Overclocking Escapades: Intel socket LGA775 Core 2 Duo E4700 G0 stepping

Well, it's that time again.

Time to see what my next CPU can do on the test bench.

Thus far, it's been interesting so say the least. I figured the VID rating at 1.35v was a bit high, so thought I'd go for 1.2v and see if I could get that stable. Turns out I can go far lower than that with the vCore, I'm down to 1.056v at the moment and it's still stable. Ran Prime95 for 3.5 hours and it didn't have a single error or warning. Also runs fairly cool, which is nice. I've been testing it for nearly two days now, and I still haven't found it's minimum voltage.

I've had a go with an E4700 once or twice in the past, but I never fully tested one. Just kind of set an overclocked speed (3.6GHz IIRC) and hoped for the best. I was kind of in a hurry at the time to get a submission in for the 2012 Team Cup, so didn't really have time for testing. I think I've got two or three E4700's though, so I'm not sure which one it was that I was using last time.

CPU: Core 2 Duo E4700
Motherboard: Gigabyte GA-EP45-UD3P
RAM: 2x1GB Crucial Ballistix .16FD3 Micron D9's
GPU: GTX 570
PSU: Corsair TX750

Intel Core 2 Duo E4700 G0 overclock settings
VID: 1.35v
2.6GHz (stock): 1.04v (1.0625v)
2.7GHz: 1.056v (1.0875v)
2.8GHz: 1.088v (1.1125v)
2.9GHz: 1.12v (1.15v)
3GHz: 1.152v (1.175v)
3.1GHz: 1.152-1.168v (1.1875v)
3.2GHz: 1.20v (1.225v)
3.3GHz: 1.232v (1.2625v)
3.4GHz: 1.28v (1.30v) 48°C
3.5GHz: 1.312v (1.3375v) 51°C
3.6GHz: 1.360v (1.3875v) 54°C-failed after three hours
3.7GHz: 1.392v (1.425v)

realtemp/coretemp not giving correct vid's for LGA775's anymore?

hmm, i guess after seeing better results with Q6600 G0's i would havent figured the dual cores to be worse. though its possible some or all of the dual core G0's are failed quad cores, who knows at this point.

Evilsizer said:

realtemp/coretemp not giving correct vid's for LGA775's anymore?

hmm, i guess after seeing better results with Q6600 G0's i would havent figured the dual cores to be worse. though its possible some or all of the dual core G0's are failed quad cores, who knows at this point.

Click to expand...

The Allendale/Conroe chips just seem to typically have a higher VID than Wolfdale and Kentsfield (the one's I've tested anyway).

I'm running an older version of Core Temp (0.99.8), and an older version of RealTemp (3.60). So, it's not a software error due to older hardware with newer software.

might be splitting hairs here but you did say G0 stepping which to me means it is kents based core. the conroe core being B2, then replaced by the allendale core L2,R0,M0. i think intel had a ton of failed quad core G0 dies left and decided to dump them.

as the only other G0 stepping cpus besides that E4700 are all quads from what i see on ark.intel.com, maybe i missed another dual core with G0 but i dont think i did.

There were a few different ones with the G0 stepping actually.

E4700
E6540
E6550
E6750
E6850

The E6540/6550/6750/6850 say they're Conroe-based cores. But it might be possible they were based in some way on Kentsfield, though it would be the first I've heard of it.

E2140
E2160

well if i recall right when the stepping update happened the first one to get the G0 was the Q6600 SLACR. which just made me think that the E4700 was released way later and they used the dies. never know, i just thought it if was that it should have done better ocing wise. considering how much less heat it put out, though they could have gone with the tim+glue method like used on the E4300. as that cpu ran hot no matter what i tried when ocing it vs the older E6400 i had. for the same clock and voltage, the E4300 just ran hotter. if this is the case then that explains the semi lack luster ocing, imo. the E7400's being out before the E4700 were hitting 4ghz+ in most cases on air with decent temps.

for sucide run the E7400 i got from brollocks did 4.4ghz or 4.6ghz on air for a single thread superpi run. i was dumping about 1.5v into it for that high speed on a P35 chipset, might have done better ocing wise on a P45.