what kind of issues? They don't work? I don't see why, 2400Mhz 8-8-8 should be donable w/ a good pair of hyper, and that should be better than 2400Mhz 8-11-8 or even 7-10-7 on a PSCQuote:
Originally Posted by chew*
Printable View
what kind of issues? They don't work? I don't see why, 2400Mhz 8-8-8 should be donable w/ a good pair of hyper, and that should be better than 2400Mhz 8-11-8 or even 7-10-7 on a PSCQuote:
Originally Posted by chew*
APU uses system ram for video memory, Elpida Hyper artifacts and crashes in 3d making it useless even at 9-9-9, we are well beyond 2400 with psc.........Quote:
Originally Posted by Barr3l Rid3r
seems like hypers are yesterday's news:yepp:Quote:
Originally Posted by chew*
1280 7-10-7:eek:, keep it up bud:up:
chew, amazing mem scores. do you have AIDA (Everest) screen?
FCH u mean southbridge or?Quote:
Originally Posted by TsunamiJuan
yeah i belive that equates out to southbridge.Quote:
Originally Posted by Alex-Ro
and cpu-nb for 150 blck?
On most boards I have worked with currently and the APU i have in my possesion it doesn't particularly like CPU NB voltage changes. Changes to that voltage will yeild what acts like a hang before the login section of windows startup. So I have largely Ignored this. You will also notice that most of the time the system don't have that voltage enabled by default. Atleast this as been the case on all three of the motherboards i have in my possession.
No man,CPU-NB like IGP voltage,it controls mem controller also 6550D voltage,i doubt you leave it default,and in my experience plays a major role reaching figh blck meaning high vga clocks.
I guess there is something really wrong happening... if the mem is ok why would it artifact? maybe some wrong sub setting? If them can store any kind of pattern and bits w out problems (prime 95 and memtest stable) at 2200Mhz 8-8-8-24, why it wouldn t work? Can you test it better?Quote:
Originally Posted by chew*
ther is no way to run it even at 1800mhz 7-7-7?
Yes I am doing something wrong.........this is my fist time ever testing hardware.Quote:
Originally Posted by Barr3l Rid3r
You should show us how it's done in 3d with hypers.
Anothers users experience with hypers.
http://www.xtremesystems.org/forums/...=1#post4902470
Wow guess it must be reeaaally wrong to question why a set of memory modules doesn't work... :shrug:Quote:
Originally Posted by chew*
No just repeatedly push the issue.......especially when more than one user is reporting the same, then make comments when not having tested it themselves that people are idiots and don't know how to test.Quote:
Originally Posted by Smartidiot89
There is more than just one comment about it.
Hypers fail at stock
BS not possible
Hypers fail slacked
BS your testing wrong
Hypers are failing in 3d and not optimal
BS Hypers are teh best.
I can get them running on the 1:8 divider with the cpu at stock on the giga 3d stable, 1866 or any overclock = fail.
They are running better on the biostar, 1866 is good and very slight overclocking ( No more than 2000 ) I need to revisit it with that new bios but am testing the msi atm, will check both the msi and biostar again. Give up on the giga it's impossible, chew is right.
This is with multiple kits btw, kingston stt and corsair. They also seem to require more voltage than on other platforms.
I have noticed the same thing El Gappo, I have had to push the voltages much higher on the ram to get it to be stable. Let alone to even get it to clock to stock speeds. I first Noticed this on the biostar board, but it acted the same across the ASUS and jiggy aswell.
Stock volts and almost stock speed/timings for what has been referred to as one of the best damn sets of hyper around.........
http://chew.ln2cooling.com/qdig-file...m%20hypers.jpg
And then 3d at same settings.
3dmark 06 test
http://www.youtube.com/watch?v=cVx_Vr9px_8
3dmark Vantage test
http://www.youtube.com/watch?v=r7RncLsLmoM
3dmark 11 test
http://www.youtube.com/watch?v=vS7ycFSExj4
Now when I don't suggest using them for this platform you can understand it's not some grand conspiracy against hypers.........
My parts came in.. woot! Probably won't see anything fantastic from me, My RAM doesn't clock as well.
What kind of ram do you plan on running charged, you might be surprised how well some sets are doing.
Here's the AIDA results someone requested at 2560 7-11-7.
I refuse to pay for another key so the newer verion is just trial.
http://chew.ln2cooling.com/Llano/aida%201.jpg
http://chew.ln2cooling.com/Llano/aida%202.jpg
Just got my cpu and board in. Now I can try for my absolute best possible scores, which should be around 10% lower than Chew* :D
Which board did you get massman?
How will be good RipjawsX 2133 8-9-8 at this APU?
GIGABYTE A75-UD4H.Quote:
Originally Posted by TsunamiJuan
From the looks of this thread, it should be serving my overclocking needs quite well :p:
Cas 8 is 24 hours windows memtest(x4) stable at stock volts on 16GB of these Corsair XMS3 sticks. Nothing close to what I'm seeing on the flares, but not bad. Will push further later. One thing is certain, this platform sure makes memory clocking easy.
--Matt
Passing Memtest on Llano is easy ;)
Start 3DMark06 instead and make it loop GT-1.
I've not run into any circumstances where passing 24 hours of memtest in windows with one instance per CPU core yields anything less than complete memory stability, even on this chip. What memtest are you using where that isn't the case? I was playing DNF for hours before I fired up memtest, so it seems stable enough for me. Perhaps you're closer to the ragged edge of stability than I am at these current clocks?Quote:
Originally Posted by The Stilt
--Matt