What's the Highest 24/7 Stable FSB on an E4300?

[brencat]

It occurs to me that after reading tons of posts that many people whether or not they have E6600, E6400, E6420s are choosing to run 24/7 OCs typically in the 3.2 - 3.4 ghz range. Sigs with these speeds seem to be more common than those running 3.6 ghz or higher. Obviously, this means 400 - 425 FSB on a 64xx, and 375 - 400 FSB on a 6600.

My question is given this context, does it make more sense to simply buy a $115 E4300 and try for 375 FSB for a 24/7 3.375 ghz OC? I know that the Allendale has 1/2 the cache of a Conroe but if it's for a gaming PC, isn't it better to divert the $70 - $100 saved on the CPU toward a better vid card?

Thx, B

 

[AnotherGuy]

I would totally agree with u... save it on cpu n spend it on vidcard

 

[RussianSensation]

If you are on a budget, it's always better to divert more $ towards the videocard than the cpu if you are already aiming for 3.0ghz Core 2.

Overclocking is not guaranteed. But if you are trying to decide between say X1950Pro and 8800GTS vs. E6600 and E4300, definitely go for the cheaper cpu + 8800GTS and overclock that thing. I have E6400 running at 3400mhz with no problems and probably can push it further. Even at 2.8ghz, Core 2 Duo will be more than sufficient to prevent major bottlenecking for all but 8800GTX SLI setup.

 

[brencat]

Thanks guys. That's what I thought. Most certainly I plan to get a quality vid card, and I'm not on a budget per se. However, I think I should also buy a high quality mobo that will be able to handle the next-gen cpus so I don't have to swap that out in 2008 as well. So...I may end up waiting a few more weeks for Bearlake and pick up one of the new Asus or Abit boards due to roll out. Heck, the E4300 might even be a tad cheaper by then as well. But I am an idiot for not jumping on the $115 E4300 deal + free SupCom last week. I could have punted the game for $20 - $30 for a net $80 CPU cost. Oh well.

With the new Thermalright Ultra-120 Extreme, my goal is to push an E4300 to the limit...hoping for 350 - 375 FSB 24/7 stable, if that's even possible with the current batches.

 

[The-Noid]

400x9 24/7 stable at 1.55v(real). Pinmodded to 1066 fsb.

 

[brencat]

Is 400 FSB only doable with a pinmod on good air cooling? I don't necessarily need to be on the looser 1066 strap. Can current batches do 350 - 375 FSB on the 800 strap?

 

[The-Noid]

No. Not unless you board allows you to set the strap. ASUS/GB boards have problems.

965P DFI is ok. Also the pinmod is 2 minutes of work and $4 of cost. Anyone can do it.

 

[brencat]

I think with the Abit QuadGT there is a Bios line item "FSB Strap Select" or something like that. Could I simply set that to 1066 using the E4300 without pinmodding it?

FYI, although I can buy the QuadGT today if I wanted, I'm probably going to wait for the Bearlake mobos so I have more flexible upgrade path. I'm sure Abit will have an attractive offering with strap selection there too.

 

[brencat]

@Yoxxy...just to clarify, is the BSEL pinmod even necessary if my mobo allows for strap selection?

In other words, if I buy the DFI Dark or the Abit Quad GT, I won't need to do this mod correct? I can simply set the Bios to 1066 strap?

Also, does anyone know if the 650i chipset mobos (Asus P5N32-E SLI Plus, MSI P6N SLI FI) have a strap selection feature where this mod wouldn't be necessary?

 

[yacoub]

What happens if you don't/can't change the strap?

Specifically wondering about the MSi P6N Platinum since that board looks good.

 

[brencat]

Because the E4300 is an 800 mhz cpu as opposed to the conroes which are 1066, you will be chipset limited in your ability to OC the FSB (i.e. you'll hit the FSB wall much sooner) using an E4300 unless you can force it to use the looser 1066 strap. And if your mobo doesn't have a strap selection feature, you need to do the BSEL pinmod to force the 1066 strap at POST.

I posed the same question in another forum and it is confirmed that you do NOT need to do this pinmod if you have a mobo that allows you to select the strap. The DFI P965-S Dark, the Abit Quad GT, and the older Abit AW9D Max all have a strap selection feature in the Bios.

Can't really comment on any of the other popular mobos, though I believe that the strap is not selectable on the P5B Deluxe or Gigabyte DS3 for example. In addition, I've even read somewhere that the pinmod won't work if you're using the P5B Deluxe because the board takes a secondary reading and somehow knows that the E4300 is still an 800 mhz chip. Maybe others with more knowledge can comment about that board though.

 

[Conky]

Pin-mods are not advisable. Neither is more than 1.4v.

This assumes you can't afford to buy new stuff. If you have deep pockets then give it voltage until it sizzles.

 

[The-Noid]

Originally posted by: Conky
Pin-mods are not advisable. Neither is more than 1.4v.

This assumes you can't afford to buy new stuff. If you have deep pockets then give it voltage until it sizzles.

Horrible response on both fronts.

The BSEL is via conductive ink that can be taken off with alcohol. It is not even close to a pad that would have enough juice to short out the chip.

As for 1.4v, where did that come from? There is really no safe limit to how much voltage a 65nm part can take, but looking at Celerons and Pentium D that use 65nm. 1.5v is nowhere near even the limit. The 1.35v and drive strength set by Intel is to lower TDP and power draw. When overclocking power usage/output kind of goes out of the question as you are not using the stock fan.

 

[Conky]

Originally posted by: Yoxxy

Originally posted by: Conky
Pin-mods are not advisable. Neither is more than 1.4v.

This assumes you can't afford to buy new stuff. If you have deep pockets then give it voltage until it sizzles.

Horrible response on both fronts.

The BSEL is via conductive ink that can be taken off with alcohol. It is not even close to a pad that would have enough juice to short out the chip.

As for 1.4v, where did that come from? There is really no safe limit to how much voltage a 65nm part can take, but looking at Celerons and Pentium D that use 65nm. 1.5v is nowhere near even the limit. The 1.35v and drive strength set by Intel is to lower TDP and power draw. When overclocking power usage/output kind of goes out of the question as you are not using the stock fan.

Actually, Intel's own engineers state that damage starts to occur at 1.5v with the C2D. I chose 1.4v to be conservative and based on my personal experiences with air cooling.

 

[The-Noid]

Link please.

Also during your personal experiences, you could find Electro Static Migration. You are AMAZING!

I didn't know you had the money and knowledge to have a multimeter that could check drive and resistance inside an Intel C2D @ 1.4v. Please provide videos and other evidence, THIS IS A GROUNDBREAKING POST!

 

[brencat]

quote: Originally posted by: Conky
Pin-mods are not advisable. Neither is more than 1.4v.
__________________________________________________

lol - thanks Dad

Sorry to pile on here pal, but neither is overpaying for a Conroe by $70 - $110 advisable if we're only talking about a modest 200 mhz hit in total clock performance by taking an E4300 to the limit versus those who run Conroes at 3.2 - 3.5 ghz. Remember that my needs with the E4300 will be primarily gaming. Besides...it's fun to tinker!

 

[Conky]

Originally posted by: Yoxxy
Link please.

Also during your personal experiences, you could find Electro Static Migration. You are AMAZING!

I didn't know you had the money and knowledge to have a multimeter that could check drive and resistance inside an Intel C2D @ 1.4v. Please provide videos and other evidence, THIS IS A GROUNDBREAKING POST!

Wow, you're so cool. :laugh:

The link is buried in Intel's site and it's actually 1.55v upon further research(find it yourself since you know everything).

And I didn't say anything about my personal experiences with "Electro Static Migration" so stop trolling already. No seriously, stop it.

Again, 1.4v is my personal recommended limit on voltage for a C2D on air. And this has mostly to do with heat and stability issues. Intel's own engineers state that degradation starts with 1.55v and, frankly, I think they know more about this than you.

 

[The-Noid]

Originally posted by: Conky

Originally posted by: Yoxxy
Link please.

Also during your personal experiences, you could find Electro Static Migration. You are AMAZING!

I didn't know you had the money and knowledge to have a multimeter that could check drive and resistance inside an Intel C2D @ 1.4v. Please provide videos and other evidence, THIS IS A GROUNDBREAKING POST!

Wow, you're so cool. :laugh:

The link is buried in Intel's site and it's actually 1.55v upon further research(find it yourself since you know everything).

And I didn't say anything about my personal experiences with "Electro Static Migration" so stop trolling already. No seriously, stop it.

Again, 1.4v is my personal recommended limit on voltage for a C2D on air. And this has mostly to do with heat and stability issues. Intel's own engineers state that degradation starts with 1.55v and, frankly, I think they know more about this than you.

Actually I tried to send you that in PM but was ignored. I hate calling people out, but providing information that states 1.5v is Intel max, then stating in the same sentence that 1.4v is your recommendation makes it seem as if you are privy to non-public or first hand research.

If you were only providing your own recommendation for "heat and stability issues." You should preface what you say as "My recommendation though is 1.4v max as in my research heat and stability became a problem for me after this point."

Also if you would like to go with trolling, what does 1.4v have to do in a post about max stable e4300? Do you even own an E4300?

 

[Conky]

LINK

Another link (pdf file at Intel)

 

[StopSign]

I can run my E6300 at 3.2 with only around 1.25v. IMO it's not worth it to go all the way up to 1.5v just for 200-300 more MHz.

 

[aaronking]

I just put together a system with an E4300 and a Gigabyte DS3 Rev.1.3. I didn't spend much time tinkering. Simply took to FSB from 200 to 300 to 350 to 400. I haven't tested higher as 400X9=3.6 is plenty for me. I am actually running 400X8=3.2 right now until I get a better cooler. 3.2 is orthos stable at 1.3125v. 3.6 took 1.45. I have not modded anything.

 

[AnotherGuy]

nice 1 aaron

 

[bendixG15]

I pass ....

 

[cmdrdredd]

Originally posted by: StopSign
I can run my E6300 at 3.2 with only around 1.25v. IMO it's not worth it to go all the way up to 1.5v just for 200-300 more MHz.

I've had an E6400 up to 3.72Ghz on 1.45v and that wasn't much voltage at all IMO. Especially if you have really good cooling on it (was water in this case).