Wolfdale e6300 overclocking, VERY disappointed

[jhyukkang]

hi everyone, yesterday, i finally attempted to overclock my 4 months old Pentium E6300, excited to see over 4.0Ghz!!!
i read almost all the reviews on Newegg, and other websites, and i concluded that this chip will do atleast 3.5 on stock voltage. that is why i bought this chip. upgrade from e4400. i was increasing FSB, prime for 5 min, and increase FSB little more. at the point when i was 3.41Ghz and prime got me an error after 2 minutes. long story short, i figured out it was vcore that needed to bump. (i was trying to find out what other could affect stability than vcore because what i have researched online, which this chip should not have vcore shortage until atleast 3.5). trying to not believe the reality, i kept going...
here are my STABLE results
clock voltage
2.8 1.2875 (stock)
3.36 1.2875 (stock)
3.41 1.3
3.46 1.325
3.51 1.3375
3.57 1.35
3.62 1.3625
3.67 1.3875
3.72 1.4125
3.78 1.4375
i didnt go anyfurther because of too high vcore for 45nm.

with 1.4375 volt, i could not get 4.0Ghz, not even 3.8Ghz. i dont know how much performance exactly ill get at 4.0 over 3.78, but i like to see this at 4.0. maybe i like to see the "numbers" go high.
someone got 4.0 with 1.36Vcore. 3.6 at stock.

btw my specs
e6300 Wolfdale
1GB x2 , 512MB x 2, DDR2 6400
Gigabyte P35 DS3L
GTS 250 512MB EVGA
3 HDD
2 ODD

i raised FSB, GMCH, DDR Voltage also. makes no difference.

is there something BIG i am missing?
or is it just a chip from a bad batch??

 

[spentshells]

way to much voltage this s going to sound dumb doo you have the most recent version of the BIOS ?? your voltage is out of control for 45nm

3.67 1.3875 is about as high as I would go not much change except benchmarking

what cooler are you using ??

 

[jhyukkang]

i believe my DS3L supports 45nm, u mean latest even though it supports 45nm?

with 1.4375V 3.78Ghz it never goes over 60 during prime with Xigmatek 92mm 4 Heatpipes, i believe this is because of high vcore. many people are using stock cooler and got to 4.0 without going over 70C. so with xigmatek, i expect more..

edit - model for cooler HDT-SD964 sigmatek

tried many ways to lower the voltage but no luck. it would only get stable if i raise vcore... looks like if i shoot for 4.2Ghz, it will suck up like 1.6 Vcore...

 

[jhyukkang]

also i have been using this since march at stock setting with no problems

 

[jhyukkang]

-_- anyone got an idea??

 

[spentshells]

yeah update your bios

 

[jhyukkang]

i didnt know bios can affect cpu overclockablity, and stability at certain voltages,
i didnt do it cuz i did it long time ago, and i crashed it. ill try one more time;;

 

[jhyukkang]

ok i did get the latest bios which is F9 for DS3L. but no luck here. i didnt try different multiplier before bios update. but right now i found out lower multi seems to add little more stability at little lower voltage. normally is 10.5. im able to get this stable at 3.78 (8.5 multi 444FSB) with 3 step lower in vcore.

probably this have been same with older bios. i just didnt try it before.

so did i just get a bad chip?

 

[jhyukkang]

one more thing, i ran prime for 8 hours before bios update, i found out that cpu needed 1.4325V to get 3.70Ghz stable for 8 hours. (not 3.78)

i just started prime for my new settings 3.7Ghz( 435 x 8.5 ) at 1.40V. ill see how this goes

even with lower multi, this is really bad result. 3.7 with 1.4V on 45nm...

 

[jsc]

First, have you seen this:
Shadow's Gigabyte motherboard OC guide:
http://www.tomshardware.co.uk/forum/page-245679_11_0.html

Second, what are your memory settings?

Set your memory voltage to factory recommended values. Change the System Memory Multiplier (or whatever your BIOS calls it) from AUTO to 2.00, 2.00B, or 2.00D - whichever you need to set the Memory Frequency to twice the FSB. Then when you increase the FSB, the memory clock will rise in in proportion with it. At an FSB of 266 MHz, your memory clock should be at 533 MHz.

Keep your core voltage under 1.45 and your core temps under 70 C.

And third: Intel sold you a 2.8 GHz chip and your complaining that you can't reach 4.0 GHz with it?

 

[jhyukkang]

thanks for reply

first i have read that gigabyte mobo oc guide page you gave me. also this is not my first time overclocking. i did oc my e4400 2.0Ghz to 3.33Ghz with no problem. plus more experience with my friends pc overclocking also.

second my ram is already at 2.0 (or 1:1) i know all about memory frequency and fsb link, multiplier, voltage, and all that stuff already;

third haha i know intel sold me a 2.8 chip, but intel didnt sold me a one that has capability of overclocking which all other people seem to have recieved with same (or should be) 2.8 chip. (i might sound as if i dont like intel, but i actually love intel way better than amd hehe) i always buy intel to all pcs i have built.

after testing more at 3.7 i found out it wasnt stable at that clock with 1.4, so
FINAL FINAL clock is 3.65Ghz with 1.4125V. highest temp is 62C

my ram is at factory rec settings, with FSB 384 x2 so 768Mhz. which is below factory clock 800Mhz

 

[jsc]

It's not the magical 4.0 GHz, but that ain't bad for a budget chip.

 

[jhyukkang]

one of my friends computer which i built and overclocked, used same e6300 wolfdale and p31 gigabyte mobo. sadly p31 only supports fsb up to 335 Mhz. so his result was 3.517Ghz with 1.3375V stable with prime 24hrs. mine would take at least 1.3875 to get to 3.517Ghz stable.

 

[Drowning Doom]

Wow... I get much better on my 45nm E6300... Here is what I'm running it at:

333 MHz bus speed at a 10.5 multiplier (pushing 3.5 GHz)
Voltage is at 1.275 (stock voltage)

Ran stable with a 12 hour Prime95 test.

Thinking about upping the OC again now.

I'll be trying to go to 366 MHz x 10.5 next. With the voltage at 1.352

This chip should never go above 74.1 C, you should also never take the voltage above 1.4 (though Intel says 1.3625)

Max temp I hit over the 12 hour test was 55 C, though I'm not using stock cooling, I'm using the Zalman CNPS9500 AT

It must be something in your mobo... I'm running an Asus Rampage Formula x48, wish I could be of more help, but all I can provide is what I got it running stable at... :S

 

[spentshells]

Wow... I get much better on my 45nm E6300... Here is what I'm running it at:

333 MHz bus speed at a 10.5 multiplier (pushing 3.5 GHz)
Voltage is at 1.275 (stock voltage)

Ran stable with a 12 hour Prime95 test.

Thinking about upping the OC again now.

I'll be trying to go to 366 MHz x 10.5 next. With the voltage at 1.352

This chip should never go above 74.1 C, you should also never take the voltage above 1.4 (though Intel says 1.3625)

Max temp I hit over the 12 hour test was 55 C, though I'm not using stock cooling, I'm using the Zalman CNPS9500 AT

It must be something in your mobo... I'm running an Asus Rampage Formula x48, wish I could be of more help, but all I can provide is what I got it running stable at... :S

um your motherboard is .....extremely better than his but in the end it's likely just the chip

 

[jsc]

Intel's recommended max voltage is 1.45 volts, but many times you will reach thermal limits first.

 

[Drowning Doom]

Intel's recommended max voltage is 1.45 volts, but many times you will reach thermal limits first.

Where did you see that? Not that I blatantly don't believe you or anything, but if thats true then I can push my CPU a lot more... I am seeing 1.3625 on Intel's website for the chip.

 

[spentshells]

Where did you see that? Not that I blatantly don't believe you or anything, but if thats true then I can push my CPU a lot more... I am seeing 1.3625 on Intel's website for the chip.

you can exceed that 1.45 if you like. providing you can cool your chip and keep it below 70C i prefer lower than 62C myself

 

[Jak_Sparra]

anyone know if memory errors cause errors to show in prime95? i know when i have overclocked in the past using water. cpu temps were very low, but the memory was stupid hot and causing problems.

 

[crooked vulture]

are your memory sticks rated at the same latencies? Sometimes mixing sticks can cause problems. Try removing the pair of 512mb sticks... And try loosening the timings slightly. i know you mentioned you know about all of this...

 

[jsc]

Where did you see that? Not that I blatantly don't believe you or anything, but if thats true then I can push my CPU a lot more... I am seeing 1.3625 on Intel's website for the chip.

That 1.3625 volts is the max VID or the maximum "stock" voltage that the chip programs the motherboard VRM to provide.

Here's where the "1.45 volts" figure comes from.
http://download.intel.com/design/processor/datashts/320467.pdf page 17.

Hmmm. I had better save that link. Someone else is certain to ask.

You can learn all kinds of neat things poking about the technical documentation.

anyone know if memory errors cause errors to show in prime95?

Yes. P95 small fft's tests primarily the CPU. I am guessing that except for reporting, the test runs only in the CPU cache. Both the large fft's and blend tests reads and writes extensively to the RAM. So memory errors will be more likely to show up there.

 

[jhyukkang]

Wow... I get much better on my 45nm E6300... Here is what I'm running it at:

333 MHz bus speed at a 10.5 multiplier (pushing 3.5 GHz)
Voltage is at 1.275 (stock voltage)

Ran stable with a 12 hour Prime95 test.

Thinking about upping the OC again now.

I'll be trying to go to 366 MHz x 10.5 next. With the voltage at 1.352

This chip should never go above 74.1 C, you should also never take the voltage above 1.4 (though Intel says 1.3625)

Max temp I hit over the 12 hour test was 55 C, though I'm not using stock cooling, I'm using the Zalman CNPS9500 AT

It must be something in your mobo... I'm running an Asus Rampage Formula x48, wish I could be of more help, but all I can provide is what I got it running stable at... :S

ive seen lots of people who got those frequencies with stock voltages, im jealous of all ur overclocks..
i thought p35 was enough to overclock e6300 to at least 4.0. this board was one of the top 775 overclocker.

also ive seen quite of bit of people who oced 45nm pretty high with p35 chipset too. this is where im starting to think its my cpu

 

[jhyukkang]

are your memory sticks rated at the same latencies? Sometimes mixing sticks can cause problems. Try removing the pair of 512mb sticks... And try loosening the timings slightly. i know you mentioned you know about all of this...

two 1gb sticks are from pny rated at 800mhz and two 512mb sticks are from corsair rated at 667 mhz. thats why i even added +.3V because 512mb are being overclocked alot compared to 1gb ones. also i losen the timings too just in case.

i havnt tried with just 2 same sticks because i need 3GB -_-

 

[jhyukkang]

one thing that doesnt make sense to assume its memory pulling back cpu overclock is because the cpu actually does get stable when rasing vcore. i can boot up at 3.65 with 1.3875 but prime would fail within 3 min. but when i finally tried 1.4185 it was stable for 12 hrs

 

[jaspreetsinghbabbra]

the reason may be a bad board or a lazy chip.......my board takes 1.44 for northbridge oc to 2400 mhz in amd x4 b50 or may its the weak chip....