Printable View
Quote:
Originally Posted by nickfd
Have you tried memtest in other divider??, I mean 1:1... I was indeed surprised when you said that 4:5 was working. I´ve a pair of D9 rated at 1000 and the only way they work passed the 350 fsb barrier its 1:1.
Its a trc/trfd thing.
read this thread for more info:
http://www.xtremesystems.org/forums/...hlight=eva2000
Quote:
Originally Posted by Iannis
Where did you learn about overclock? Try to low all of your voltages.
I got 2 x 320 Segate 7200.10 in RAID 0 and i'm only getting 60MB/s Read. I'm on the Intel ICHR7 controller. Bios is 1901. What could be going wrong ?
To tell you the truth I can't remember how I got below results awhile back.Quote:
Originally Posted by nickfd
http://img289.imageshack.us/img289/4...0mhzhu7.th.jpg
Quote:
Originally Posted by sombrio
Have U a good guide for this mobo???
I think 6600 is beter CPU for OC than 6700
Scoop,
I have no idea why it's that slow on your side. I got the same drives here and they work fine with a decent speed.
http://img413.imageshack.us/img413/6189/hdtachxz1.jpg
What is officially the most stable bios for this motherboard? If I'm running an e6600 and some ddr2-800 and plan on sticking with vista, what bios will give me stability and performance?
Those voltages are definately high. Are you getting a bad vdroop? I had to really volt my cpu to get it stable before I did a vdroop pencil mod. Now I'm stable at 3.55 with 1.55v in bios. I lowered my clock to 340 for now because I moved the case and temps aren't that great (60ish load).Quote:
Originally Posted by Iannis
One suggestion is to drop the turbo.
Quote:
Originally Posted by iadstudio
When I 'm in 100% cpu load (Folding) I have Vdrop from 1.6125 in the bios to 1.57 (speed fan reading)
ASUS PERMANENT DISEASE -VDROP :mad: :mad: :mad:
Vdroop of less than 100mV is not a problem.Quote:
Originally Posted by Iannis
You have only 41mV, according to Speedfan, measure again with a DMM and you will notice it's even less.
If you take away the droop (vdroop mod agressively tuned) you'll get a heavy increase of PWM-noise and worse overclocking results.
You'll notice that top-OC-people take care of the filter caps to avoid that...
So a vdroop mod might help, but don't push it, leave a little droop as it's part of the regulation circuit.
The voltage regulator of the P5W DH is limited to 1.60V (specs :fact:), better to lift that up with a vcore mod and leave the droop alone.
But if you're on air, 1.60 is already over the safe side for 24/7 use, imho...
And, btw, what you call a permanent disease is actually on Doctor Intel's prescription (check the datasheet of your cpu);-)
Those doctors invented the term "droop"...
Quote:
Originally Posted by chron
1707 or 1901, performance is the same, I didnt have any stability problems
Do you have firmware 3.AAJ in these HD's ?Quote:
Originally Posted by Scoop
The thread is pretty interesting, but it shows already what I know.. Getting over 1000mhz fsb is really difficult. OP in that thread used 2:3 with lower base fsb to do higher clocks in mem, then later went up to 1066mhz mem..Quote:
Originally Posted by monza1412
What's strange is that I got home from work last night, changed dual dimm positions (was in the two orange slots and moved to the two black slots), and it worked. I have no problems in 4:5 and 1:1 now.Quote:
Super Pi @9x355fsb = 3200Mhz 2:3 533mhz 4-4-4-12 at 2.4v
32M Single = 15min 20.453s
32M Dual = 16min 36.657s / 16min 33.500s
So looks like for Super Pi v1.50 32M at least I've managed to get 2x1GB Gskill PC6400PHU2 2GBHZ Micron D9GMH to 533mhz 4-4-4-12 at 2.4v.
CPUZ Validation:
http://valid.x86-secret.com/show_oc.php?id=114620
When I did my memtest'ing yesterday, it was in every memory mode that I had problems; 1:1, 4:5, standard, and overclocked. Like literally memtest86 was spitting out errors thousands by the second.
I'm going to try and debug this later today to see if it is in fact memory, or maybe the dimmslots on the board. Whatever it may be, I'm happy I ordered the other memory and hope that I'll be able to sell this corsair stuff. :)
Another problem I'm having is that most of the times when I reboot it results in an overclocking failure. Is this normal?
Nopez, that's not normal...Quote:
Originally Posted by Mr.X
Maybe you left something in bios on auto
or one of the voltages you chose (vcore, vmch, vmem, vich)
is a little too low...
Perhaps a lowly E6300 is 'special' (or maybe Week 25 is great), but I'm using AUTO on vCore and have no issues whatsoever @2.87Ghz. Other values like vMCH are higher, though. "Trial and error" is the only real way to find out what works, but not changing something after every failure seems bad...Quote:
Originally Posted by OneyedK
400mhz x 8 = 3200Mhz
1:1 mem
vcore 1.35v
Vmch 1.55v
Vich 1.05v
Vdimm 1.9v
FSB termination: Auto
Should I change Vich and/or FSB ter. settings? I haven't messed with those much because I'm not sure what they do.
I have Vich on 1.20 and vFSB on 1.40 myself. Working perfectly on a 9x356 (3.2ghz) OC. :)
Try this:Quote:
Originally Posted by Mr.X
Vmch-1.6/1.65 depends of your motherboard revision
and if is still unstable try raising fsb termination- 1.4/1.5
That should work
imho, most settings are fine...Quote:
Originally Posted by Mr.X
maybe vcore is a tad low, 1.375V could be enough
and give the OCZ's 2.2V, that's what they ask for
FSB termination voltage of 1.20V is enough, never needed more, not even @ 440MHz (but it depends on you cpu, so higher could be better, not with my E6400)
vmch and vich are perfect in my opinion (no need to overheat those chips)
higher is not always better for stability
What revision is your P5w ? Mine was working all the way up to 400 with 4:5 divider ( ddr1000) with VMCH=1.85. I've already tried it with G.Skill 8000 ZX ( D9GKX ) and Mushkin Redline ( D9GMH ) and both passed dual instances of S-Pi 32M. But whether it's Orthos stable or not, that I don't really know.Quote:
Originally Posted by monza1412
I'm runnin' my E6600 @ 420 and my board won't even post at all other than 1:1.
But I agree with you that the lack of trc/trfd control really hindering this board for memory performance.
i've never seen this one.
i was wondering why my gigabit network slowed down, looked at the local area connect icons (i havethe 2nd nic turned off) but the icon on the first nic shows what seems like a share hand at the bottom left. it had the network running at 100mbps insted of 1 gig.. looked back at status i was back at 1 ghz. i have never seen this before and dont even know what 2 put in to search forthis wierd thing. anyone seen this before and if so what the hell is it & how do i turn it off?
thank u
Quote:
Originally Posted by clee
Mine is 1.03 version (1.85v). Please list your memory timings and your voltages settings (Vmch,Vfsb,Vmem)
I ve tried fsb 400/ mem 1000------No boot
fsb 400/ mem 1067------No boot
Quote:
Originally Posted by monza1412
Performance mode=auto
vFSB=1.3
vMCH=1.85
vICH=auto
vDIMM=2.3v
timings shown in screencap. I just did a quick S_pi test as I'm about heading out to work.
I responded to your post because you said you had problem reaching > 350 fsb in 4:5 even with pc8000 D9s, that really puzzles me. Perhaps it's related to the memory internal timing optimization, but it's just a wild guess to get off my chest.
My other G.Skill pair with "supposingly better" D9GKX chips needs higher voltage (2.4v) and more relaxed timing 5-5-5-15 to reach ddr1000 in 4:5.
[IMG]http://img01.picoodle.com/img/img01/...om_eea9310.jpg[/IMG]
@clee :
you think is possible for me to obtain the same result with my ram and mobo?
mine is rev 1.02... (1.65 v).. or norbridge low voltage is a limit?
Ps.: my ram are D9 surely stable @ 925 mhz 4 4 4 8 2.2v.. ( more I havent try..)
I tried yesterday but nope...:( ( 2,4 v 4 4 4 10 & 5 5 5 15.. )
Tanks