Overclocking Escapades: Intel socket LGA775 Core 2 Duo E8400 C0 stepping

I'm a 775 noob and I will use a crappy old Rosewill air cooler. Come on man it will be fun!

Well, some developments have come to light in my testing.

Stock (3GHz 333x9): 1.1v (1.13125v)
3.1GHz: 1.1v (1.13125v)
3.2GHz: 1.12v (1.15v)
3.3GHz: 1.17v (1.18750v)
3.4GHz: 1.18v (1.21250v)
3.5GHz: 1.22-1.23v (1.25v)
3.6GHz: 1.25v (1.2750v)

Was going to continue my testing for maximum overclock yesterday, but it was 85°F/29.5°C outside and 80°F/27°C inside. That doesn't make for good conditions for testing an overclock at max load with Prime95, as I was hitting 60°C pretty quickly even with the CPU cooler fans (Delta EFB1212SHE) at maximum speed.

Going for 3.7GHz now and it is requiring a huge voltage boost. 3.6GHz was 1.25v (1.2750v), but 3.7GHz is taking 1.36v (1.38?v) in order to not fail the test within 1-5 minutes.

You can make open contest in benching section. Core 2 low clock or something. I bet other users will join too. Can add post in CPU section so others know about it.

This C0 stepping E8400 is really weird.

I've been completely unable to get 3.7GHz stable thus far. So, to test if it was a FSB wall or an inability of the CPU to run at 3.7GHz I dropped the multiplier from 9x to 8x and increased the FSB from 412MHz to 420MHz.

It's a bit odd, there's apparently a voltage "black hole" it will load to the OS at 1.4v but will blue screen after a minute or two of being at the desktop. So, I tried the 1.425v and 1.45v settings, then it wouldn't make it into the OS at all (blue screened as soon as it tried to load).

I had to increase the voltage to the 1.475v setting (effectively 1.44-1.46v in OS HWMonitor), then it would get to the desktop in the OS. I don't know that this setting is stable yet, and in fact it probably isn't, as I haven't tested it at all. EDIT: wasn't stable, crashed as soon as I started a run of SuperPi 1m.

This CPU apparently needs shed loads of voltage just to try to run at 412-420MHz (effectively 3.7-3.8GHz with maximum multiplier/stock multiplier). It also puts off tons of heat with any voltage setting past ~1.35V and up to 1.45V (max I've gone so far). At the stock frequency and voltage it idles at 37-39°C (44-46°C under load), at 1.45v it idles at 42-44°C (58-60°C under load). EDIT: I was completely wrong about this, it does not need tons of voltage for 3.7GHz, my EP45-UD3P just really sucks.

For a while it seemed like I had 3.7GHz stable, it would run Prime95 for 45mins to 1hour 15mins, and then error out. So, I tried increasing the CPU voltage, then it crashed even faster. Then tried increasing FSB voltage and MCH voltage, neither helped.

That's one weird chip.

Not like the i5-2400 you did

Silver_Pharaoh said:

That's one weird chip.

Not like the i5-2400 you did

Click to expand...

Yeah, I think I liked the i5-2400 better. It was at least relatively predictable.

This one is just plain odd. Usually (with 775 chips at least) there should be a point where you need a small voltage boost to the vCore to gain an extra 100-150MHz over what you can achieve at stock voltage, and then you might get another 200-300MHz with a larger voltage boost beyond that level.

This one doesn't have that, it has a point where you need a giant voltage boost (1.45v it seems) to get 100MHz more past what is achievable at stock voltage, and then it won't go any farther. I hit a solid wall at 3.7GHz and it absolutely won't run 3.8GHz at all (I went up to around 1.46-1.47v and it still wasn't stable). Yet strangely it will run at 3.6GHz with the stock 1.25v.

I just seem to have one of the worst E8400 C0 stepping CPU's in existence as far as potential to overclock, since in all of the professional reviews I read the review sites were able to easily get 4.0-4.5GHz with ~1.45-1.5v.

I don't think it would be worth the extra heat and voltage/wattage to run beyond 3.6GHz for 24/7 operation anyway (with this particular CPU at least, as it seems to run hotter than other Wolfdale CPU's I've used (even at stock voltage and frequency)).

I'm glad I have an E0 stepping as well, because this C0 chip is really, really disappointing. I already know my E0 can run 4.0GHz at least.

Tech Tweaker said:

This one doesn't have that, it has a point where you need a giant voltage boost (1.45v it seems) to get 100MHz more past what is achievable at stock voltage, and then it won't go any farther. I hit a solid wall at 3.7GHz and it absolutely won't run 3.8GHz at all (I went up to around 1.46-1.47v and it still wasn't stable). Yet strangely it will run at 3.6GHz with the stock 1.25v.

I just seem to have one of the worst E8400 C0 stepping CPU's in existence

I'm glad I have an E0 stepping as well, because this C0 chip is really, really disappointing. I already know my E0 can run 4.0GHz at least.

Click to expand...

I never had that problem with my E8400 C0, which I got in late 2012 and OC'ed it to 3.825 on August 25, 2013 with my Maximus II Gene. (450x8.5)
I didn't need more than stock Vcore at 3.825 with my C0. (Linpack stable, too)

Even some E0's require quite a bit more vCore to reach thier potentil see here: http://valid.canardpc.com/pzl6tc

Well, I managed to get it running faster, I just needed a board that was a better overclocker.

Now at 3.7GHz, in the OS I'm seeing 1.296-1.32V, and I set 1.35V in the BIOS. And I'm running with a stock cooler right now. Temps aren't too bad actually Core 0 is 41-44°C at idle, Core 1 is 43-45°C at idle. Core 0 is hitting 53°C max under load, and Core 1 is hitting 57°C max under load.

Seems completely stable thus far. Ran Cinebench r11.5, SuperPi 1M, PiFast, wPrime 32M, Aquamark3, 3DMark03, and 3DMark06, and completed all of these benchmarks successfully with no errors or issues.

Which board are you using in it now?

if i didnt sell mine id join in sold her last summer she ran 4.2 24/7 on a cheap air cooler

Silver_Pharaoh said:

Which board are you using in it now?

Click to expand...

EVGA nForce 780i SLI (132-CK-NF78)

Well, I have to say that 3.8GHz is pretty much the limit of stability for this particular E8400.

Above that, going for 3.9GHz or 4GHz temps just go crazy and a ridiculous amount of voltage is required (especially for 4GHz).

Even then, I couldn't get 3.9GHz or 4GHz fully stable. I could run Prime95 and sometimes it would run for an hour and then have an error, sometimes it would be after two hours. Seemed to be temperature related, as increasing voltage wasn't really helping with stability, and was just increasing the load temperatures even further. Seemed like any time load temps were going above 60°C it would cause an error, and I can't really get it much cooler in the room since it's summer here (outside temps are 80-90°F/27-32°C lately) and room temp is 72-78°F/22-26°C typically (depending upon how high I have the AC turned up and whether or not the door is closed (closed door=cooler temps)).

Thats the c steppings for you I have had 4 or 5 I cant remember e0s all did 4ghz with pretty much stock volts most did 4.8 24/7 on water

Tech Tweaker said:

For a while it seemed like I had 3.7GHz stable, it would run Prime95 for 45mins to 1hour 15mins, and then error out. So, I tried increasing the CPU voltage, then it crashed even faster. Then tried increasing FSB voltage and MCH voltage, neither helped.

Click to expand...

That sounds like the strange crap my E2180 was doing at 3.0 and 2.928.
Prime95 would fail in 5 hours, then it suddenly started failing a lot faster. Even when I turned the core clocks down to 2.928!

The board's likely faulty! I never expected an E8400 to lose stability like that at only 3.7!

I was testing this CPU again over the last couple of days.

A small update, 4.1GHz is apparently possible for benchmarking runs, but I can't get it fully stable for a 24/7 type of overclock. The voltage is really up there for 4.1GHz, I'm at 1.456-1.464V in HWMonitor and Prime95 is still throwing out errors at around 35-38 minutes into testing.

Core 0 is maxing out at 65°C under load, Core 1 maxing out at 66°C under load. The CPU is currently being cooled by my Thermalright SB-E Extreme and 2x TY-140 fans. So, the cooling used really shouldn't be an issue. It's too hot to remain stable though, just too close to its limits.

It seems as though the second E8400 C0 I got isn't much better than the first.

It has a lower VID, but tops out at the same core speed. The only upside is that it's able to achieve that speed at a lower voltage than the other CPU.

thats no good.
i have an E0 stepping e8400 that did 4.5ghz prime stable with ~1.4-1.5v (its been a few years, dont remember exact numbers) with custom wc loop.
would bench at 4.9ghz or so, but could never get it prime stable at that speed.