Llano OC & benchmark collection thread

[Alex-Ro]

Hmm,ill bump it up.Well now CPU-NB voltage sets both IMC voltage and GPU voltage?

[PcCI2iminal]

Hmm,ill bump it up.Well now CPU-NB voltage sets both IMC voltage and GPU voltage?

i dont know well how it works on llano,but i still trying many voltages and for a high bclk NB voltage seems to help the stability
and now im pushing the memory clock speed and I just changed these two voltages(NB and vddr) and everything else was default or auto

[imamage]

I can't get mine A8-3850 beyond 120 x 29 , pushing VCore / CPU-NB voltage hard doesn't work
It boots up but Windows always show BSoD

I guess the Windows is corrupted

[mattkosem]

Mine seems to be doing fairly ok so far. Haven't pushed the cpu freq yet. Just focusing on the IGP so far.

Here's what I'm getting in vantage so far:
stock (29x100) ram @ 1866 9-9-9-24 - http://3dmark.com/3dmv/3301950
3059 (23x133) ram @ ~1733 9-9-9-24 - http://3dmark.com/3dmv/3304624
3045 (21x145) ram @ 1932 9-11-11-36 (way safe timings just for playing) - http://3dmark.com/3dmv/3306154

Hard to complain so far. These are all at roughly stock voltages. I didn't change anything on the first two. The third one is at 1.450 CPU and 1.250 NB.

--Matt

[mattkosem]

My ram seems right at home on this board. The nic on it really blows though. I've got two of the same board sitting here and the nic drops out on both. It's able to sense the link state but stops receiving data after a while.

--Matt

Edit: Here's a vantage compare URL from these clocks: http://3dmark.com/3dmv/3306390

[El Gappo]

http://valid.canardpc.com/show_oc.php?id=1901358

Had the hypers up to around 2400 cl9 on the x8 multi, 9.33 is a bit more flaky. Now to find some stability...

FSB = half of what cpu-z thinks it is hence rejection.

[crazydiamond]

awesome el gappo, glad to hear hypers play a little better with 32nm, glad i kept mine

[mAJORD]

Im thinking, 1.4V in reviews is too much and maybe its a BIOS problem.

I can't post detialed results as i'm about to go out, and my laptop is off atm, but something seem's off with the 1.4v VID's

I can (just) achieve prime stability at 1.11v (measured) , default VID is 1.412 on mine.

Currently doing a stable 3.2 @ 1.235v

So leaving that aside, power consumption is way way too high at 1.4v. At this voltage, measuring at the DC input of a pico ( so 100% efficent on the 12v rail) The platform pulls around ~25w at idle, and ~140w with just CPU loaded (prime) .

I couldn't actually load the GPU also as I was already exceeding the pico-psu's maximum 12v draw! (had to measure within seconds) , but you'd be looking at 160+w if I could going by measurments at lower Vcores

I'll post up much more details tonight when I get home

[imamage]

Weird
1. ASUS F1A75-V PRO with A8-3850 , GPU-Z shows SUMO 9640
2. Biostar TA75A+ pair with the same A8-3850 , GPU-Z Shows HD6550D

I guess the ASUS board BIOS need an update

[imamage]

Weird
The Biostar board "O.N.E." function is broken
I adjust the Bclk to 110 , 105 , 120 and save BIOS

After reboot it shows 2.9GHz in BIOS , CPU-Z and HWINFO32 show the same result.....

[FlanK3r]

Gappo: with aircooling!? Nice, again oldschool "Whau" for you

[El Gappo]

Nah that's on phase cooling mate.

Did drop a proper gpu in and gained a bit of headroom with the cpu. Voltage is definitely not LLano's friend.

Will try and get a round of 2d in today and hopefully some 3D. Could do with some PSC to play with but have none atm.

Would love to know what max fsb others are seeing, feel like I'm bashing my head against a brick wall at 160.

[Daveburt714]

Isn't the PCI-e controller on the APU as well?
If that is locked to the HTRef clock it would seem to me that could be the limiting factor.
High PCI-e clocks could easily reek havoc with other things on the bus, Nics, HDDs, USB and stand alone VC's.

Personally I think we need unlocked multis (Black Editions) before we can really see what these chips can do.

[FlanK3r]

Dave:I think, u has right about it. With SSD and OC Llano is more problems than with classic HDD ( I read it).

[mattkosem]

Dave:I think, u has right about it. With SSD and OC Llano is more problems than with classic HDD ( I read it).

I've had 4 SSDs and an HDD on mine OCd for days without any hiccups. Even had the ref clock at 150 for more than 24 hours now.

--Matt

[El Gappo]

Few quick runs






Imageshat is resizing now



max ram http://valid.canardpc.com/show_oc.php?id=1902289

max cpu http://valid.canardpc.com/show_oc.php?id=1902490

[mattkosem]

Would love to know what max fsb others are seeing, feel like I'm bashing my head against a brick wall at 160.

I'm only using AIR ATM, but 160 seems to be where the fun stops for me too. I can get it to post at 160 with .1v more than 150 takes on the CPU and NB, but it's not stable.

Nice numbers BTW. Seems to really like the phase. Have any 3dm numbers? I bet you'd be over 6k easy in vantage with 150 and a 4Ghz core clock. Probably closer to 6.5 or 7 no?

--Matt

[El Gappo]

Doing that bit now Cheers for that.

[mattkosem]

New AIDA shows more info than the last one, and higher scores at a glance.

--Matt

[El Gappo]

Still on the old one, write is way down O.o

Need to re run that.
Having real issues with hypers in 3D, just isn't happening.

[mattkosem]

Still on the old one, write is way down O.o

Need to re run that.
Having real issues with hypers in 3D, just isn't happening.

Nice Read speed! Copy seems a touch low too, but nowhere near as low as write.

--Matt

Edit: New bench for today. There might be a touch more in these 4GB sticks.

--Matt

[mattkosem]

Seems to run fairly tight timings with these 4GB sticks. I can almost match those numbers at lower clocks.

--Matt

[chew*]

Neither too late nor early to post IMO.

First from coolaler, some OC result.

Second pic Bugged run GT2 bugged the most.

[Formula350]

I was able to bench for a few on 3780MHz or so, before BSoD, caused by running the first FPU test in AIDA heh Did that each time as well, fine on all the CPU and Memory tests, then BLAMO on FPU :\ Mind you, I'm only on a A55 with 4+1 (or might be 3+1+1), but it still does quite well considering the phase power MOSFETs have no heatsink. Hopefully we'll get some other boards, in A75 flavoring, to see what my A8-3850 will really do. Oh, also keep in mind I'm using a similarly budget Scythe Katana III heatsink and not anything high end (though it IS fitted with a 92mm Delta Vane server fan [12V @ 1.5A] at ear splitting speeds lol). Anyways, I didn't bother rebooting to try and nab a screenshot of the highest reached, since unstable speeds aren't important in a review :P

Stable run:CPUz + IBT-Overclocked 3.58Ghz, 1112Mhz, DDR3-2306.PNG

DDR3 2306 10-11-10 2TSANDRA Memory Bandwidth - Overclocked (3.58Ghz, 1112Mhz, DDR3-2306 10-11-10-35 2T 90ns) BIOS 1.0.png

More interesting to come, but I've been up forever, my ability to function is diminished and... yea... I'm going to bed lol

[filmbot]

Formula350, can you give me an idea what your temps were during mid burn in?