Jump to content
HWBOT Community Forums

Recommended Posts

Posted

... I don't want to freeze my ass off when benching (if it's not necessary) ...

 

Been testing F4d now. Got upto 4900 in OS now. Seems a little bit better.

 

//edit: bclk is WAAAY better now. The ratios are actually working now :)

 

4764-129-1202.png

  • Replies 151
  • Created
  • Last Reply

Top Posters In This Topic

  • Crew
Posted

"For overclocking itself we reached 4800 MHz in a matter if minutes and 5 GHz after a while. That last step did need some fine-tuning though. Mind you though that we work with vanilla picked engineering sample CPUs, we can never guarantee you'll achieve the same results at home."

 

You have been Guru3d'ed !!

Posted

no it is possible, they just have to have a CPU that does higher MHz than ours. Over at OCN we gotta guy on water hitting 5.0ghz 5.1ghz on the UD7 with retail CPUs, i get jealous of him lol, anyways he got a cooler express and went on the leaderboard and posted 5.4ghz screenshot and beat all the 5.2s and 5.3s from R4Es, but he killed the board with his cooler express directed insulation directions.

 

BTW i can boot fine with quad. Is your CPU C0 or C1?

Posted (edited)

they did one better and ran their entire bench suite at 5ghz on the UD7. i mean I can believe it, but i like how they say how cherry picked their CPU is.

 

Conclusion says some about OCing too.

 

hey does C1 not show anything under stepping? His cpuz screen shows nothing, but mine shows C0?

 

BTW this is F4D quad: http://valid.canardpc.com/show_oc.php?id=2141871

 

best my memory or i guess CPU can do. But i can push BLCk to 128 easily so at least its not limited by that. I used some crap cheap crosair DDR3 2000 XMS kit, I can tighten up those timings a bit. But I think my GPU or OS got corrupted, now any BLCk change wil mess up picture, so i guess i need to swap out the GPU or reformat.

 

with 1.00 i can get 106blck: http://valid.canardpc.com/show_oc.php?id=2141832 opposed to like 102 with previous BIOS, but with slight bump in voltage

And 128 here too: http://valid.canardpc.com/show_oc.php?id=2141909 i think ima try for higher tho with this divider.

Edited by sin0822
Posted (edited)

LLC on new BIOS is superb!

 

CPUz shows slight change, but DMM slows maybe a few mv(2-3mv) change with 30%, that is like virtual no change with such a huge load. Excellent LLC. These are good improvements, still more improvements are needed, but they got the LLC right!

 

Chart added:

LL

Edited by sin0822
Posted

Nice details of sharing sin0822 and thanks for you share it ;)

 

so what good for LLC setting of your suggestion ? but if see from your chart i think around 30% to 60% will good for select.

 

btw. well new bios is fixed for 5GHz or still stuck at 4.8GHz ?

 

Thanks. :)

  • Crew
Posted (edited)
no it is possible, they just have to have a CPU that does higher MHz than ours. Over at OCN we gotta guy on water hitting 5.0ghz 5.1ghz on the UD7 with retail CPUs,

 

Not the point Sin, if the same CPU can do 5Ghz on eg R4E it should also be able to do it on any other board too...

 

X79DSI walls at 4.7Ghz here, boots at 5ghz on sabretooth... testing MSI and hopefully the UD5 tomorrow...

 

Limiting factor is not the CPU...maybe it's Massman :P

 

for laughs : I know hardwareheaven tell us they have a 5.2Ghz prime stable FX 8150 CPU... never saw a screenshot or sign of proof... words mean nothing...

 

Btw I don't get the LLC chart : 100% is no loadline if I see correctly ? 1.5 is 1.482 idle... that's the opposite of what I would logically would select in a bios... 100% LLC would mean zero droop/drop to me... darn these taiwanese with their logic keeps surprising me...

Edited by Leeghoofd
Posted (edited)

I didn't say anything to go against what you or massman were saying, i was just pointing out its possible to hit 5ghz on air on the UD7 lol.

 

 

the LLC on Gigabyte LGA2011 boards is reserved. the % actually refer to mohm rating that is the slope of the equation for vdroop(the first release BIOS F2 has mohm instead of %), so 0 slope would be no droop under load. 0% is strongest and 100% is weakest. 80% is standard 0.8mohm slope per intel spec, so smaller slope=less droop under load. its confusing i know, its just reversed 100% is most droop. 15% int he chart above is least. look at 30% and 15% and you will see the trend.

 

BTW those watts aren't measured at the wall, they are measured on the 8-pin connectors for the CPU VR, its not total system power consumption. I think conversion is 85% so .85(the values in the chart) would be more realistic.

Edited by sin0822
Posted

I updated the ud3 mobo with f5 bios.

This version is better than the previous bios version. ( f5c ).

I test that 3930k raise the clock up to 5Ghz.

I can run the super-pi 1m, impossible super-pi 32m or 3dmark11 test.

I discoverd the trick that it is same as the massman's trick.

But ud3 mobo a little bit different from msi x79-gd65.

PLL Override Function is the state of disabled is more stable.

Posted

Testing F4 final bios now.

 

It seems to be, again, quite an improvement over the beta bioses released earlier. Still the clock wall, though, limiting this chip to around 4800MHz on (crappy) air even with only two cores enabled. The BCLK overclocking capabilities have improved, giving me a perfectly functionaly 1.25x ratio! Really, no sweat pushing the BCLK anymore. Performance seems solid as well; here's a 4xCF run.

 

20866.png

 

As you can see, there seems to be no Physics low bug on this one. Before I switched to the UD7, I did some runs on the R4E and my physics at similar clocks was around 14700. Now I sometimes get 15500.

Posted

Hi

Do you have the same problem with ET6? The name of the board and the frequencies .. : (

et6klpen.jpg

 

et6_17yqzm.jpg

 

It is the X79 version of ET6. The problem I have with F4D and F4.

 

Schnuppl

Posted
That is the latest downloaded from the website? Did you fully uninstall the previous version before installing this one?

 

Yes, the last version. I had uninstalled again and installed a newly installed version. In your case, the display is correct?

Posted

Did I miss this 6A reboot loop or something? It seems everytime I hit the multi wall (54x I found) I drop in to this deep 6A boot loop that it seems only time can get it out of... Still on 4d...ud7 is going on the shelf for a bit I think.

  • 2 weeks later...
Posted

Flashed F8d this morning and I'm not too impressed with the latest BIOSes to say the least. Recovery after failed OC is a joke as it takes so insanely much time for the board to get ready to but finetuned again. The BIOS is changing from #1 to #2 and back randomly, it takes two hard reboots to get into the bios (if it doesn't switch in between again) ... sigh.

 

Gear ratios and CPU frequency seem fine, though, so once you get the board up and running it should work fine. But the BIOS definitely needs more work. They made great progress from F2 to F4; I hope they can repeat the same for F7 to F9 :).

Posted

So, figured I might as well show what kind of balls I have and give this UD7 a real xtreme OC test. 1.6V+ vcore, 5400+ cpu speed, 3DMark11 4way and Wprime. Here's what happened ...

 

94.536t.png

 

22706t.png

 

... ehr, nothing happened besides completing the benchmarks and giving me a result to submit to HWBOT :nana:.

 

I was running about -40°C (IHS on CPU); no problem with CB or CBB anymore. My chip is rather picky on changing BCLK at temps lower than -25°C, which results in a 6A boot failure. Just up temp to -25°C, set the BCLK you want, save and reboot, up the ratio and bench.

 

System:

 

- Core i7 3960X

- GIGABYTE X79-UD7 F8d bios

- 4x MSI GTX 580 Lightning Extreme

- 2x 2GB Gskill RipjawsX

- 2x 2GB Gskill Flare

- Coolermaster Silent Pro M1000

- Corsair AX1200

- Huge balls (for trying)

 

Youtube:

 

 

BIOS settings:

 

th_110120042009.jpg th_110120042005.jpg th_110120042001.jpg th_110120041954.jpg th_110120041950.jpg th_110120041944.jpg th_110120041939.jpg th_110120041932.jpg

 

Some pictures:

 

2012-01-01-20.06.24.jpg

 

2012-01-01-20.06.38.jpg

 

2012-01-01-20.06.49.jpg

 

2012-01-01-20.07.53.jpg

 

2012-01-01-20.20.13.jpg

 

2012-01-01-20.20.37.jpg

 

So, I guess that's settled then. Onto other stuff.

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


×
×
  • Create New...