Core i7 Overclocking by Batch #

[iboomalot]

+15c at idle sounds fishy IMO

see if you can get another one

[Sparda]

Hi mate,

Yeah I know that those voltages are really good, this is why I don't understand why I am experiencing these huge differences in temps between this chip and my previous one.

I mean surely if I am pushing less voltage through this chip, it should actually run cooler than my previous one right?

I have tested up to 1.355v Vcore and 1.4v QPI for 4GHz, but when I ran LinX the temps hit 100c on the hottest core, so I stopped testing immediately.

And I have tried running 4GHz using the same voltages as my last chip, but it doesn't even manage a single pass of LinX before it gives me a BSOD with Stopcode: 0x00000101.

When I tested it with 1.355v Vcore, it done 2 passes before I manually shut it down because of the temps.

It must be the IHS on the chip that is warped, because I have tested 3 different coolers on it, 1 of which was the cooler that I used for my last chip, and the temps were still ridiculously high.

As you can see from those pics above my last chip @ stock idled at 34c on the hottest core, but this new one idles @ 50c, 16c of a difference, that is not normal.

Also, I have tried reseating the coolers loads of time, and reapplying AS5 to, infact, I had half a tube of AS5 when this chip arrived, I now have none left because of the amount of times I have reapplied it to see if it was due to there not being enough/to much TIM.

Anyway, as I said, I will be getting another chip pretty soon, so I am hoping I get a decent one this time.

If I didn't care about my warranty, I would just lap this chip to make the IHS nice and level, as I am pretty sure this chip has an uneven IHS, when I look at it closely, it actually looks like it is concaved, and whenever I remove my coolers, there is a line across the IHS about 1cm wide that doesn't have any TIM on it, this happens everytime I have removed my coolers, I tried spreading out the TIM so it was nice and even, and then I also put another little blob in the center before fitting the cooler, but this didn't have any effect whatsoever.

I also tried putting a washer underneath the mounting bracket so that there was more pressure on the heatsink, and this knocked a measly 2c off the temps, which wasn't worth my time tbh.

If you already mount the hs few times and still got worse temp than the old chip given that the ihs is ok then it probably a leaky cpu.
I got one last time and it certainly hotter than normal one when I poking with it.. >> Core i7: high leakage & average chips @ low VID & high VID
~30Watt make quite a difference. It certainly not kind of chip you want if you using air cooling.

The only easy method I know to see whether you chip is a leaky one is swapping 2 cpu and watch the watt meter. Or probably you can try use a clamp multimeter on cpu aux power cable..

[DavyBoy]

Thanks for the replies guys.

@iboomalot, I have talked with the retailer, and they have agreed to swap this chip for me because of these issues, so I am hoping that the new chip will run as cool as my previous one did.

@Sparda, thanks for sharing that info with me, I guess this would explain these issues I am having with this chip, although I haven't got one of those things to tell me how much power my system is drawing, nor do I have another chip to compare it with, like I said, it is being replaced anyway, so I just hope that the new one will be better.

It definitely isn't a poor mount that is causing this problem, it is either what Sparda has mentioned, or it is the IHS.

[Chicken Patty]

Just got my new D0 today. Same voltage settings as my C0 chip, trying to see how much more I can go. This voltage achieved 3.8 GHz on my C0, so far 4.2GHz on the D0

Batch # 3849A866

[KingOfsorroW]

i7 920 D0 Batch #3910A308

Prime


Linx


(nevermind the dates, I always forget to set it correctly)

[DavyBoy]

i7 920 D0 Batch #3910A308

I had a chip from this batch, and it was garbage, I had 3910A501, and it wouldn't clock any higher than 3.8GHz, and the temps were insane, I tested the chip with 4 different HSF's, and nothing I did would get the temps down.

Stock idle temps were 51c and loads were 72c!!!

Pretty much the same temps that you are getting when you are clocked to 4.2GHz!!!

Thankfully the retailer has agreed that this isn't normal and is sending me out a replacement one.

I just hope that it isn't anything like the previous one!

[KingOfsorroW]

BLunTSmoKeR Good luck with your next cpu, it seems that they are not all that bad.
Don't forget to post the results when you get it

[DavyBoy]

@KingsofSorrow, that's my replacement chip being shipped out to me now, and I also ordered my 2nd chip this morning as well, so they will both be here with me tomorrow, I will be sure to post my results on here as soon as I test them out.

[FabricioGS]

Batch #3844A716

10x Linpack stable with 1.235v

Here at almost 4000 uncore with 1.35vtt and 1.29v vcore



I'm limited with my air cooler. With a WC I would esily run it at 4.4/4.5.

[genec57]

[QUOTE=unclewebb;3938042]Scooby: Your screen shot clearly shows turbo throttling. Your multipler is at 20X at full load which equals about 4308 MHz.

At idle it jumps back up to 21X or 4523 MHz but it's not running at that speed at full load.
You need to make an adjustment in your bios to get the 21X multi to stick at full load.

Check out RealTemp 3.30 and i7 Turbo V8 so you can watch your multi at full load.

Unclewebb,
i7 Turbo has EIST grayed out and core 1 throttles. EIST is disabled in the bios of my GA EX58 UD4P. How do I get the 21X to stick?
Thanks

[DavyBoy]

Unclewebb,
i7 Turbo has EIST grayed out and core 1 throttles. EIST is disabled in the bios of my GA EX58 UD4P. How do I get the 21X to stick?
Thanks

You need to also disable C1e in the BIOS to make 21x multi stick on Gigabyte X58 mobos.

[genec57]

Both C1e and EIST are disabled.

[..::G80::..]

Just got my new D0 today. Same voltage settings as my C0 chip, trying to see how much more I can go. This voltage achieved 3.8 GHz on my C0, so far 4.2GHz on the D0

Batch # 3849A866

Very good.

[DavyBoy]

Both C1e and EIST are disabled.

Hmmm, then I don't know what could be causing this to happen then mate.

Are you using the latest BIOS for your mobo?

[genec57]

Yah
the interesting thing is that EIST shows with a check mark in i7 Turbo 6.5 and is grayed out.

[Kylzer]

Heres my batch D0

#: 3913B213

[DavyBoy]

Heres my batch D0

#: 3913B213

Not bad, but what is your highest clocks with HT enabled?

[Kylzer]

Not bad, but what is your highest clocks with HT enabled?

I can run 4.2 with HT but its a bit toasty.

4ghz is ok though.

[drewxbk]

passed 20x linx
3904A361 D0

[Chicken Patty]

Very good.

here is my latest stable settings. Currently crunching on them as well, I think this CPU is freakin' amazing!

[beast200]

Heres my batch D0

#: 3913B213

hi
i have the 3913B213 batch
@4515
cpu '1.385
vtt 1.38
seem warm at idle 41 ish but full load 77 so ok
under water

[Bei Fei]

This is my 3849B028 D0 on my new uATX build

http://valid.canardpc.com/show_oc.php?id=677707

[Kylzer]

hi
i have the 3913B213 batch
@4515
cpu '1.385
vtt 1.38
seem warm at idle 41 ish but full load 77 so ok
under water

Nice i think i'll have to put this on water some time

[oggie]

I see some people on here show their overclocks with HT off, i though whole point of buying i7's is to have more then 4 cores

[zalbard]

I see some people on here show their overclocks with HT off, i though whole point of buying i7's is to have more then 4 cores

Not necessarily. Some applications aren't very HT friendly. And x58 is a very nice platform in general.