The official GIGABYTE "Pi Is Returned" Contest thread.

[I.nfraR.ed]

1.5Ω for 6800K and 0.7Ω for 5800K

 

I didn't use the slow mode. I was booting at 1.55V for the cpu and 1.45V for the NB, then gradually increase cpu voltage in ET.

It was the second session, benched several hours on the first one and about 2 hours for the second.

It died at 1.55V when I tried to increase bclk from 117 to 118 in ET (no voltage change in settings, just the bclk). Then tried to power on the board, but it seemed to have a short, because psu protection was kicking in.

 

Stopped benching and dried everything carefully checking the board, no visual burn on the mosfets. Then tried it with both cpus and guess that's what killed the 5800, stupid me

 

The old board turned off immediately with 8-pin cpu power connector attached, the new one keeps running.

 

Hopefully the new board is still alive, otherwise I will have a hard time with RMA'ing everything if it gets accepted at all.

Do a working board show something on the debug display when powered on without a cpu?

[The Stilt]

A working F2A85X-UP4 does not display anything in the segment display if there is no APU installed.

Take a bare board (with nothing attached) an measure the resistance between one of the OS-CON capacitors located between the inductors (CPU loop, six inductors located on the left from the socket).

 

The resistance should be around 1.6k (climbs for a while).

The resistance for the NB loop (two inductors located above the socket) should be between 2.5 - 2.6k.

[I.nfraR.ed]

Both seem about right.

Thanks for the help and good luck in the competition!

You have the tools and the knowledge

[Wizerty]

I can't join the competition :

 

http://hwbot.org/submission/2426428_wizerty_superpi___1m_a10_6800k_10sec_531ms?recalculate=true

 

http://hwbot.org/submission/2426429_wizerty_superpi___32m_a10_6800k_10min_3sec_47ms?recalculate=true

Edited September 16, 2013 by Wizerty

[Remnant]

Bought a 8350 and 990fxa UD5 for this competition. UD5 had cb at -120 and cbb-80. Returned both for new 8350 & 990fxa-ud3 rev 2.0. It died from Condensation due to improper insulation around chokes and mosfet. Bought another 990fxa-ud3 rev 3.0 and it died during bios flash. Exchanged for 990fxa-ud3 Rev 4.0 and it wont work with 8350 or 1090t before or after bios flash. Works only with 965BE and its cb was -50 and cbb -30... Not sure why so low, or if the Hardware Gods hate me. Lot of time invested in this competition, nothing to show for it

[dinos22]

The Stilt cleaning up

[BeepBeep2]

Well, had efficiency issues then dropped a channel then wouldn't post...must have been doing it right.

 

Still got 8 gig validation, fun stuff.

[Xtreme Addict]

Congrats The Stilt

[Massman]

The UD7 is a little tricky like that. I have tested 3 different boards, and they all behave the same. At least with the latest non-UEFI BIOS.

 

Air: 275 max stable, sometimes doesn't boot, sometimes loses channel

LN2: 335 max stable, after that freeze or no post

 

To get to 300HTT+, you need to dial in the memory subtimings manually. Once you figured out a right problem, save it. Also, you will never be able to boot at the top HTT frequencies. My procedure usually is

 

BOOT 240 -> 275 -> 295 -> 300 (-> 315 (if I want to get close to 330 HTT)

 

This has worked on all three boards.

 

Fyi, the UD3 can do 300HTT on air, and 350HTT on LN2. Less layer seems better for signaling ...

[tiborrr]

I could run 990FXA-UD7 at 410 something on a Sempron 145. HTT wall also depends on the CPU, I have seen variances of about 40MHz from worst and best.

[Massman]

Might be related to architecture or HTT spec. Sempron is still K10.

 

I've tried multiple boards, multiple BIOSes, multiple CPUs, multiple memory kits, and they all have the exact same walls

[Wizerty]

Can you please fix it... FM2 submition fail every time

 

 

http://hwbot.org/submission/2426592_

 

 

http://hwbot.org/submission/2426602_wizerty_superpi___32m_a10_6800k_9min_54sec_812ms?recalculate=true

 

 

thanks

[Massman]

Fixed.

 

You don't submit straight on the competition page? I don't understand why this is failing for you, but working for other people.

[Wizerty]

yes from "normal" sublit request... but it works for AM3+... anyway, thanks

[ivanov]

Fine competition Too bad I'm fourth in both AM3 categories ;P

[Wizerty]

Thanks A LOT to The Stilt, you really deserve to win. Congratz to every body, nice competition, AMD FTW.

[Calathea]

Congrats to all winners, well deserved and overall sick results in this competition. Just thinking of how fragile 6800k seems to be, how many was killed this comp? 1 Here, 1 for infrared, 2 for dfordog.. maybe more? Never killed a cpu before, this is kind of shocking.

[BeepBeep2]

The UD7 is a little tricky like that. I have tested 3 different boards, and they all behave the same. At least with the latest non-UEFI BIOS.

 

Air: 275 max stable, sometimes doesn't boot, sometimes loses channel

LN2: 335 max stable, after that freeze or no post

 

To get to 300HTT+, you need to dial in the memory subtimings manually. Once you figured out a right problem, save it. Also, you will never be able to boot at the top HTT frequencies. My procedure usually is

 

BOOT 240 -> 275 -> 295 -> 300 (-> 315 (if I want to get close to 330 HTT)

 

This has worked on all three boards.

 

Fyi, the UD3 can do 300HTT on air, and 350HTT on LN2. Less layer seems better for signaling ...

I think this should just be called "Gigabyte syndrome", this issue goes back to 790FX.

I'm running Rev 1.1 UD5, it would not POST over 272 HT Ref on LN2, around 270+ it kept looping "Updating backup BIOS to latest version..............."

 

Easytune locked up the machine when I tried to change HT Ref in windows, luckily PSCheck worked to achieve higher frequencies, but I threw any possibility of running SuperPi out the window.

Edited September 16, 2013 by BeepBeep2

[kirbster]

blew away the old records! Nice work!

I will have to settle for fastest water cooled 5800K...

No money! ;-)

[The Stilt]

ust thinking of how fragile 6800k seems to be, how many was killed this comp? 1 Here, 1 for infrared, 2 for dfordog.. maybe more? Never killed a cpu before, this is kind of shocking.

 

The APUs are really not that fragile.

It is usually the voltage spike (e.g. a PState change with a large voltage offset) or a motherboard VRM failing or shorting due excess load or condensation.

 

It is naturally possible to kill the APUs with voltage.

They are more fragile than the FXs for example.

 

About half of the die is reserved for the GPU so the power density on APUs is massive when they are pushed to the limits. I would guess it is rather the substrate-die interconnects which might be failing rather than the die itself.

 

The CPU can usually take up to 1.95V on LN2 (depending on leakage) when only one of the CUs is being loaded. The GPU can take up to 1.7V, after that it dies.

[Hero]

6800K stronger than 5800k,so I love 6800k so much,but nb voltage more than 1.65v,CPU pwm mosfet will be burned!CPU will also be implicated.

[I.nfraR.ed]

I never went above 1.5V for NB, but it still died and afterwards the board killed 5800K from the last country cup, because I was stupid enough to try it.

[Calathea]

Same here, it died with nb 1.55 volt, vcore 1.96 during POST. Probably shouldn't try to boot that high vcore Anyway really fun competition.

[ATMINSIDE]

The man himself came in and put out a hurting!

 

Some great scores out there guys!