How well does the e5200 overclock?

[faxon]

so, between me and a friend, we have 2X1gb of DDR2 1066 Crucial Ballistix ram, a 680I motherboard, a case, plenty of hard drives, a copy of XP PRO, and a 1240x1024 Dell monitor. all we need is a CPU, GPU, PSU, and a DVD drive, plus a good cooler for the CPU if needed. the budget is ~$300, so i picked this as a shopping list

http://www.newegg.com/Product/...x?Item=N82E16827106073

http://www.newegg.com/Product/...x?Item=N82E16814102792

http://www.newegg.com/Product/...x?Item=N82E16817103939

http://www.newegg.com/Product/...x?Item=N82E16819116072

http://www.newegg.com/Product/...x?Item=N82E16835233003

$310 after tax and shipping

now, what im curious to know is how stable this chip is at say 4GHz with what voltages and CPU temps (average of course), and how much people find the 2MB of L2 cache hurts gaming performance over the 6MB that is standard on mainstream to enthusiast grade Wholfsdale chips. im not counting the e7xxx chips because the 12.5x multiplier on this chip makes it so easy to overclock that there really isnt a comparison between it and the e7xxx chips in OCing at a cheap price point.

so, has anyone else purchased one to see how the performance on them is? im more curious as i would like to make this purchase tomorrow and have the comp set up and ready to go tomorrow as well if frys has all the parts at the same price newegg does.

also, i didnt yet see any official benchmarks in anandtech articles on this chip, so if there is an article on it i missed i would like a link



ed: food for thought, this is replacing a 3.2ghz Prescott, 1GB of DDR400, AGP 7800GS. the performance difference is going to be off the charts no matter what we do to it. also, yes i saw the other thread, and the guy running it at 4.3ghz at almost 1.7v. im more interested to see what you can do inside the "acceptable voltage range" of up to 1.4v

 

[Foxery]

It should be very similar in both regards; the only physical difference between an E5200 and an E8400 is the L2 Cache size - 2MB vs 6MB. There have been a lot of reviews showing that the performance difference between Core2 chips with different caches is <10%. (The 8000 series has more than it needs, as the Core2 architecture simpy isn't cache limited.)

I'll have a 5200 in hand tomorrow, but no 3rd-party cooler to really push it...

 

[edplayer]

Originally posted by: Foxery
(The 8000 series has more than it needs, as the Core2 architecture simpy isn't cache limited.)

If that were true, there would be no measurable benefit of the 6MB of cache vs 2MB


This is Anandtech [the source for hardware analysis], not Consumer Reports

 

[faxon]

hehe good to know either way. i just put in the order for everything listed above + a G11 keyboard, i should have everything in hand sometime next week since UPS Ground doesnt normally ship on weekends. will probably make a post in the video card forums after i get everything set up to let you guys know how the sapphire HD4670 does, what with it being new and all

 

[Acanthus]

3.5 on a e5200 + TRUE with a DFI P35-T2RL.

 

[error8]

Originally posted by: Acanthus
3.5 on a e5200 + TRUE with a DFI P35-T2RL.

Doesn't go any higher then that?

 

[faxon]

3.5 will still have me outperforming my buddies rig which has a Q6600 in it at stock. he keeps trying to OC it but either he got a bad OC chip or his cooler sucks, because it keeps getting to hot even at just 3GHz, and he has to increase the voltage a lot to get it to OC. expect him to start making posts here about it as soon as we get this system set up and we can concentrate on his main rig again lol. until then i get to keep yelling at him for buying nvidia boards when he doesnt use SLI

 

[nyker96]

from what I can see, the E5200s are so so overclockers. They can do 4 but needs too much juice to accomplish that. Probably the 8xxx series are still the best for dual core OCed. the E5xxx seems pretty low bin stuff. Still at that price I wouldn't care too much. However as for people upgrading to it from say E2xxx that already running at 3 or so I doubt there will be incredible gain to be had.

Faxon, I think your friend might be suffering from temp problems when OC that Q6600. Get new case/fan/HSF I would say. That chip is good for 3+. And at 3.2 or above I think it will give your 3.5 E5200 a run for the money. especially in video encoding.

 

[faxon]

yea thats what i have been telling him lol. he got a new mobo (from the XFX 680i to Evga 780i FTW) because the NB was acting unstable OCing it, and it is better now, but its running really hot like i said before. he is using a CM Gemini 2 with dual scythe fans, and i took a look at it, it isnt really doing much as far as cooling goes because of the older, poor fin design compared to current HDT coolers. since the 680i board needs a bios update to work properly with the e5200 im going to drop the S1283 on it to get an idea of the difference in temps while we flash the bios, since he is already looking into getting himself a cooler as is, and he isnt sure how much of a difference it will make or if he just got a crappy chip


ed: he is also getting himself new ram, which is why we have 2x1gb sticks avalable for the system im building with him

http://forums.anandtech.com/me...227334&highlight_key=y

if anyone wants to comment

 

[VirtualLarry]

Ok, I have an E5200 in my friend's IP35-E board now.

It's at 3.875Ghz, 1.42575v (BIOS), 1.390/1.400 (fluctuates, CPU-Z). Priming (25.6 small FFTs) as we speak. Looking good, real good. These chips are a steal! PS. 68C on both cores with a CoolerMaster HyperTX2.
Edit: One core failed prime. I don't want to give the chip any more vcore, until I can find out what's safe and what's not, so I'm going to start decreasing the frequency.
Edit: Priming 3.8Ghz.
Edit: Came back the next day, it had failed prime after an hour. Now trying 3.75Ghz (300Mhz FSB). The AC is off, now load temps are hitting 76C. Priming for an hour thus far.
Edit: Primed for 12hr, and run WoW for an hour, all while listening to internet radio. I think it's stable!

What's the max safe vcore for these babies?

 

[21stHermit]

Originally posted by: VirtualLarry
Ok, I have an E5200 in my friend's IP35-E board now.

Edit: Priming 3.8Ghz.

Well, that would seem to answer the Abit IP35-E E5200 BIOS question.

 

[error8]

Originally posted by: VirtualLarry
Ok, I have an E5200 in my friend's IP35-E board now.

It's at 3.875Ghz, 1.42575v (BIOS), 1.390/1.400 (fluctuates, CPU-Z). Priming (25.6 small FFTs) as we speak. Looking good, real good. These chips are a steal! PS. 68C on both cores with a CoolerMaster HyperTX2.
Edit: One core failed prime. I don't want to give the chip any more vcore, until I can find out what's safe and what's not, so I'm going to start decreasing the frequency.
Edit: Priming 3.8Ghz.

They seem to overclock exactly like E7XXX series. Great budget chips from Intel, as usual.

 

[Acanthus]

Originally posted by: VirtualLarry
Ok, I have an E5200 in my friend's IP35-E board now.

It's at 3.875Ghz, 1.42575v (BIOS), 1.390/1.400 (fluctuates, CPU-Z). Priming (25.6 small FFTs) as we speak. Looking good, real good. These chips are a steal! PS. 68C on both cores with a CoolerMaster HyperTX2.
Edit: One core failed prime. I don't want to give the chip any more vcore, until I can find out what's safe and what's not, so I'm going to start decreasing the frequency.
Edit: Priming 3.8Ghz.
Edit: Came back the next day, it had failed prime after an hour. Now trying 3.75Ghz (300Mhz FSB). The AC is off, now load temps are hitting 76C.

What's the max safe vcore for these babies?

1.42v was my line.

I could get into windows up to 3.8 but had to back it off to 3.5 for prime stability.

 

[Shmee]

hmm, hey, I am the friend faxons been talking about. anyways, I heard anything over 1.4v is bad for yorkies or wolfsdales, so idk. I think its odd that the e5200 wouldnt oc well, is it different stepping or something?

 

[Foxery]

1 week in: My E5200 is Folding@Home stable at 3250 MHz, 1.24v (BIOS, less vdrop). Temperatures aren't even breaking 60C with the shitty stock cooler.

3330 MHz @ 1.26v (BIOS, less vdrop) not stable. It's not a high-end motherboard, and I haven't tried higher voltages yet.

 

[VirtualLarry]

The E5200 in my friend's IP35-E is 12hr Prime95 25.4 small FFTs stable, and one hour WoW stable, and about 6 hours internet radio stable.
3.75Ghz (300Mhz FSB), 1.42575v (BIOS), 1.392/1.400v (CPU-Z). Load temps 76C (no AC)/71C (with AC).

IP35-E, 4x1GB DDR2-667 HP RAM, 6600GT, ThermalTake 500W PSU, 250GB Seagate PATA HD, IDE DVD burner, CoolerMaster HyperTX2 heatsink, Xion XON-403 case.

Another friend of mine has an E5200 in a Gigabyte G31 micro-ATX mobo, he got his up to 3.76Ghz (301Mhz FSB), unsure about voltage (he had it on auto at one point).

 

[Tempered81]

that puppy needs lots of volts! I don't think i've ever seen one stable at > 4ghz with < 1.44v load. still a nice bargain larry! since you dont care about running your conroes to 90C, try upping the vcore on your 5200 to 1.7 @ 4.2g

 

[nerp]

Hmm. 3.7ghz E5200 sounds really killer and I have the IP35-E, could be an easy plug and play type situation for me, but for just 500mhz and my use, I'm STRUGGLING to justify jumping on it.

Part of me thinks just to splurge, get an E8400, E8500 or E8600 and be done with it.

 

[faxon]

Originally posted by: Shmee
hmm, hey, I am the friend faxons been talking about. anyways, I heard anything over 1.4v is bad for yorkies or wolfsdales, so idk. I think its odd that the e5200 wouldnt oc well, is it different stepping or something?

to answer this for everyone, from what i was reading about intel,s binning process, these chips are the ones that came out binned badly in that they need more voltage to do the same clocks as their higher binned counterparts. also as far as i know, they are the same E0 stepping as the e8600, the Q9650, and any other penryn achitecture chip fabbed in the last month and a half, give or take

 

[Shmee]

weird. so looks like were getting the leftover crap. still, if it CAN run stable and not burn at high volts, then it is still a good overclocking budget chip. so long as it can take the extra v needed.

 

[edplayer]

Originally posted by: faxon
to answer this for everyone, from what i was reading about intel,s binning process, these chips are the ones that came out binned badly in that they need more voltage to do the same clocks as their higher binned counterparts. also as far as i know, they are the same E0 stepping as the e8600, the Q9650, and any other penryn achitecture chip fabbed in the last month and a half, give or take

I thought it used the M0 stepping like the E7200

 

[VirtualLarry]

Originally posted by: jaredpace
that puppy needs lots of volts! I don't think i've ever seen one stable at > 4ghz with < 1.44v load. still a nice bargain larry! since you dont care about running your conroes to 90C, try upping the vcore on your 5200 to 1.7 @ 4.2g

Err, if it were in my rig, I would, but this is a friend's rig, so I didn't want to risk anything so I stuck to 1.4v actual.

Edit: yeah, these E5200 are M0 stepping, not E0.

 

[Foxery]

12.5x 275 = 3437 MHz @ 1.30v (BIOS)
CPUz reports 1.31v under load (??)

So far so good, 3 hours in. Temps are still only 60C! (using RealTemp 2.7) The lack of transistors with the small cache make this thing very low-power and low-heat. I love it.

 

[nyker96]

I'd personally wait for the E0 refresh of this line. These probably little older stock chips from the M0 era.The E0 might allow slightly better OC with less volts. At their present state I wouldn't think I want them.

 

[VirtualLarry]

Originally posted by: Foxery
12.5x 275 = 3437 MHz @ 1.30v (BIOS)
CPUz reports 1.31v under load (??)

So far so good, 3 hours in. Temps are still only 60C! (using RealTemp 2.7) The lack of transistors with the small cache make this thing very low-power and low-heat. I love it.

Sounds good. Maybe you can make it to 3.8 @ 1.4v?