"ASUS P5Q Deluxe discussion".

Try this:

FSB 450
PCIE 100
PSB strap to NB 400
DRAM freq 900
DRAM manual timing: 1st info 5-5-5-18-3-55-6-3

DRAM static - disabled for ram overclock
DRAM read training Auto
Mem OC charger Enabled
Ai clock twister Moderate
Ai transaction booster Manual
Common performance level 12

CPU voltage 1.3500 (BIOS)
GTL AUTO
GTL AUTO
PLL 1.54
FSB 1.34
DRAM 2.02
NB 1.3
NB GTL Auto
SB AUTO
Sata AUTO
Load line calibration Enabled

If that doesn't work try to bump your vNB up to 1.35v and retest.

If it passes, reduce the vcore 1 notch at a time until you find what it needs.

Quote:

---

Originally Posted by Mumid

as soon as you put anything in the second slot it will drop them both to x8

---

Are you able to tell me where you acquired this information? It seems odd that they would have two x16 slots and make it impossible to use them both. I know marketing departments are not above such scams, just want to make sure you have not pull that 'fact' out of the air.

1.35v NB for 450fsb :confused:

Quote:

---

Originally Posted by Runkett

Are you able to tell me where you acquired this information? It seems odd that they would have two x16 slots and make it impossible to use them both. I know marketing departments are not above such scams, just want to make sure you have not pull that 'fact' out of the air.

---

I tested it myself but it is fairly obvious. There are only 16 electrical lanes for the first two slots so you can see that its impossible for them to both run at 16x

Quote:

---

Originally Posted by gurusan

1.35v NB for 450fsb :confused:

---

He is running a quad so that takes more vNB to run at a given FSB than a dual core. It's a moderate bump up from where he is at so he can eliminate the MCH as his restriction.

It may also be his CPU taking more than the average cpu of that type, so bumping up both can help eliminate things as the cause of the instability.

Quote:

---

Originally Posted by BeastNotro

Looks good 8 x 475 and Linkpack stable .. woot woot

Now i always state do Orthos to for 8-10 hours or 8-12 hours.

I can make everything out, but if a person has bad eyes (no offence) they may not.

You may as well wait until new bios comes in to Orthos and then Linkpack it again.

I ran 3 Linkpacks all at different memory FREE . like 1887, 2000 and 2200 .. and all passed.

2200 that is like about over half of the memory i have.

---

What picture hosting site is the best for pictures people can see? That site there wants you to pay money for better visible pictures, f them.. Thanks,

imageshack.us.

use the thumbnails.

Quote:

---

Originally Posted by truehighroller

What picture hosting site is the best for pictures people can see? That site there wants you to pay money for better visible pictures, f them.. Thanks,

---

As said above, http://www.imageshack.us/ is the way to go.

also try www.imagecross.com, works nice too. i use it for most things, limitations are pretty high that most wont even come close to being bothered by them

Thanks guys. I noticed everyone had better looking pics then me and wondered who to use. I will post another better pic leter tonight when I get home.

Quote:

---

Originally Posted by CryptiK

He is running a quad so that takes more vNB to run at a given FSB than a dual core. It's a moderate bump up from where he is at so he can eliminate the MCH as his restriction.

It may also be his CPU taking more than the average cpu of that type, so bumping up both can help eliminate things as the cause of the instability.

---

dunno, I've had about 6 different quads with this board and never needed above 1.28v for 500+fsb, and I am running 4x1gb dimms as well.

Quote:

---

Originally Posted by Runkett

Are you able to tell me where you acquired this information? It seems odd that they would have two x16 slots and make it impossible to use them both. I know marketing departments are not above such scams, just want to make sure you have not pull that 'fact' out of the air.

---

You can find it on Asus' site. When running one GPU, the slots are x16/x1/x4 from top to bottom. When running Crossfire, the slots run x8/x8/x4. The bottom slot will always be x4 as it is run through the southbridge.

THe x8/x8 Crossfire is a P45 limitation... if you want two x16 electrical (not just physical) slots, you're either going to need X38/48 for Crossfire or a 780/790 board for SLI.

gurusan: what vNB do you reckon I need to hit over 485MHz-FSB?

I'm pretty much stuck at that speed (E6300) but I have my vNB pretty low (1.12v-1.14v).

Will form out template now for full technical assist but as your talking about vNB vs high FSB I thought I would pick your brains! :)

Quote:

---

Originally Posted by gurusan

dunno, I've had about 6 different quads with this board and never needed above 1.28v for 500+fsb, and I am running 4x1gb dimms as well.

---

Fair enough, perhaps it is something else holding him back. It was simply a suggestion to attempt to alleviate his issues, only him testing it can determine what is holding him back. However I have never run a quad on this board, so I don't know from personal experience what vNB is required for them to hit 450 - 500 FSB. Thanks for your input.

I have hit 590FSB on only 1.26v vNB so I wouldn't think it would be an issue for him but just suggested he bump it up to eliminate it as a possibility.

Quote:

---

Originally Posted by Big.Wayne

gurusan: what vNB do you reckon I need to hit over 485MHz-FSB?

I'm pretty much stuck at that speed (E6300) but I have my vNB pretty low (1.12v-1.14v).

Will form out template now for full technical assist but as your talking about vNB vs high FSB I thought I would pick your brains! :)

---

Try 1.24 or 1.26v...maybe even lower if you just have 2 dimms full.

Quote:

---

Originally Posted by CryptiK

Fair enough, perhaps it is something else holding him back. It was simply a suggestion to attempt to alleviate his issues, only him testing it can determine what is holding him back. However I have never run a quad on this board, so I don't know from personal experience what vNB is required for them to hit 450 - 500 FSB. Thanks for your input.

I have hit 590FSB on only 1.26v vNB so I wouldn't think it would be an issue for him but just suggested he bump it up to eliminate it as a possibility.

---

well yeah your reasoning is correct and a good way to eliminate possible culprits.

Which LAN port is on the PCI-E bus in BIOS? The top one or the bottom one?

Receiving my P5Q-D tomorrow and I wanna be ready! :D

Quote:

---

Originally Posted by gurusan

Try 1.24 or 1.26v...maybe even lower if you just have 2 dimms full.



well yeah your reasoning is correct and a good way to eliminate possible culprits.

---

I am at 1.28v nb to get it intel burntest stable at 3.8Ghz. I bumped it up from 1.26v because it wasn't stable.

Quote:

---

Originally Posted by Polizei

You can find it on Asus' site. When running one GPU, the slots are x16/x1/x4 from top to bottom. When running Crossfire, the slots run x8/x8/x4. The bottom slot will always be x4 as it is run through the southbridge.

THe x8/x8 Crossfire is a P45 limitation... if you want two x16 electrical (not just physical) slots, you're either going to need X38/48 for Crossfire or a 780/790 board for SLI.

---

The wording just seems dishonest - Clearly states that crossfire is limited to x8 + x8 - no issue there, but it also says it has 2x x16 slots and only says they drop to x8 in 'link mode'. It seems intentionally worded to imply that using those slots for a purpose other than crossfire would not reduce the lanes. So if there is no way to use the 2nd slot as x16 (bar leaving slot 1 empty or disabled) I'm not sure how they can advertise 2 x x16 pci-e 2.0 slots.

Its not the end of the world or anything (my intent was using an 'old' 8800gts in the 2nd slot as a PPU- thus not crossfire or 'linked') just cant help feeling I have been lied to and cheated just a little. And of course they make the third slot so low you cant use double width cards (x4 would have been fine for a ppu) and get around it that way :shakes:

I always understood that pcie2.0 and 8x/8x wont be a bottleneck...Thats about equal to pcie1.1 16x/16x.

They say it's got 3x x16 slots, but that just means physical size. The electical size (x16/x1/x4 or x8/x8/x4) are what really matters and what you have to pay attention to. They just include the x16 length so people with RAID cards or other expansion cards have a little more flexibility like running an x8 length card in an x16 slot at x4 speed (for instance, in the bottom slot).

Quote:

---

Originally Posted by RAdoxx

Which LAN port is on the PCI-E bus in BIOS? The top one or the bottom one?

Receiving my P5Q-D tomorrow and I wanna be ready! :D

---

Its the bottom one mate, and if you want to disable the PCI one in the bios, it is labeled as 'Marvell Lan 2'.

http://img503.imageshack.us/img503/7...aug2008ia8.jpg

Intel E6300 + TRUE
Asus P5Q-Deluxe (#803)
Crucial Special 10th Anniversary 2GB (2x1GB)
Radeon HD3450 256MB
Corsair HX620W PSU

http://img103.imageshack.us/img103/8...aisedabbr3.jpg

ASUS P5Q DLX Template :

JumperFree Configuration Settings
AI Overclock tuner: Manual
CPU Ratio Setting: 07.0
FSB Strap to North Bridge: AUTO
FSB Frequency: 485
PCI-E Frequency: AUTO
DRAM Frequency: DDR2-971
DRAM CLK Skew on Channel A1: AUTO
DRAM CLK Skew on Channel A2: AUTO
DRAM CLK Skew on Channel B1: AUTO
DRAM CLK Skew on Channel B2: AUTO
DRAM Timing Control: Manual

1st Information :

CAS# Latency: 5
DRAM RAS# to CAS# Delay: 5
DRAM RAS# Precharge: 5
DRAM RAS# Activate to Precharge: 15
RAS# to RAS# Delay : AUTO
Row Refresh Cycle Time: AUTO
Write Recovery Time: AUTO
Read to Precharge Time: AUTO

2nd Information :

READ to WRITE Delay (S/D): AUTO
Write to Read Delay (S): AUTO
WRITE to READ Delay (D): AUTO
READ to READ Delay (S): AUTO
READ to READ Delay (D): AUTO
WRITE to WRITE Delay (S): AUTO
WRITE to WRITE Delay (D): AUTO

3rd Information :

WRITE to PRE Delay: AUTO
READ to PRE Delay: AUTO
PRE to PRE Delay: AUTO
ALL PRE to ACT Delay: AUTO
ALL PRE to REF Delay: AUTO
DRAM Static Read Control: AUTO
DRAM Read Training: AUTO
MEM. OC Charger: AUTO
AI Clock Twister: AUTO
AI Transaction Booster: AUTO

CPU Voltage: 1.41250
CPU GTL Voltage Reference (0/2): AUTO
CPU GTL Voltage Reference (1/3): AUTO
CPU PLL Voltage: AUTO
FSB Termination Voltage: 1.24
DRAM Voltage: 1.84
NB Voltage: 1.14
NB GTL Reference: AUTO
SBridge Voltage: 1.20
PCIE SATA Voltage: AUTO

Load Line Calibration: Enabled
CPU Spread Spectrum: Disabled
PCIE Spread Spectrum: Disabled
CPU Clock Skew : Auto
NB Clock Skew : Auto

Advance CPU Settings
CPU Ratio Setting: 0.7
C1E Suppport: Disabled
Max CPUID Value Limit: Disabled
Intel® Virtualization Tech: Enabled
CPU TM Function: Enabled
Execute Disable Bit: Enabled


Ok guys could use a little pointer or two, kinda stuck trying to get my old but nice E6300 much over 485MHz-FSB, I am approaching this very methodically slowly but surely adjusting voltages but I'm getting mixed results. I have never been able to get this chip fully prime stable over 3.4GHz on my previous boards (P5B-Deluxe and P5K-E) so I suspect either the chip has some kinda FSB wall or perhaps I am doing something wrong?

Using only Prime Small FFT's I have found the approx vCore needed to keep the chip stable but once I switch to blend it errors from 5 mins to 2 hours. I finally managed to get the 8 hour prime shot above by upping the vTT (or vFSB) from 1.22 to 1.24, the vNB from 1.12 to 1.14 and the vDimm from 1.80v to 1.84v

Thought I cracked it so being a stickler I then lowered the vTT and vNB to see if it started to error again which it did, so I put the two settings back to where they were for the prime pic above but it failed?

http://img503.imageshack.us/img503/6426/umkn3.th.jpg

I need to speed up my testing a little so would appreciate a little insight from those that have walked the path before me. I am now gonna pump the vFSB/VTT and vNB and vDimm. I have also tried dropping the multi from x0.7 to x.06 to take the strain of the CPU but that didn't seem to help.

The PC won't POST with the FSB set over .. um about 488MHz-490MHz FSB so its like a full brick wall and where I am at present (485MHz-FSB) is the beginning of the *ragged edge*

p.s: I'm aware that I need to take PLL from auto but thats how it was when I started testing and I didn't want to start again, if you guys think that may be part of my problem then I'm happy to change it now!

p.p.s: I have previously tested my memory up to about 533MHz (DDR2-1066) so I don't think thats the problem? however If I don't get some success soon I will back down my Overclock and then set the memory on a divider and give it a good testing just to be sure. I think the problem is either the CPU or incorrect mobo config so thanks in advance!

I think you're going to want to change that CPU vPLL to something other than Auto... Auto overvolts like mad.

Sounds like you are nearing the limit of the cpu, but I wouldn't give up yet. I had an e6300 in a P5B vanilla before this current setup, and it was stable at 490 FSB at 1.37v vcore. vFSB didn't help it at all, it just added heat, but your cpu may be different.

Try bumping up the vNB and vdimm a little to eliminate them as possibilities. I would use 1.26v vNB, 1.90v vdimm. Leave your timings at 5-5-5-15.

Also try and set vPLL to 1.52v, vFSB to 1.24v - 1.26v

Also bump up the vcore 1 step.

Try and see if that gives you better stability.

I have also seen gains in e6300's by adjusting the GTL values, i would try raising them by single steps to see if they may help you.

Quote:

---

Originally Posted by Polizei

I think you're going to want to change that CPU vPLL to something other than Auto... Auto overvolts like mad.

---

Yeah I heard what people have said about this PLL thing, not 100% convinced that jVidia's findings are relevant to a dual-core Conroe?

Let's say it was over-volting the PLL would that lead to the errors I am encountering? I have no idea about PLL and wouldn't know what to change it too (it has been suggested 1.6v is a good place to start).

Just means I have to start my testing from scratch but unless someone chimes in with a fix I will have no choice :(