I don't know. But I have tried some different tREFI from 7936 up to 25000 the bandwidth is still the same.
Isn't 2400 peak at 19200 but I got more than that?



tREFI: 10360


tREFI: 12000


tREFI: 14000


tREFI: 25000


Until I adjusted tRRSR from 6 to 4

tRRSR: 4




tRRSR: 5


The higher the tRRSR the lower the bandwidths.

tRRSR: 7


Maybe you could give some advices on this.

Btw, welcome back bro.

Thanks for the PMs.
Well I can't make it worked but nevermind.

1200 x 7.8 = 9360



As for SuperPi32m 5Ghz Ivy All Out Challenge!, the tREFI might not be a meaningful number as it doesn't really affect much on performance. But I didn't test stability performance on different tREFI values perhaps later once I got some time.

Insane chip they got there. With that voltage I'm pretty sure that they still have backup score waiting for Asus' answer.

Yupe, it's all depends on your luck. In my case, I have two of the good chips at the moment:

1) Malaysia: lower VID but IMC 2777 max
2) Costa Rica: higher VID but IMC 2933 max

I think I will not buy Ivy anytime soon since Haswell is around the corner.

It is fine to use TIM as long as they don't put the black adhesive glue too thick or the air gap between the IHS to the die is not too large that would be okay already.

This post has been edited by owikh84: Apr 23 2013, 12:48 PM

Holy %$&%*&*@^ what an insane chip!
But don't know whether it is a correct vCore reading until he tries it on another mobo and verified by digital multimeter.
If true then it should be a 7.5GHz chip with 117MHz BCLK.
If false then Gigabyte...

False sensor?
No idea on the CPU stepping but from your pic it looks like Intel is going to release C3 stepping.
So far we only have details on the Z87 chipset stepping which will be launched with C1 stepping. C2 stepping which fixed the USB3.0 problem will be in mass production in mid-August.

Looks like that bro, those early adopters will be forced to exchange their board C1 to C2 stepping starting from Aug two months after launching.
But this time at least we know this issue earlier before release date.

Sorry I quoted wrongly i was referring to stasio's post earlier.
Alright, I don't know what's wrong with your PC, if not OCed do you have this boot issue.
What did you mean by PC can't start normally? Can post or not?

I think there's no better way than RMA it.

Like raydenex said, you may try another PSU, RAM, GC, or even CPU if you have spare
Now, two of my 7970s are not working properly, not detected correctly as HD7900 series in device manager, causing sudden shutdown after booting into windows.
I suspect it's either my GC, mobo PCIe slots, or PSU problem.

Good job! We got another success story here.
My chip also has some scratches on PCB, but fortunately it's still alive & rocks.
Pls share some OC results or cooling improvement of the delidded CPU. CLU is the best TIM after CL Pro that I have ever owned. CLU outperforms Arctic MX-4 by 9-10C: Proof
Good luck!

Nice try. But prime95 25.x is kinda old and doesn't have AVX instruction to stress the overclocks accurately.
Without this AVX libraries it's like stressing half of the performance of the Ivy. The latest one is posted in Post#1.

The buzzing noise came from coil whine, which is normal for mos/some PSUs.
I had 4x Corsair AX1200, 2x HX1000, TX750 all suffered from this buzzing sound.
XFX Black 750W and Aerocool Strike X 1100W didn't have this issue.
Can I say this is due to Corsair or OEM poor manufacturing, poor PCB contact, poor soldering & poor glue contact? HAHA

AX850/AX750/AX65 OEMed by Seasonic: Source

AX1200 is Flextronics: Source

Coincidentally, gold price is also dropping worldwide lol.

How about try Platinum PSU?

Could be bent ATX pins etc.
I have solved my 7970s issues, one of the PCIe cables that I sleeved myself has loosen pins LOL. Maybe I force it too hard due to short cable routed from top case of LD V8
Now 3-way CF is working perfectly again.

Still no 7.5GHz?

Actually you can, just that it might take longer time to sand if you start from a higher grit.

When you are switching to different grit you are actually just sanding out the scratches left from the previous grit. If you start with lower grit then you tend to get really deeper scratches.

Yes, that's in correct order. Always start from lower grit to higher.
Don't sand it too long then you should be fine. You might wanna google for CPU lapping tips/guides for more info.
Good luck!

Well done. Nothing wrong with the settings and temps except for the minimum ones.
Are they really your idle temps or you launched RealTemp during priming?
Your CPU's vCore is like any other average chip.

Higher CAS and other timings if set correctly generally will improve stability, but at lower performance.

A quick alternative to cut down the vCore is to grab the ASRock OC Formula and expect to get at least 0.06v lower than any Asus mobo.

On your current board, you might try to increase the CPU PLL volt from 1.80v to 1.85v/1.90v. You might get lower vCore without hurting the performance.