same week/bin. But i promise you the max chip A to B will be off by serval mhz.

NO CHIPS ARE EVER THE SAME.

Well, not necessarily. They produce so many CPU's that there must be 2 that overclock to the same level. Think about it, if you plotted out the overclocking results of every single E4300 ever produced, you would find the mean and standard deviation of the overclocking results, and you would see that there are so many clustered near the mean that, due to the number of CPUs they made, there must at least 2 that have the exact same limit. Of course, this is impractical as you would have to overclock every single one on the exact same system (just keep swapping out the CPU's), but generally speaking, they mostly overclock pretty similarly, with some outliers (some that overclock relatively well, some that overclock relatively poorly). The closer they are to the average overclock, the more likely it is that there's another one out there with the same max OC as most of them are near the mean (which is why it's the mean).

 

[The-Noid]

If you can't hit 3 ghz you are doing something wrong.

Raise NB Voltage, Get better cooling on the NB, update bios.

I would blame the MB a lot more than I would blame the cpu.

 

[Diogenes2]

Originally posted by: PuffDaMagicSamus
Sounds like it's something other than your chip affecting your OC. First off I'd def. get a better cooler. Considering I have my chip at 3.2 on 1.32v and got up to 3.6 on 1.5 you shouldn't be having a problem getting to 3.0Ghz. Post your setting in your BIOS. List what options you have enabled/disabled and what all your voltages are. Also what are your RAM timings?

Thanks for the concern, but I'm not asking for help. I just posted this as an observation, in light of the assumptions by some, that every C2D is a 3g chip by default...

I've been overclocking since the celeron 300 , and I know a little bit about it.

I spent about a week ( when I could find some time ) tweaking this combo
in every possible way.
It would take a lot more time than I have to spare, to list all the settings I've tried.

Like I said, I may try it in my other DS3 that is stable at 3.4, just to make sure this
MOBO doesn't have a problem . It is just not a good time for me to take it ( the other rig ) down.

I' ll let you know if I find it's not the chip, after all ..



If you are sure it's not the chip, make me an offer ..

 

[n7]

Honestly, even though it does sound like you got a bit of a poorer OCer, i somewhat agree with Yoxxy.

DS3 is NOT a good consistent OCing mobo, contrary to what many people seem to think.

I realize people don't always want the best, but if you get what's pretty much an entry level OCing mobo, i don't think you can expect the same results as from a higher end option...

 

[cmdrdredd]

Originally posted by: n7
Honestly, even though it does sound like you got a bit of a poorer OCer, i somewhat agree with Yoxxy.

DS3 is NOT a good consistent OCing mobo, contrary to what many people seem to think.

I realize people don't always want the best, but if you get what's pretty much an entry level OCing mobo, i don't think you can expect the same results as from a higher end option...

This, I totally agree with. I've built 5 systems around a DS3 and every one of them had some strange issues when overclocking. I've built a few systems (probably around 3) using the board I have. It costs alot more so people don't want it. However, each one has been easier overall. Then I have built a few 680i and 650i based systems. Each of the 650i systems was far and beyond easier to overclock and get stable than any other I have worked with. Even the 680i had issues with certain memory and timings. I don't know what it is about the 650i that makes it less troublesome.

I don't even do this for a living, I have friends who I built systems for and they tell their gamer friends that I can build their PC cheap (if they provide parts) and I offer my advice on what parts they need based on what they're doing. From there I build a system for them. Before you say something about me just going ahead and overclocking their system that's not what I do. I always ask if they want to overclock. I explain what it does, how much performance can be gained, and I always explain the risks and any potential loss of data/damage. I do a thorough 24hour test of each one and a couple hours of intense gaming before I say it's stable. I just thought I should point that out because when I tell people I build overclocked systems they think I'm building a PC that will break in a week.

 

[The-Noid]

I will trade you for an E6850 that idles at 60C.

 

[simonnance]

6850 @ 60C idle..... lap that beast!

as for the DS3.. maybe a FSB hole? Try bumping it up by a reasonable chunk, rather than by a few mhz. Some mobos have buggered FSB strapping at certain frequencies, yours may be one of them. 2.8Ghz is 350FSB..... maybe push it to 375+ on 1.4/1.45 and see if it totally craps out? Should otherwise be ok on that voltages at that speed, as long as it isnt a REALLY bad chip.

 

[GuitarDaddy]

I definately think Intel is binning more agressively since the first of the year and since price cuts have been announced. It makes sense as prices are getting cheaper and they are selling more high end chips to system builders now fully onboard with C2D.

I'm not seeing the monster overclocking reports for low end parts like early on. And the 6400 I got in february is pretty unimpressive. It runs decent temps on a mild vcore bump up to 3050mhz, then it requires increasingly more vcore as you go up. The absolute top I can get the chip orthos stable is 3.3ghz and that takes an insane amount of vcore and temps load in the mid 60's with an Ultra 120 extreme.

 

[amenx]

New F11 Bios released for DS3:

1- Fix PCI-E overclock issue
2- Support Intel E6420 & E6320 CPU

http://www.gigabyte.com.tw/Support/Motherboard/BIOS_Model.aspx?ProductID=2424

 

[mithrandir2001]

I don't think Intel intentionally disables 2MB of cache to make a E6300/E6400. These chips were instead tested and found to have 2MB of bad cache but were otherwise good, i.e. these are chips that likely could have been binned E6600 or better if it wasn't for the 2MB bad cache. That's why the OC is so good.

Now the Intel has fewer problems with bad cache and all C2D CPUs will have 4Mb of cache, Intel can bin them by their tested speed.

 

[gobucks]

wow, i really figured that the new e6420s would be way better overclockers than my e6400... i guess i don't regret my purchase after all. I've really never heard of a conroe that can't hit 3GHz.

BTW, i don't know if it's relevant, but at least initially DS3s had issues with OCZ ram. My rev 1.0 DS3 had all sorts of problems with the OCZ memory, and it limited my overclocking quite a bit, so i returned it. I think the issues are supposed to be fixed by now, but it still sounds like there are some problems for some users, so if you know somebody with another brand of RAM, I'd at least pop it in and see if your overclocking results are different.

 

[aigomorla]

Originally posted by: Yoxxy
I will trade you for an E6850 that idles at 60C.

holy crap... is that with water and after you laped the cpu?


@ Op: And i really dont think your CHIP is the issue.

Can you do me a favor?
I need the batch #

it looks likes

L631B120 <- that was my E6600 ~ 3.6ghz
L640F048 <- my xeon X3220
Your chip should have simular numbers stamped on top. I can then compare it on serveral data bases, and see what people are getting with simular week/bins.

If you truely have a sub 3ghz C2D. Feel bad for ya. However if someone has the same stepping as you do, chances are your board is erroring.

 

[cmdrdredd]

Originally posted by: gobucks
wow, i really figured that the new e6420s would be way better overclockers than my e6400... i guess i don't regret my purchase after all. I've really never heard of a conroe that can't hit 3GHz.

BTW, i don't know if it's relevant, but at least initially DS3s had issues with OCZ ram. My rev 1.0 DS3 had all sorts of problems with the OCZ memory, and it limited my overclocking quite a bit, so i returned it. I think the issues are supposed to be fixed by now, but it still sounds like there are some problems for some users, so if you know somebody with another brand of RAM, I'd at least pop it in and see if your overclocking results are different.

the memory issue is based on the IC and voltage used by default. Usually Micron ICs need more voltage (2.0v+).

 

[Diogenes2]

Hey everyone, thanks for the input.

I forgot about the problem with DS3 and OCZ, but I figured if it passed MEMTEST at 1000mhz ( with one or two sticks ) it wouldn't be my problem.
I have some Patriot I'll give a try.

Also, good reminder about the FSB holes.

The F11 BIOS is worth a try also.

There is a lot of experience here that is always worth tapping into ..

Thanks again everyone. Maybe my pessimism with this chip is premature ..

 

[amenx]

Originally posted by: mithrandir2001
I don't think Intel intentionally disables 2MB of cache to make a E6300/E6400. These chips were instead tested and found to have 2MB of bad cache but were otherwise good, i.e. these are chips that likely could have been binned E6600 or better if it wasn't for the 2MB bad cache. That's why the OC is so good.

I think they do/did actually disable the 2mb cache for the e6300/e6400 just to fill in the lower marker segment and so it wouldnt compete too closely with their e6600. From what I've gathered their initial production runs were so good that many of these lower end CPU's could have easily been passed off as e6600's, but were intentially crippled just to fill in the lower end markets.

 

[SamzAthlon]

Are you using prime 95? and for how long?

 

[SamzAthlon]

My new system is about a week old now and I still running on stock speed. Reading this thread has made me nervous about discovering a dud if I overclock.

I used prime 95 max heat torture test on stock settings with the following temps..

CPU 59C
CPU Core 1 54C approx
CPU Core 2 53C approx

I mean if I oc the cpu the cpu temp will definitely go up to 70c's, which isnt good right????