E4300 Overclocks?

[heatpipe]

Update: 1.43vcore (in BIOS), +.19v NB, 2.2v Mem

401mhz cc at 7 multiplier - won't post

350mhz cc at 9 multiplier - (3.15ghz) Orthos stable for 3 hours now. Looks good.

That 350mhz at 9 (with voltages goosed somewhat) might be a reasonable starting point for the E4300. Though I'm sorry some of you guys can't even get that.

Thanks for your numbers Bznotins, they helped. Now we'll see if she can do a little more, or if I can lower voltages somewhat. Then on to adjusting the memory.

 

[tylerw13]

Originally posted by: heatpipe
Update: 1.43vcore (in BIOS), +.19v NB, 2.2v Mem

401mhz cc at 7 multiplier - won't post

350mhz cc at 9 multiplier - (3.15ghz) Orthos stable for 3 hours now. Looks good.

That 350mhz at 9 (with voltages goosed somewhat) might be a reasonable starting point for the E4300. Though I'm sorry some of you guys can't even get that.

Thanks for your numbers Bznotins, they helped. Now we'll see if she can do a little more, or if I can lower voltages somewhat. Then on to adjusting the memory.

i must have gotten a really good chip....im able to overclock really well if it werent for my crappy mobo....damn thing

 

[Zap]

Originally posted by: Duvie
I have never found an FSB wall on the S3....

How about a FSB hole? Since the S3 is one of the more budget boards (last I checked, $108 on Newegg) I've got my eye on it for my future E4300/E4400 build if I want a board to mess with on my test bench (rather than my usual mATX board that makes it into my systems).

 

[bznotins]

BTW I tried 8x401 @ 1.45v and it wouldn't even POST. All I got was a blank screen with the fans running. I had to use the CMOS jumper (what a pain that is with a double-width vidcard) to get it to POST again.

I don't think I tweaked the NB voltage and that might have been the problem there. But FWIW nonetheless.

 

[dokomo]

I'm curious, for those that can do it, are there any advantages to running the fsb at 400 and the multiplier at 8? Instead of the fsb at 355 and the multiplier at 9?

 

[mroverclocker]

yes faster fsb=faster overall scores. I generally dont notice a diff going from 3.2-3.5

 

[magreen]

Originally posted by: Avalon

Originally posted by: InterHmai
I am using a GA-965P DS3 with a Tuniq 120, and can get it to 362x9 @ 1.43 vcore (might be 1 vcore setting higher). I can't seem to get it stable past 370fsb. I can get it to post up to 390fsb by bumping the vCore to 1.5 but Orthos fails left and right.

Is that 1.43v actual or BIOS setting? Do you have any way to check your vdroop? You may have more room for voltage than you think if your board droops a good amount. Still, 3.25Ghz doesn't sound too bad. If you want, try lowering the multiplier by 1 and raising the FSB up. A few of the initial reports said these things like high FSB, so that may make a difference. It's worth a try.

I second that call for lowering the multi and raising FSB. The P965 on the DS3/S3 is known to have an FSB hole between ~370-400 FSB until it changes to the 1333 strap.

 

[InterHmai]

I tried lowering the multiplier to 8, but seemed to hit a FSB limit of 390. Anything higher wouldn't POST at all. I think I was able to get it once to 412x8, but that required a rather high vCore (1.5ish?), so I didn't want to bother.
Even at 390fsb, it was terribly unstable. Orthos would cause a program fault when I tried to run a test.

That being said, with a 9x Multiplier I seem to have capped out at 372fsb @ 1.46875v. Load temps were right around 58c. Going any higher would push the CPU temp past 60c and cause Orthos failures. CoreTemp/Speedfan/TAT never reported the same temps for my CPU, each being higher than the other respectively. TAT seemed to be the most accurate.

So, anyone care to guess the lifespan of my CPU @ 1.468v if it was on 24/7?

 

[TripperJoe]

Originally posted by: Rio Rebel
You guys who are buying the S3, here's a tip in case you haven't seen this in other threads:

If you are going to increase the FSB significantly, there's a bios setting which lets you cap the frequency of the pci-e slot. Cap it at 100mhz. It can mean a huge difference in how high you can go.

I was only going about 350 stable on the fsb until I changed that setting. Now I can go over 400 easily. I was stopping at 2.5ghz until last night, where I was able to run windows at 3ghz. I'm not sure where my ceiling is now, but it is much higher.

Found this over in Hot Deals. Thought it might help.
http://forums.anandtech.com/messageview...&STARTPAGE=5&FTVAR_FORUMVIEWTMP=Linear

 

[fire400]

805 _ 4.1 on water with OCZ
6300 _ 3.7 on water with XMS
4300 _ 3.9 on water with XMS-D

X6800 _ 3.8 on water with OCZ

 

[magreen]

Originally posted by: InterHmai
I tried lowering the multiplier to 8, but seemed to hit a FSB limit of 390. Anything higher wouldn't POST at all. I think I was able to get it once to 412x8, but that required a rather high vCore (1.5ish?), so I didn't want to bother.
Even at 390fsb, it was terribly unstable. Orthos would cause a program fault when I tried to run a test.

That being said, with a 9x Multiplier I seem to have capped out at 372fsb @ 1.46875v. Load temps were right around 58c. Going any higher would push the CPU temp past 60c and cause Orthos failures. CoreTemp/Speedfan/TAT never reported the same temps for my CPU, each being higher than the other respectively. TAT seemed to be the most accurate.

So, anyone care to guess the lifespan of my CPU @ 1.468v if it was on 24/7?

Well, 3.35 GHz on an e4300 is nothing to sneeze at.

On the other hand, something seems fishy here... If your cpu can do 3.35 GHZ then it should be able to do it at 9 x 372 or at 8 x 418. And since your board is a DS3, it should have no problem with the 418 FSB... the DS3 can go up to 500+ FSB.

There's a chance you could go higher if you solve this mystery... maybe increase the FSB voltage by +0.20V to +0.30V. I've heard of that curing stability problems on the DS3 board above 400 FSB. Then you might be able to lower your cpu voltage back down and you wouldn't run into the heat problems you mentioned.

 

[InterHmai]

Originally posted by: magreen
So, anyone care to guess the lifespan of my CPU @ 1.468v if it was on 24/7?

Well, 3.35 GHz on an e4300 is nothing to sneeze at.

On the other hand, something seems fishy here... If your cpu can do 3.35 GHZ then it should be able to do it at 9 x 372 or at 8 x 418. And since your board is a DS3, it should have no problem with the 418 FSB... the DS3 can go up to 500+ FSB.

There's a chance you could go higher if you solve this mystery... maybe increase the FSB voltage by +0.20V to +0.30V. I've heard of that curing stability problems on the DS3 board above 400 FSB. Then you might be able to lower your cpu voltage back down and you wouldn't run into the heat problems you mentioned.
[/quote]

Yeah, I was baffled by this as well. I tried 400-425fsb but can't seem to get it to post at all, even with the extra FSB, MCH overvoltage, or extra vCore. Even if it does post, it'll probably just be hella unstable and crash in windows like it did with 390fsb. Not quite sure whats up with that.

 

[HeroOfPellinor]

Well, I had about 5 minutes of free time last night, so I bumped mine up to 2.8GHz and a 1:1 divider and set it to prime large fft on each core for an hour and came back with no errors. I'd bumped the core up to 1.40 just to make sure it would be stable.

System runs noticeably faster than stock speeds. This, in fact, will probably be where I run this permanently after going for the max OC just for fun.

 

[Bry]

I just got my system set up. I have the Gigabyte S3 (not DS3), E4300 (stock cooler), and Corsair PC6400/800mhz memory. I was stable at 3ghz (333mhz), but failed Orthos at 3.3ghz (366mhz). I can probably bump up the vcores to get that stable.

What I can't figure out is why I can't POST at 400mhz fsb. I set my CPU multiplier to 6x, so the CPU should be running at 2.4ghz. I tried bumping the fsb and mch vcores to +0.3v, PCI express +0.1v, DDR +0.4v, and nothing. I even added a 40mm fan to the northbridge.

What am I missing? Anyone else have this problem?

 

[spinejam]

what bios are u using Bry? F6 has worked well for me.

 

[Conky]

Originally posted by: Bry
I just got my system set up. I have the Gigabyte S3 (not DS3), E4300 (stock cooler), and Corsair PC6400/800mhz memory. I was stable at 3ghz (333mhz), but failed Orthos at 3.3ghz (366mhz). I can probably bump up the vcores to get that stable.

What I can't figure out is why I can't POST at 400mhz fsb. I set my CPU multiplier to 6x, so the CPU should be running at 2.4ghz. I tried bumping the fsb and mch vcores to +0.3v, PCI express +0.1v, DDR +0.4v, and nothing. I even added a 40mm fan to the northbridge.

What am I missing? Anyone else have this problem?

Just a thought but maybe you need to lock the PCI-E at 100. I did this with my Gigabyte board and then everything started behaving properly.

 

[Conky]

Originally posted by: tallman45
So much for the E6300 and E6400 once everyone sees what an E4300 can do

If I didn't already have an E6400 and was in the market for a C2D I probably would've gone with the E4300. The local Fry's has them for $149... I'm half-tempted to get one just to play with, lol.

But I am pretty happy with my E6400. Apparently I got a good one that thinks it's a 3.2GHz chip and does this stable at default voltage. The E4300's sweetspot seems to be around 3.0GHz which is still faster than the top of the line X6800.

 

[Bry]

I was originally on the F8 bios, then flashed to F7 without seeing any difference. Have you seen that things improve on the F6 bios? Maybe I'll give that a try tonight...

Yes, I am locking the PCI-E to 100.

 

[HeroOfPellinor]

Well, I'm not having success going much higher than 2.8. But I'm using the new ABit IB9 which needs probably a BIOS revision or two before we start getting decent OCs on it.

Anything above 2.8 and the system is hanging before it loads into Windows. It will freeze while running through the IRQ list or while detecting IDE devices or whenever. It also froze in the BIOS once and once before evening going into the BIOOS while the splash screen was still up. Tried 1.45 vcore and NB at +.25 and memory at 2.1. PCI-E is locked at 100Mhz.

 

[Bry]

Hrmm.. What kind of temps are you guys running? My CPU core is idleing at around 60*C, and getting into the upper 80's under load, and this is at stock vcore (stock cooler with AS5). This is when measured using hmonitor or Intel TAT. Are these accurate? I've heard people say things like it shouldn't be past 65 under load, and I'm almost there just at idle. I'm looking at my CPU fan and it isn't even spinning, presumably because it thinks its cool enough already (it stutters).

Edit: HMonitor says the cpu core is in the 80s, but the mainboard is 44 and cpu1 is 33. On other forums, some people have said that the intel cooler is "ear deafening", but on my setup it doesn't sound particularly loud. Under load, I'm able to hold my finger to the heatsink and not burn myself. I would think 80C would be enough to hurt?

Anyways, at stock voltage it's orthos stable at 2.7ghz so I may just have to live with that.

 

[Soulkeeper]

i just got a 4300 setup 2 days ago
it is running happily at 2.65GHz on stock voltage
i havn't tryed to go higher yet as i don't wanna lower my ram speeds and am happy with this speed for now
also my board doesn't have vcore adjustment for some reason ... asus p5l-mx

i like this cpu so much i'm gonna buy another one next week to go with the other 775 board and mem i picked up already

 

[HeroOfPellinor]

Originally posted by: Bry
Hrmm.. What kind of temps are you guys running? My CPU core is idleing at around 60*C, and getting into the upper 80's under load, and this is at stock vcore (stock cooler with AS5). This is when measured using hmonitor or Intel TAT. Are these accurate? I've heard people say things like it shouldn't be past 65 under load, and I'm almost there just at idle. I'm looking at my CPU fan and it isn't even spinning, presumably because it thinks its cool enough already (it stutters).

Edit: HMonitor says the cpu core is in the 80s, but the mainboard is 44 and cpu1 is 33. On other forums, some people have said that the intel cooler is "ear deafening", but on my setup it doesn't sound particularly loud. Under load, I'm able to hold my finger to the heatsink and not burn myself. I would think 80C would be enough to hurt?

Anyways, at stock voltage it's orthos stable at 2.7ghz so I may just have to live with that.

At 2.8 my idle temps are 35° according to AbitEQ. That sounds way too high.

 

[Conky]

Originally posted by: Bry
Hrmm.. What kind of temps are you guys running? My CPU core is idleing at around 60*C, and getting into the upper 80's under load, and this is at stock vcore (stock cooler with AS5). This is when measured using hmonitor or Intel TAT. Are these accurate? I've heard people say things like it shouldn't be past 65 under load, and I'm almost there just at idle. I'm looking at my CPU fan and it isn't even spinning, presumably because it thinks its cool enough already (it stutters).

Edit: HMonitor says the cpu core is in the 80s, but the mainboard is 44 and cpu1 is 33. On other forums, some people have said that the intel cooler is "ear deafening", but on my setup it doesn't sound particularly loud. Under load, I'm able to hold my finger to the heatsink and not burn myself. I would think 80C would be enough to hurt?

Anyways, at stock voltage it's orthos stable at 2.7ghz so I may just have to live with that.

Dude, that is way too hot. TAT is very accurate as is Coretemp.

If your cpu fan is not spinning then I would advise to hook the fan connector directly to your power supply and hope that you have not damaged the chip.

And touching the heatsink is not necessarily a good way of testing it's efficiency. If it is not even making contact it will be cool.

For what it's worth, the stock Intel fan is not loud and anyone that says it is must be some sort of freak who wants a completely silent system. But back to the issue at hand... get that fan working!

 

[Soulkeeper]

lol
i just updated my bios to the latest and pushed mine to 3.1Ghz on default voltage
tryed 350fsb but it wouldn't post right even with lower mem speeds
this looks like the limit for my chip without more vcore and/or better cooling
possibly the chipset/motherboard's limit also

awsome chip

 

[Bry]

OK, disabled the "smart fan" functionality in the bios and the fan is going fine... I'm now running at 1.25v cpu @ 2.7ghz and orthos has passed for the last 30 mins or so. I think my problem is the poor contact between the CPU and heatsink, and I'm debating on whether or not to plunk down another $30 for a coolermaster hyper tx (seems to be the best hsf without having to reinstall the mobo).

Anyways, here's the useful data for y'all:
1.25v @ 2.7ghz = orthos stable, may be able to go lower, but don't see the point
1.35v @ 3.0ghz = orthos fails after 1min
1.3625v @ 3.0ghz = orthos fails after 3min
1.4v @ 3.0ghz = orthos stable
1.425v @ 3.15ghz = memtest stable, orthos fail, but at this point my cpu cooler is crapping out

I think an aftermarket hsf will get me to at least 3ghz, not sure if it's worth it tho (2.7ghz is pretty damn good already, probably can't tell the diff anyways).