View Full Version : Best Mobo for QX6850? Problems :-/
Nightbane
08-06-2007, 02:25 AM
Hi!
I'm new @ this forum, and want to make few questions, Hope you can help me ^_^
Got here a QX6850 ES + Scythe Ninja, OCZ 9200 and a crappy p5wdh.
In first test I got a nice 3dmark 2006 with 1.48v @ 3.9ghz (300x13, ram @ 900) with a default 8800GTX ~13300 marks.
All was fine, but I cannot pass from that result. No more voltage help chip, not less fsb and more multiplier helped too... So i'm a bit stucked, and think maybe is matter of mobo.
Maybe a p5k3 will help this? If i am honest I tryied lowing multiplier and with same config but lower MHZ 3dmark got pass without problems. All problems come with more than 3.9ghz and always in CPU test. Within a few seconds from start it reboots (no blue screen here).
So it could be too matter from temperature :(
What do you think?
Thanks!
welcome to OCX mate :ocx44:
you must have forgotten, that 3.9GHz@1.48 through 06 is already very good :eek: and if you want more - may be the chip just can't do it ;)
if you want a better result - OC the graphic card, or run 3DMark with something like 430x9 or 390x10 (more FSB = better CPU score)
Nightbane
08-06-2007, 02:33 AM
Thanks Buttertoast.
I'm sure is a good result, so thats why im disappointed, 3.9 with only 1.48v (read from bios, real vcore was something about 1.5350) and I cannot pass from this?
Maybe its problem of temperature, ninja cannot got 4 cores at more than 1.5v . Bios reporting temperatures were something about 120ºC, so I cannot trust them... probe got crazy xD
Don't know, but want more :D
Thanks
Dynasty
08-06-2007, 09:56 PM
get an P5K-Prem. for DDR2 or P5K3-del. for DDR3......
you will love it with either one using qx6850 G0.
here is some stuff i did on air with one of my qx 6850 G0's under P5K-prem.
PI 1M sucks so dont put too much into PI time...:p
http://img258.imageshack.us/img258/3448/41039mhzqx6850g0156vc1mrc2.gif (http://imageshack.us)
Nightbane
08-07-2007, 04:25 AM
Ill take your advice and get a p5k prem, got here some nice ddr2 to play with :D
Thanks dinasty
if you want to save some money - you can get a P5K-E (which basically is a bit "simpler" version of P5K-Premium with all the OC potential left in) :)
Nightbane
08-07-2007, 04:30 AM
Better tho, Cannot get p5k premium here atm, but p5k-e is in stock now ^^
So if its basically the same, ill get a p5k-e
Thanks Buttertoast :D
the only diff I can spot is CPU/Mem feeding chain, chipset cooling, and number of PCB layers ... which IMO doesn't worth the extra cash
Cpt Twitchy
08-07-2007, 01:32 PM
the only diff I can spot is CPU/Mem feeding chain, chipset cooling, and number of PCB layers ... which IMO doesn't worth the extra cash
Different amount of PCB layers is a big thing. That means the board is designed completely different.
Eldonko
08-07-2007, 02:36 PM
I think it will be pretty much the same, both use the ICH9R SB and have the 8 phase power.
Nightbane
08-09-2007, 03:15 AM
I'm still testing the chip with a ninja heatsink and a p5wdh...
I think the problem is the vdroop (1.49~1.45v with 1.55v set in bios) and the temperature by air (bios reads 120ºc XD but i'm sure is something around 60ºC in full)
With more vcore than 1.6v in BIOS (1.52v real) the system dont got into BIOS, it reboots constantly...
The max i got atm is quad spi 32m at 3850 1.45v (1.50v in bios) and 14619 3dmark 06 with a slightly oc'ed 8800GTX (672/1035).
Thanks all :D
Different amount of PCB layers is a big thing. That means the board is designed completely different.
Yes,
theory behind it is ground noise isolation, inductance and signal crosstalk reduction, how it pans out in the real world is interesting and perhaps worthy of a TTR style comparison of 6 and 8 layer models. For overclocking DDR3 to high mhz, 8 layer boards are pretty much a theoretic pre requisite.
I expect all the better DDR3 boards will be 8 layer based, DDR2 benefits are not likely to be as prominent, though noise sensitive chipsets like the 680 could certainly have gained from the increased trace isolation..
Of course pcb layers start to become a little benign if the pcb design engineer is at fault for bad trace routing in the first instance..
regards
Raja
Nightbane
09-01-2007, 08:53 PM
Hello again! :)
Today I tested the chip with a home made single stage (-50ºC evap temp with 6600@4.7 1.6v) Cannot got temperatures coz no UEI here, and BIOS only shows 4ºC.
Got Spi 1M @ 4500mhz 1.56v, 3D03 @ 4450 mhz 1.58v (43983 with stock 8800GTX) and now I'm going to try some 3d 06.
Unfortanetly that single stage is not prepared to a quad monter at 1.6v, I'm not sure but I think temperature in chip will be something near -15 or -20ºC (35º amb temp).
BTW chip is going right, inst?
Regards
vBulletin® v3.8.0, Copyright ©2000-2010, Jelsoft Enterprises Ltd.