RD600 first look and general questions

well as my luck seems to be running this new RD600 will not run my crucial 10th anniv at 500...memtest86+ 1.70 test 5 has errors after 7 passes at 2.2v

old board 5-6-5-5-2t @ 2.2v 500mhz...50 passes of test5, winmemtest 1000% no errors, P95 v251 blend 20hrs...

new board....memtest86+ 1.70..fails after 7 passes

this board seems to run hotter chipset too...same cooling as old board(old board idle at 47/48C with 1.42v, load was 54C)...new board at 1.42v idle at 56/57...loads 62c..stock HS with 120mm fan blowing on it and ram

voltage settings..
vcore 1.51875 bios setting=1.53v bios reading old board 1.52v
NB 1.42 bios setting=1.45v bios reading.... old board 1.44v
NB 1.36 bios setting= 1.39..... old board 1.38v
vdimm 2.2v bios setting=2.22 v...old board initial showed 2.2v as well..but then after awhile showed 2.24 v

SG show at 1.51875 v....under load alot of flucuation on the new board..go from 1.48-1.52v....old board 1.48v with occ. 1.5v flucuation

So I am not sure if voltages are more or less stable

Quote:

---

Originally Posted by nealh

very nice temps..my ambients if Florida tend to run 25/26C

I would guess a higher fsb will add more heat...not sure though

---

Heheh i live in the north of Sweden still winter here, 1m snow and below 0c most of the time:)

Quote:

---

Originally Posted by nealh

Are you getting 55C on NB at idle with water

I only have SG as a guage for chipset temps....idle with 1.42v at 59/60..seems much hotter than the first mobo...quick load temp was 63/64 C

---

Er....I said 55 on Stock Cooler...I know this Cos I had to use it before my NB DD Block arrived. So i'd be guessing thats about right for your NB Temp. The DD Block gives me about 38Deg Idle. These are Bios Readings as well.

Anyways, seems like my board has thrown a Pink Fit somewhere!? For the life in me, I can't Re-Install WinBl0ze. I keep getting copying file Errors. I have tried 3 different XP Pro Original Disks, all with the same cr4p. I swapped CDROM's, replaced Cables...same sh1t!
Okay, not entirely thwarted, I boot from a Win98 Floppy. Thankfully, I have a couple of FAT32 Partitions around my drives.
So, back to DOS cr4p, we go:
Smartdrv /v (enables Caching of Drives @ DOS Level - need cmd on the boot floppy tho')
Mkdir D:\I386
Copy F:\I386 D:\I386
One coffee break l8r, 5,709 files Copied!
Rest Comp, boot from Floppy. At the cmd prompt:
D:\I386\winnt

& off it goes....copies all the files & restarts the puter. At the F6 prompt, it takes the drivers for the HighPoint SATA & then still complains it can't copy certain files whilst trying to set itself up?? This is from the hard drive to itself dammit so it cant be the CDROM side of it unless the files it did copy are corrupted & DOS didn't say so! It says the file image it has copied is corrupted, ENTER to try again, Esc to Skip etc even if I install from CDROM or the Hard Drives!! Trying to skip the files is futile cos windows doesn't set itself up right & Device manager doesn't handle the system devices correctly.
Hard drives are fine & pass extensive Software tests @ DOS.
Funny thing is, I now ALWAYS loose the USB Keyboard & USB G5 Mouse just as I have to say YES to Authorise the RR2310 Driver for the First time. It never did this to me the first time I Installed on this RR card. The Image I have restored now is the very one I created after f'kin Installing it without Issues????
I remember my DFI nF4 SLI DR Expert board running 3 CDROMs for about 2 months. Then my Plextor stopped functioning, along with a Pioneer 120 Slot Drive DVD Reader. The only CDROM device that worked was the Pioneer 109 DVDRW & that was the MASTER on the Primary header. Bios Updates NEVER corrected it....wtf?
I have an RMA already & now scared the Supplier wont find anything wrong with the board when they install on the Boards Headers. I can't for the life in me understand why this set up without Issue on the RR First time round only to go sh1t faced on me for trying a 200 Strap. I wonder if it lies with knowing 200 STRAP @ 286fsb was max possible for me @ BIOS level, but I did manage to get 336FSB using AMD System Manager. Since that point, all I have seen is random crashing & System errors in Explorer, IE, WINRAR. Orthos wont run without wanting to send an Error report to Micro$0ft.
Gunno try installing from the Promise headers to see if I get file copy issues so I guess it's time to fire up the Old Amiga again. :cool:

Kin Hell..I figured %%c on WB would be a real issue...

I am sad with the new board...the NB runs hotter at idle by 5C..load seems 3-4C higher at a LOWER NB voltae of 1.36 bios setting with bios reading of 1.39!!!!(I am back to same cooling as before stock HS and 120mm fan...)

I want a waterblock but very concerned about maintaing contact on the NB with MCW30 and wire clips....and really do not want to spend $40 on blaock that screws down but will only work on this board

I am considering getting a Swiftech MCX 159...but I heard the fan is noiser vs the Microcool whipser XE..which I have but can not with the shim get reasonable contact on the chipset...also the post that attach to the wire clips do not fit correctly the hole does not line up correctly and the s,mall screw that tightens the mount took 45min to get installed.....

Quote:

---

Originally Posted by nealh

Kin Hell..I figured %%c on WB would be a real issue...

I am sad with the new board...the NB runs hotter at idle by 5C..load seems 3-4C higher at a LOWER NB voltae of 1.36 bios setting with bios reading of 1.39!!!!(I am back to same cooling as before stock HS and 120mm fan...)

I want a waterblock but very concerned about maintaing contact on the NB with MCW30 and wire clips....and really do not want to spend $40 on blaock that screws down but will only work on this board

I am considering getting a Swiftech MCX 159...but I heard the fan is noiser vs the Microcool whipser XE..which I have but can not with the shim get reasonable contact on the chipset...also the post that attach to the wire clips do not fit correctly the hole does not line up correctly and the s,mall screw that tightens the mount took 45min to get installed.....

---

Yeah with the MCW30, it really is a cr4p fixing. I got the Danger Den RD600 Block. It uses white metal T-Bar Pegs with a bit of a hook shaped in to the bottom Leg of the "T". You push these T-Bars through Small Springs which have PTFE Collets each end. Whilst trying to keep the Block Level to avoid crushing the NB Core, hook the T-Bar onto the Boards cr4p clips! Eventually, it all goes in but the Left Leg ot the "T" is pointing uptowards the Graphics card & clears by about 1mm. Dead dodgey, Pressure is much better than the MCW30. With all the Tygon Piping in my V2000B case, it's bloody hard trying to get hands in such confined spaces.

On my side of things, I cant run Microsoft Memory Test from floppy without getting errors on ALL tests. The board seems to be getting worse & the Images I was re-installing are now NOT working!? I'm writing this from a Safe-mode boot with Networking enabled....

At the current rate of depreciation, the board should be DEAD & un-useable by teatime tommorow. :rolleyes:

I am at a complete & Utter f'kin Loss with this cr4p! :confused: :( :slapass:

Quote:

---

Originally Posted by Kin Hell

Yeah with the MCW30, it really is a cr4p fixing. I got the Danger Den RD600 Block. It uses white metal T-Bar Pegs with a bit of a hook shaped in to the bottom Leg of the "T". You push these T-Bars through Small Springs which have PTFE Collets each end. Whilst trying to keep the Block Level to avoid crushing the NB Core, hook the T-Bar onto the Boards cr4p clips! Eventually, it all goes in but the Left Leg ot the "T" is pointing uptowards the Graphics card & clears by about 1mm. Dead dodgey, Pressure is much better than the MCW30. With all the Tygon Piping in my V2000B case, it's bloody hard trying to get hands in such confined spaces.

On my side of things, I cant run Microsoft Memory Test from floppy without getting errors on ALL tests. The board seems to be getting worse & the Images I was re-installing are now NOT working!? I'm writing this from a Safe-mode boot with Networking enabled....

At the current rate of depreciation, the board should be DEAD & un-useable by teatime tommorow. :rolleyes:

I am at a complete & Utter f'kin Loss with this cr4p! :confused: :( :slapass:

---

I think it is the sync mode and 1T...lower back to async and 2T...I bet the memory is corrupting everything(not that the sticks are bad but the stupid board cant handle what it promises)..or boost your voltages for the NB

Like I said ...my old board loved my ram at 500 with 2.2v..5-6-5-5-18-24-42-2t (winmemtest for 1000% 4hrs, 50 passes test 5 and 20 hrs of P95 blend without a failure)...new baord 500 with 2.2v 5-6-5-5-18-24-42-2t fails memtest 7 in 7 passes......

Quote:

---

Originally Posted by nealh

I think it is the sync mode and 1T...lower back to async and 2T...I bet the memory is corrupting everything(not that the sticks are bad but the stupid board cant handle what it promises)..or boost your voltages for the NB

Like I said ...my old board loved my ram at 500 with 2.2v..5-6-5-5-18-24-42-2t (winmemtest for 1000% 4hrs, 50 passes test 5 and 20 hrs of P95 blend without a failure)...new baord 500 with 2.2v 5-6-5-5-18-24-42-2t fails memtest 7 in 7 passes......

---

I hear ya dude....I really do. But nothing is working anymore @ defualt settings.

Despite my board apparently being totally JFK'd, it is still arse-ing about with it's usual sh1t. you know, where it crashes & you have to Rest CMOS before applying your over-clock settings! I mean....wtf? At least I'm at desktop with a restored Drive Image again & this text is a Copy/Paste from desktop notes as I've crashed 3 times trying to f'kin post this Sch-peel.
These boards are obiviously having some real bad issues with themselves & the Bios not being Straight isn't helping matters either.

If you got one that works, be sure you are using the fastest DDR2 Ram possible & enjoy the Bull-sh1t "Async" memory timings. Do not persue 1:1 timings as the booard WILL kill itself despite very very safe voltages being used on the overclocks & all whilst under water....so Heat is well out of the equation RE this anomally.

I'm just dissapointed I didn't get any smoke out of this puppy....I so wanted to see it Smoke its sad 4rse! It more than desrves it. :mad:

One last thing to do before RMA time. I Just gotta try & install off the standard headers. I refuse to Accept this PCI-e High Point 2310 Raid card is causing issues or that my RAM is JFK'd. Everything was fine untill I tried some 200 Strap Runs from 286FSB up. JFHC @ this cr4p!

*edit*

Can't install on Native SATA Headers...still corrupts files when Copying from CDROM, still crashing randomly.....bleh! :slapass:
My RAM works fine in another board & the other boards RAM which is fine is Fuxord in this RD600.
This boards Ass is Grass on Monday! :mad:

Quote:

---

Originally Posted by Kin Hell

I hear ya dude....I really do. But nothing is working anymore @ defualt settings.

Despite my board apparently being totally JFK'd, it is still arse-ing about with it's usual sh1t. you know, where it crashes & you have to Rest CMOS before applying your over-clock settings! I mean....wtf? At least I'm at desktop with a restored Drive Image again & this text is a Copy/Paste from desktop notes as I've crashed 3 times trying to f'kin post this Sch-peel.
These boards are obiviously having some real bad issues with themselves & the Bios not being Straight isn't helping matters either.

If you got one that works, be sure you are using the fastest DDR2 Ram possible & enjoy the Bull-sh1t "Async" memory timings. Do not persue 1:1 timings as the booard WILL kill itself despite very very safe voltages being used on the overclocks & all whilst under water....so Heat is well out of the equation RE this anomally.

I'm just dissapointed I didn't get any smoke out of this puppy....I so wanted to see it Smoke its sad 4rse! It more than desrves it. :mad:

One last thing to do before RMA time. I Just gotta try & install off the standard headers. I refuse to Accept this PCI-e High Point 2310 Raid card is causing issues or that my RAM is JFK'd. Everything was fine untill I tried some 200 Strap Runs from 286FSB up. JFHC @ this cr4p!

*edit*

Can't install on Native SATA Headers...still corrupts files when Copying from CDROM, still crashing randomly.....bleh! :slapass:
My RAM works fine in another board & the other boards RAM which is fine is Fuxord in this RD600.
This boards Ass is Grass on Monday! :mad:

---

Problem is my new board will not run my ram at 500mhz like the first board..the setup was 100% stable with mmemtest86+, winmemtest and P95 blend

new board 500mhz just errors out..so I back to 400mhz 4-4-4-18-2t

this boards seems to overvolt slightly more than my other board..so I am going to try a slightly lower vcore...I just finished 26 hr of P95 blend at 3500(9x390) with ram 2.0v 400mhz 4-4-4-18-2t, async, NB 1.36v

woah.. my board is sweet with me.. :D I hope it will live till I have my upgrade..


hope you guys can sort things out.. my problem is my proc.. its concave.. :( so I cant push it to the limit..

Well im on 360*9 with 1.365v to the processor and every other voltage stock.

Tried going above but orthos fails within mins and I'm taking it slowly. Next step will be GTREF Settings before any more voltage increases. Took CPU 1/2 3/4 up to 84 and couldn't improve stability at 360 before upping the CPU volts.

Anyone else played with GTREF settings? If so, what did you find?

RAM is async 450 4-4-3-7 2.2v but the orange slots don't want to run dual channel? Yellow ones are fine.

Can't use the new bios despite using a PS/2 keyboard which I have from the start.

Quote:

---

Originally Posted by Mekrel

Well im on 360*9 with 1.365v to the processor and every other voltage stock.

Tried going above but orthos fails within mins and I'm taking it slowly. Next step will be GTREF Settings before any more voltage increases. Took CPU 1/2 3/4 up to 84 and couldn't improve stability at 360 before upping the CPU volts.

Anyone else played with GTREF settings? If so, what did you find?

RAM is async 450 4-4-3-7 2.2v but the orange slots don't want to run dual channel? Yellow ones are fine.
Can't use the new bios despite using a PS/2 keyboard which I have from the start.

---

have you checked the biossetting on memory channels..set to AUTO or A&B

I would just up vcore..set gref to 48/48/128..set bios vcore to 1.4v and NB to 1.36v see how you do..but again..lower memory timings or you will confound the cpu overclocking..you need ram out of the equation

you can go back later and tweak the ram at given fsb

I really believe as you increase the fsb this stresses the chipset and lower ram timings will help get more cpu speed at a given voltage....cpu speed is still most important...

I can tell you running 3500 with 400mhz(4-4-4-18-2t) on my setup is as fast in superpi 1M, 3dmark03/05/06(these are slightly faster) as compared to 3500:500(5-6-5-5-18) on my old board

my new board will not run my 10th anniv at 500mhz with memtest errors at 2.2v

BTW I was 20hrs blend stable on old board at 3500:500...3500:400 is 26 hrs blend stable in P95 when I stopped it yesterday

Yeah tried auto and A & B so think my orange DIMs are fubar.

I'll slacken the timings up.

Hard posts zu read, I still testing the system and waiting for the new rma OCZ PSU Evo then I will start to overclock my system. The first start with a old Tagan runs fine :)

Quote:

---

Originally Posted by nealh

Problem is my new board will not run my ram at 500mhz like the first board..the setup was 100% stable with mmemtest86+, winmemtest and P95 blend

new board 500mhz just errors out..so I back to 400mhz 4-4-4-18-2t

this boards seems to overvolt slightly more than my other board..so I am going to try a slightly lower vcore...I just finished 26 hr of P95 blend at 3500(9x390) with ram 2.0v 400mhz 4-4-4-18-2t, async, NB 1.36v

---

100Mhz Difference in boards is taking the Pi$$ (unless the last one JFK'd your Ram) & just says VOLUMES (DFI need need to Read & Digest this cr4p....We as Paying, "-oh-my-god-lets-bend-over-real-hard" Punters, ....are!!)

Anyway, I was right...my board is Totally Dead now, so RIP RD600 #3 & "Adios Amigos" to DFI. Will I be buying the DFI 680i @ the End of April?
Doh!.......mmmmmmm......Doughnuts! :nono:

For you fellas "hoofing" around with the Ram Slots, Orange (DIMM1 & 2) are the ones to use for Dual Channel at Max Bandwidth potential. You will have a slightly reduced bandwidth for running Dual Channel on the Yellow Slots.
If you use an Orange & a Yellow one, you wont be running the Ram in Dual Channel mode at all & the bandwidth Drop is huge.
Highest Multiplier & Fastest FSB possible is the way for the best performance on this board. IMHO, an E6600 aint big enough on the Multi. I'd recconmend 10X Multi's & up on this board & wish a whole pile of Luck your 10 Multi will take a fast FSB with it.
For all you newcomers to Over Clocking, Use AUTO on the STRAP setting until you fully understand how Straps work. I'm still convinced my board dying for a 370FSB 1:1 @ stock voltages on a 266 Strap is total BS & just bad Engineering on DFI/China's part.
If my comment's here upset anyone in the Industry, I'm speaking the truth & if it hurts, that's your Problem & you need to deal with it. Punters like us have been paying for this cr4p over the last 9 months...... Enough is enough! :cool:

Best of luck to all you "Die-Hards" for RD600, they are good untill they roll over & die. Stock Voltages takes the proverbial to see it die, so having Died "3" times already, I'm afraid I just can't stay "Hard" anymore for this board! ;)

Quote:

---

Originally Posted by Kin Hell

100Mhz Difference in boards is taking the Pi$$ (unless the last one JFK'd your Ram) & just says VOLUMES (DFI need need to Read & Digest this cr4p....We as Paying, "-oh-my-god-lets-bend-over-real-hard" Punters, ....are!!)

---

I am not sure if that is the case..I guess it could be and I worried about that being the case(my crucial 10th anniv at 500 was real sweet overclock)..but I am at 390 fsb a lower NB core..so I am guessing it is board variation which really sucks

Also my ram is at 2.0v @800

at some point later I will retry 500 again..but I fed up with this crap...too many variable too many issues...I a m100% stable....and my quick benchies seem very close with ram at 400 4-4-4 vs 500 5-6-5..not true with my last mobo..so who knows

between me and another 2 friends we had 4 dead boards

mine was a stupid thing.....changed a strap....never booted again
on another one the display went off....the board never worked again


i am really disapointed with this mobo.......
slow perfomance compared with 965......very buggy
i am running an abit quad gt right now and finally i reached stability


when it returns from rma it will be useless

i guess new bios would solve the majority of the issues......but who gives a crap about us.....
:slapass:

I got to admit I've given the board some beating and haven't had a problem yet touch wood!

Scares me to read all these things though.

Quote:

---

Originally Posted by nealh

I am not sure if that is the case..I guess it could be and I worried about that being the case(my crucial 10th anniv at 500 was real sweet overclock)..but I am at 390 fsb a lower NB core..so I am guessing it is board variation which really sucks

Also my ram is at 2.0v @800

at some point later I will retry 500 again..but I fed up with this crap...too many variable too many issues...I a m100% stable....and my quick benchies seem very close with ram at 400 4-4-4 vs 500 5-6-5..not true with my last mobo..so who knows

---

Don't get me wrong nealh, there will always be differences between one board & another, but surely not not by 100Mhz...totally unacceptable tbh! Maybe you just need to throw a bit more Voltage at the Ram with this one!?
Happy that you had a fair turn-around & I hope your replacement is a better example than the last, tho' 100Mhz down on Ram is a BIG suck!

I just gotta decide which board to go for next......almost cba tbh! :rolleyes:
I'll see how long I can do without....Waiting & Returning RD600 kinda took me out of owning a "Pee-Cee" for nearly the last 9 months, so a little more time won't hurt & things might have matured for the better. :slapass:

Dunno if "fhpchris" will read this, but just in case; ;)
DFI 680i appears to be using Hooks again! :D

Well, I must have the only working board. But remember guys...drivestrengths are open to play with(stock should be 4-5 not 5-5)...so ram issues make much sense to me. My board has been running FAH 24/7 @ 1.65v NB 400FSBx9 sync, 1.55v CPU, on air. I really want to blame user error...anger found in postings show unlevel, non-persistent mental attitudes. This board is not made for this sort of person. It is a tweaker's board only.:lol2::fact:

Quote:

---

Originally Posted by cadaveca

Well, I must have the only working board. But remember guys...drivestrengths are open to play with(stock should be 4-5 not 5-5)...so ram issues make much sense to me. My board has been running FAH 24/7 @ 1.65v NB 400FSBx9 sync, 1.55v CPU, on air. I really want to blame user error...anger found in postings show unlevel, non-persistent mental attitudes. This board is not made for this sort of person. It is a tweaker's board only.:lol2::fact:

---

drive strengths... refresh me on the settings you are referring to please....

if you look above..one person changed a strap setting and that was the end of his mobo

My first board was a good board..I had some frustrations with it....but got stable 3500 mhz and 500 ram but I decided to tinker ..I lowered timings and I all did was go to 1:1 sync at 380(lowered ram timings and kept 2.2v) with same NB voltages rebooted and board failed to boot...reset and that was the end of my mobo....(mind you 390 1:1 sync booted once)...

that was very frustrating..my new mobo....will not run my ram at same timings and vdimm...it runs hotter on NB at lower voltage....I have tried a Microcool whisper xe and and removed my stock HSF, cleaned up crap and reapplied AS5 and use a 120mm fan blowing on it anf my temps are 6-7C higher at idle and load

so nothing irrational..board is good for tweakers but that all I need was tweak a bit and it died...nothing would bring it back from C1 error..not new ram, not any method of CMOS clearing

quality is not great IMHO..I am not extreme in anything I try....

ok. Let me be blunt. Changing strap could only at worst corrupt bios. If it was done via windows, totally user error. If not, parts weren't capable, and that's part of the risk you take when you overclock. BTW, my cpu's, all of them(and that's very many) operate @ 266fsb strap or less only. 333 causes no boot, 300 causes errors. I don'tt know why, nor do I care...this strap does not limit the cpu(had most @ 500FSB), merely the mobo.:fact:

I AM EXTREME!!!!lol. :lol2: I've tried changing EVERY SINGLE setting, had C1's, etc, and never failed to get the board alive again. But i have gamut of parts at my disposal, so things such as swapping out ram to get the board to boot is always an option for me. And we are not talking about jsut having different IC's...becuase sometimes it's simply SPD interpretation causing issues.

500mhz 1T prime stable? done that too...hard as hell tho. Took me about 3 days of 14-16 hrs a day to get it prime stable. 1:1...would never happen.

ATI wanted 1.5v default for NB, in order to pull 1:1 in operation, however this was quashed by Intel, if i got the story right. We got the story it will kill the chipset by eletron migration, but after having many different configs using this board, it seems plain to me that over 1.5v northbridge hurts cpu, and not the chipset. 2.0v, and your gonna cause serious issues. Something has to convert that chipset voltage to a votlage usable by the cpu...

I see alot of users here with complaints, but mostly this comes from people without many parts, nor time to reign in this board. Tony warned everyone way back when...would be good to heed his word! He's got nothing to gain from helping...

Yes, this chipset is hot. I have always stated this, defended it even. So take that into consideration...I did and seem to have no issues. Sure, this means I've got to deal with a loud system, but that's the price you pay.

266mhz, folks. That's stock. 400mhz, that's 150% of stock. I know this is extreme systems, but let's be a bit realistic here...not every part is gonna cooperate when you push things this far. 500mhz...187%. Almost a full double the stock speed. Go through the threads here and find how many parts(not just boards..i mean everything) can do that on traditional cooling. Be happy with the free performance you got. You could have less. No sweat off my back either way...My board works just like it should. Not too sure why, given some of the comments in this thread, but it does. But then you hardly hear from people without issues...:fact:

You need 1.65 for 400*9? Is that sync with ram though? Thought 1.5 got you up to 450. Maybe thats with Async though.

Edit: I have no problems except sometimes I boot up and get a screen as if the video card memory has gone berkerk and can't get into the new bios for toffee. Can anyone else who has a usb mouse confirm this? I think and USB input devices stop the new bios from allowing you into setup.

Quote:

---

Originally Posted by cadaveca

ok. Let me be blunt. Changing strap could only at worst corrupt bios. If it was done via windows, totally user error. If not, parts weren't capable, and that's part of the risk you take when you overclock. BTW, my cpu's, all of them(and that's very many) operate @ 266fsb strap or less only. 333 causes no boot, 300 causes errors. I don'tt know why, nor do I care...this strap does not limit the cpu(had most @ 500FSB), merely the mobo.:fact:

I AM EXTREME!!!!lol. :lol2: I've tried changing EVERY SINGLE setting, had C1's, etc, and never failed to get the board alive again. But i have gamut of parts at my disposal, so things such as swapping out ram to get the board to boot is always an option for me. And we are not talking about jsut having different IC's...becuase sometimes it's simply SPD interpretation causing issues.

500mhz 1T prime stable? done that too...hard as hell tho. Took me about 3 days of 14-16 hrs a day to get it prime stable. 1:1...would never happen.

ATI wanted 1.5v default for NB, in order to pull 1:1 in operation, however this was quashed by Intel, if i got the story right. We got the story it will kill the chipset by eletron migration, but after having many different configs using this board, it seems plain to me that over 1.5v northbridge hurts cpu, and not the chipset. 2.0v, and your gonna cause serious issues. Something has to convert that chipset voltage to a votlage usable by the cpu...

I see alot of users here with complaints, but mostly this comes from people without many parts, nor time to reign in this board. Tony warned everyone way back when...would be good to heed his word! He's got nothing to gain from helping...

Yes, this chipset is hot. I have always stated this, defended it even. So take that into consideration...I did and seem to have no issues. Sure, this means I've got to deal with a loud system, but that's the price you pay.

266mhz, folks. That's stock. 400mhz, that's 150% of stock. I know this is extreme systems, but let's be a bit realistic here...not every part is gonna cooperate when you push things this far. 500mhz...187%. Almost a full double the stock speed. Go through the threads here and find how many parts(not just boards..i mean everything) can do that on traditional cooling. Be happy with the free performance you got. You could have less. No sweat off my back either way...My board works just like it should. Not too sure why, given some of the comments in this thread, but it does. But then you hardly hear from people without issues...:fact:

---

Ok...I agree with what you say to a point

Im my case I got 390x9 stable, async 500mhz on my ram using 1.52v on cpu/1.42 v on NB

Kin Hell and I were talking about sync..for grins I gave it a try...I lowered ram timings from 500:5-6-5-5-18-24-42-2t sync to 400 4-4-4-4-18-24-42-2t sync mode still with 2.2v on ram, 1.52v on cpu, 1.42 v on NB..set 380 on fsb and sync mode 380....
strap was auto
I reboot..systems post goes to second screen and hangs at updating DMI pool. I hit reset button..posts gets to second screen and says somethingabout DMI pool. I reset to get to bios and change settings and the board is done, dead...

I get C1..ok no issue..I move ram to all slots, single sticks etc, I try everything in manual and suggested to reset cmos....nothing(BTW my 10th anniv crucial booted initial at bios default 1.8v with no issue)
So I go buy cheap stick of 667 DDR 1.8v..ram nothing

The board is absolutely dead...my only fault here IMHO is after the inital no boot on my rese I sould have gone right into bios ..reset to known working settings and booted again...never again will I allow a second reset with this board

I had no issues before with a second reset ...many times I would not get an initial boot into winxp..hit reset at the same second screen where just hung if it failed after the second, hit a third reset and it would post to get into bios..not this time

I found sometimes a second reset was need to get it load winxp on some OCing attempts

Quote:

---

ATI wanted 1.5v default for NB, in order to pull 1:1 in operation, however this was quashed by Intel, if i got the story right. We got the story it will kill the chipset by eletron migration, but after having many different configs using this board, it seems plain to me that over 1.5v northbridge hurts cpu, and not the chipset. 2.0v, and your gonna cause serious issues. Something has to convert that chipset voltage to a votlage usable by the cpu...

---

where did you get this info

BTW how high do you thinkchipset temps can be 24/7..you seem to have alot of knowledge on the board limitations

Quote:

---

Originally Posted by nealh

BTW how high do you thinkchipset temps can be 24/7

---

My max temp is approximately 55c fully loaded regardless of settings or voltages. That's using the stock heatsink with a fan sitting on top of it.

I had a HR-05 on the NB but removed because I didn't like the way it was attached. 55c max I can live with.

I've had this board since the day of release and have pound on it hard. So for, it keeps bouncing back and asking for more.

Quote:

---

Originally Posted by nealh

1.42 v on NB

---

This way your problem. As I have stated before, it took small voltage to get 450mhz, but much more to get 400mhz 1:1. Typically, crashing beofre CMOS has been checked means chipset volts are too low(or ram). I have tested my ram on otherboard so know it's max, and never reboot a failed overclock attempt such as you did. This could have been your mistake..could have been not enough voltage, but I have done the same as you, and not had any issues. However, the one difference I see right away is that my cheapy sticks are 512mb d9's with 400/533/677 SPD(ram is stable 533 3-3-3-9 @ 1.8v), and they always seem to boot this board @ 533mhz after a CMOS clear.

BTW...best way I have found to deal with a C1 is too do a power-reset CMOS clear. If after the clear you still have a C1, then you need to swap to other ram.

Quote:

---

Originally Posted by nealh

where did you get this info

BTW how high do you thinkchipset temps can be 24/7..you seem to have alot of knowledge on the board limitations

---

I read it somewhere, cannot remember where, but the person posting the info was not someone who's posts I take lightly. Unfortunately I cannot remember the source. I do however remember that this info does conicide with what I hav found with my use of the board.


My chipset was 62-63c on load with stock heatsink. Changing to AS5 only dropped temps a couple of degrees. Currently I have an Thermalright HR-05 on the chipset, and under 1.66v in bios i get 50c max loaded. If i had stock cpu voltage, this would be far too much, as I do believe there is a "buffer" where cpu voltage and chipset voltage must be in order for proper operation. If the values are too far apart, damage will occur that will not make itself known for a while. The same can be said for all the other voltages...they all need to be within a specific ratio of one anotheras the components on the board are only capable of shaping these voltages in certain ways. Same for every board tho.

[QUOTE=cadaveca;2080382]This way your problem. As I have stated before, it took small voltage to get 450mhz, but much more to get 400mhz 1:1. [q]

I agree after the fact I wished I pushed this to 1.48v with 390 1:1

Quote:

---

Typically, crashing beofre CMOS has been checked means chipset volts are too low(or ram). I have tested my ram on otherboard so know it's max, and never reboot a failed overclock attempt such as you did. This could have been your mistake..could have been not enough voltage, but I have done the same as you, and not had any issues. However, the one difference I see right away is that my cheapy sticks are 512mb d9's with 400/533/677 SPD(ram is stable 533 3-3-3-9 @ 1.8v), and they always seem to boot this board @ 533mhz after a CMOS clear.

---

I bought KVR 1 gb stick..1.8v DDR2...but it did not work as well...no matter ram slot

Quote:

---

BTW...best way I have found to deal with a C1 is too do a power-reset CMOS clear. If after the clear you still have a C1, then you need to swap to other ram.

---

I tried this as well as everything Tony recommended at DFI forum site, battery removal....I tried for 3 days various combinations of things that should have cleared cmos for a boot...somehow the bios got pooched
I wish I had kept the mobo and tried to get a new bios chip

my new board just will not run my ram at 500mhz async like the first one..2.2v 500mhz 5-6-5-5-18-24-42-2t, auto strap
this ran memtest 86+ 1.7 test 5 for 50 passes, winmemtest for 1000%, blend prime95 for 20 hrs

I error out after 7 passes on test 5 now

Quote:

---

I read it somewhere, cannot remember where, but the person posting the info was not someone who's posts I take lightly. Unfortunately I cannot remember the source. I do however remember that this info does conicide with what I hav found with my use of the board.
My chipset was 62-63c on load with stock heatsink. Changing to AS5 only dropped temps a couple of degrees.

---

My chipset sucks?? Microcool whisper xe gave idles of 59C..poor contact with shim??

I went back to same cooling on my old board ..1.42v idle was 47/48c and load was 53/54...stock HS, AS5 and 120mm fan blowing on it

Now same cooling at 1.36v..stock HS,AS5 and 120mm fan(lower ambients as well)..idle at 53/54...load was 58C(rare noted 60)

Quote:

---

Currently I have an Thermalright HR-05 on the chipset, and under 1.66v in bios i get 50c max loaded. If i had stock cpu voltage, this would be far too much, as I do believe there is a "buffer" where cpu voltage and chipset voltage must be in order for proper operation. If the values are too far apart, damage will occur that will not make itself known for a while. The same can be said for all the other voltages...they all need to be within a specific ratio of one anotheras the components on the board are only capable of shaping these voltages in certain ways. Same for every board tho.

---

now this is a very important point...cant recall what other board this was an issue...if you set one voltage too low and one too high the difference was a chip and board killer

my cpu vcore is at 1.52 and NB volt. at 1.36...I think this ok...not sure what difference is too great

I realyl want to blame your ram for most of your issues. This may not seem to make such sense, but seeing so many issue with Micron IC's(D9's), it's hard to not place blame there. Some of my Crucial kits won't boot any more @ default(400mhz 5-5-5-15 1.8v)...actually only one does. The others cause C1 issues. C1 i beleive is memory initialization, so either memcontroller or mem is at fault under these circumstances. However, you should be able to get a non-stable boot long before C1 occurs...so you adjusted too many options at once, IMHO, to get a C1.

I set my ram not based on speeds, but based on what timings give me a certain amt of bandwidth. This is key...as 3400mb/sec @ CAS4 or 3400mb/sec @ CAS5 are virtually the same thing to the memory IC...but at the same time, different for the memory controller. With this in mind, I have very tangible goals when tweaking, as well as limits.


In regards to your chipset, i cannot say it sucks without playing with it myself. Sometimes hotter chips scale farther, sometimes not. depends on whether the heat is leakage related, or current related. Leakage on hot chip generally means cooling will bring it in line, but if the current is making chip hot, then not much can help it. this is the same for pretty much all silicon...chips with high leakage can handle higher current. current is not Volts, BTW, but amperage. Pretty much like "b" conroe's...pretty sure the only difference is the IC's on bottom of chip pcb, that allow more current to the chip...easily explains why "b" chips tend to go farther...they have more current available. Also explains the heat of those chips too.

Quote:

---

Originally Posted by cadaveca

I realyl want to blame your ram for most of your issues. This may not seem to make such sense, but seeing so many issue with Micron IC's(D9's), it's hard to not place blame there. Some of my Crucial kits won't boot any more @ default(400mhz 5-5-5-15 1.8v)...actually only one does. The others cause C1 issues. C1 i beleive is memory initialization, so either memcontroller or mem is at fault under these circumstances. However, you should be able to get a non-stable boot long before C1 occurs...so you adjusted too many options at once, IMHO, to get a C1.

---

ok but why did the cheapo ram not allow a boot....Again when I made my changes that killed the board...async to sync, lower fsb to 380, and lowered ram speed and used known good timings with no change in vdimm

BTW..this same crucial booted right up on the new board at default bios settings(1.8v 667 5-3-3-3)

Quote:

---

I set my ram not based on speeds, but based on what timings give me a certain amt of bandwidth. This is key...as 3400mb/sec @ CAS4 or 3400mb/sec @ CAS5 are virtually the same thing to the memory IC...but at the same time, different for the memory controller. With this in mind, I have very tangible goals when tweaking, as well as limits.

---

what gets you 3400mb/sec at cas4? my cas4-4-4-4-18-24-42@400 settings per memtest 86+ 1.7gives a bandwidth of like 2700mb/sec
cas5-6-5-5-18-24-42 at 500 gives me about 3110 mb/sec

Quote:

---

In regards to your chipset, i cannot say it sucks without playing with it myself. Sometimes hotter chips scale farther, sometimes not. depends on whether the heat is leakage related, or current related. Leakage on hot chip generally means cooling will bring it in line, but if the current is making chip hot, then not much can help it. this is the same for pretty much all silicon...chips with high leakage can handle higher current. current is not Volts, BTW, but amperage. Pretty much like "b" conroe's...pretty sure the only difference is the IC's on bottom of chip pcb, that allow more current to the chip...easily explains why "b" chips tend to go farther...they have more current available. Also explains the heat of those chips too.

---

interesting info..

Blah..room temp is 76F/24.5C..my idle temps on my chipset now with 1.36v ...54/55C

i run 4-4-3-4-12-16-28-3-4-3-3-2-2-5, 401mhz, sync, for 3356mb via memtest.

Don't get me wrong, not blaming the issue on you, but like you say, it doesn't sound right. Could be the board...very possible...however if something corrupted BIOS as to make it not bootable, and it was before, only user error is the cause. I'm not saying you were stupid in your settings or anything, but that it was your actions that killed the board. All settings you may have changed may have been valid...but your fingers still pushed the buttons that made it die. Had everything stayed at stock ,I bet the board would still be alive.

And yes, i know this board is meant to be overclocked. But it's the overclocking that killed it. Bad chipset, bad B.O.M., bad luck, doesn't matter.

I mean, if i take your approach, then it MUST be YOUR fault, as my board works fine. Without both boards in my hands, there's really no comparison to be made. Now, if i got 4 brand new boards(not boards given as RMA, they are never new, merely returns, 99% of the time), and all of them died, 3 of them died, heck even if one of them died, I'd be a bit upset, but that's not the case here. We've got one guy returned his board and got back two duff boards. To me, this is no surprise. They didn't pull new boards off of the shelf...DFI doesn't even do that. I sent in RMA board direct to DFI, marked with invisible ink, got the same back. Didn't overclock worth a damn for me when i got it back, but whatever. Couldn't get 300FSB out of it(NF4). Dude who bought the board got 300mhz easy as pie. My fault? Did he know some trick i didn't? Nope. Just used different parts.

So, we got different parts, my board works, your doesn't. I say it's your other parts. Based on the C1 error on diagnostic LED, I'd blame your ram. Do i know this is the cause 100% for sure? Nope. 10%? Nope. But it was something that made it die...

Quote:

---

Originally Posted by cadaveca

i run 4-4-3-4-12-16-28-3-4-3-3-2-2-5, 401mhz, sync, for 3356mb via memtest.

---

nice..401 sync...I wonder if the 401 value is what helps the sync..I have seen alot of talk regarding how some fsb just dont work and others do

I tinkered a tiny bit last night...all async-2t with NB at 1.36v
tried 451 mhz on my Crucial 10th anniv..@2.1v 4-4-4-4-18-24-42-3-6-3-3-2-2-5 with cpu at 390 fsb(3500) I get 2974 mb/sec
changed cpu to 3465(lowered speed to see effect)...bandwith was essentially the same 2924 mb/sec
then lowered cpu to 3455...ram at 451 as above
bandwidth was 2999 mb/sec...passed 27 passes of test 5 on memtest 86+ 1.7

Quote:

---

Don't get me wrong, not blaming the issue on you, but like you say, it doesn't sound right. Could be the board...very possible...however if something corrupted BIOS as to make it not bootable, and it was before, only user error is the cause. I'm not saying you were stupid in your settings or anything, but that it was your actions that killed the board. All settings you may have changed may have been valid...but your fingers still pushed the buttons that made it die. Had everything stayed at stock ,I bet the board would still be alive.

And yes, i know this board is meant to be overclocked. But it's the overclocking that killed it. Bad chipset, bad B.O.M., bad luck, doesn't matter.

---

I cant disagree with this...I did make the changes and hit the buttons..so I agree user error in the end is the issue
But they sell the board knowing it will be overclocked and I payed a premium(my fault), I expected a bit more tolerance, esp since it did not seem as comapred to others experiences I was pushing too hard

Quote:

---

I mean, if i take your approach, then it MUST be YOUR fault, as my board works fine. Without both boards in my hands, there's really no comparison to be made. Now, if i got 4 brand new boards(not boards given as RMA, they are never new, merely returns, 99% of the time), and all of them died, 3 of them died, heck even if one of them died, I'd be a bit upset, but that's not the case here. We've got one guy returned his board and got back two duff boards. To me, this is no surprise. They didn't pull new boards off of the shelf...DFI doesn't even do that. I sent in RMA board direct to DFI, marked with invisible ink, got the same back. Didn't overclock worth a damn for me when i got it back, but whatever. Couldn't get 300FSB out of it(NF4). Dude who bought the board got 300mhz easy as pie. My fault? Did he know some trick i didn't? Nope. Just used different parts.

---

Again based on your comments, I have to agree with you...it was my fault but I still expect if the product is advertised to techies like us for overclocking...I hope it will have better tolerances

My analogy is a car...if you buy Honda your expectations will be different than if you buy a Corvette..you will pay differently as well

I am pretty sure from the process the RMA board was a new one....I got Newegg to cross-ship it....the packaging on the mobo anti-static bag...was untouched...The rep claimed they were shipping a new board

HS still had gumball thermal paste(some users would never remove it and reapply but many buying this board will)
can I be 100% sure it is a new board no, of course but based on box appearence and contents, antistatic bag...pretty sure it is brand new

Quote:

---

So, we got different parts, my board works, your doesn't. I say it's your other parts. Based on the C1 error on diagnostic LED, I'd blame your ram. Do i know this is the cause 100% for sure? Nope. 10%? Nope. But it was something that made it die...

---

I agree again...even with the same exact parts....we will have a different although close experience and your skills probably are better than mine
I always get a little less performance/overclock vs others...which is ok

Yeah nealh, your new board is working. It might need some spanking to get in line, but then it really sounds as if your ram is starting to die. I'm sure you've seen the other threads here on those sticks...your probably came form teh same batch...ordered when the "huzzah" was going around. I did too...those sticks are toast.


And yes, this board is a bit disappointing. But it's an old chipset hanging with the top guns, and not doing all that shabby, so i don't mind much. But 975x is much faster, in everything, it seems.

There are large holes is performance...if you check back to some of Tony's posts from when this board first came out, as well as eva2000's, you'll find that there are 3-4mhz holes where performance falls behind, due to the offset of the cpu bus and memory bus on the chipset. you must pick numbers that divide evenly, or performance suffers.

hmm, I have a problem. I'm currently at 426*9 1.61V on idle in windows. Now when my CPU gets on 100% load by running orthos, after some time CPU-Z reports multi-drop to 6 so the CPU runs @ 2556MHz, everest reports it too, but orthos and coretemp don't. It didn't show before, @ 3.7Ghz it is always at set speed, no drop. the C1E function is disabled in BIOS, is there anything else that could be causing this? thnx

Quote:

---

Originally Posted by cadaveca

Yeah nealh, your new board is working. It might need some spanking to get in line, but then it really sounds as if your ram is starting to die. I'm sure you've seen the other threads here on those sticks...your probably came form teh same batch...ordered when the "huzzah" was going around. I did too...those sticks are toast.

---

UGH....are you saying this becasue it will not run at 500mhz anymore..or did you note something else in my posts

I was impressed last night when I was able to do 451 4-4-4-4-18 at 2.1v, when I was only on my old board to 500 with 2.2v 5-6-5-5-18
I should note if I give the ram 2.23v=bios 2.25v reading(vs 2.2v bios setting on old=2.22v) ..I was able to run a few passes at 500 5-6-5-5-18 with 3115mb/sec with cpu at 3500...but I really want to keep the ram from max voltage as I saw so many sets die

my biggest concern is going to be chipset temps....Ambient temps are going up...my room last night 76F/24c...idle chipset temps are now 54/55..by summer in florida I will have 25/26C ambients

I am serious looking at WCing block but I really hate the idea of rd600 specific block and while I like a MCW30..I am not sure the clip menchanism will keep a good tight fit with my tube routing

I could not get my Microcool Whisper XE to fit correctly with shim in place and was concerned I would damge the core without it

Considering a Swiftech MCX 159CU..now

Quote:

---

And yes, this board is a bit disappointing. But it's an old chipset hanging with the top guns, and not doing all that shabby, so i don't mind much. But 975x is much faster, in everything, it seems.

---

I realize the chipset is older technology..though realy did not understand this until after buying...only 975x board I like was the Badaxe2....but reports did not seem to show consistent 400+ fsb(though for me in the end it does not matter as I cant run 3600) and the lack of lower multiplier

Quote:

---

There are large holes is performance...if you check back to some of Tony's posts from when this board first came out, as well as eva2000's, you'll find that there are 3-4mhz holes where performance falls behind, due to the offset of the cpu bus and memory bus on the chipset.

---

this seems to be a bug a boo for all intel boards....

Quote:

---

you must pick numbers that divide evenly, or performance suffers.

---

are you referring I assume to fsb and ram speed? do they need to divide evenly as a given number ie cpu speed 384 vs 385..or cpu vs memory speeds ..384 vs 451(cant do 450 in bios)

what i mean is that the fsb and ram speed should , hopefully, follow the default ram multi's, 2:3, 1:1, 4:5, it does not have to be exact...but i find that say 432 for ram, is divisible by 4. So to me, fsb should be divisible by 3, giving 3:4 split, almost. If it's not you get a slight latency hit because the ratio is whacked, and this drops bandwidth.


BadAxe 2 is ok, but it's slot arrangement doesn't suit my needs. I have one tho...and have no issues getting over 400FSB. Mind you I have no issues doing such on P5W(68MOAG, 450FSB no problem) or XBX1 too(over 50% using windows, 417FSB, rebootable. more is only prevented due to clockgen), and lots of others DO have issues reaching such speeds.

It's really weird, because I have bought many parts...spent about 15k already this year on Core2Duo cpu's, boards, and DDR2. Very few of the issues I see related on forums are ones that I have had. All my e6600's(23 now) reach 3.6ghz. all of my E4300's(15) cannot do 400FSB, but 333FSB for 3ghz, @ stock volts and cooler each can do without problems. So either I got EXTREMELY lucky on all the parts I have(which i HIGHLY doubt), or something else is amiss here. I got a couple of quads when they first came out...ran into same FSB issues most had...but when it comes to Core2Duo clocking, I've yet to get a really bad chip. Mind you I don't mind pumping 1.6v for 3.6ghz, if needed.

Can anybody get past 530FSB w/ this board?? I have been stuck there w/ my E6600. Another question, can anybody tell the difference between the 300 & 333
bootstrap? I did not see any differences when running Super Pi or PC Mark05....

ok, can anyone tell me how to disable EIST in this board's BIOS? I can't find a setting that would look like it does that, I ran through the whole BIOS a couple of times.

Quote:

---

Originally Posted by cadaveca

what i mean is that the fsb and ram speed should , hopefully, follow the default ram multi's, 2:3, 1:1, 4:5, it does not have to be exact...but i find that say 432 for ram, is divisible by 4. So to me, fsb should be divisible by 3, giving 3:4 split, almost. If it's not you get a slight latency hit because the ratio is whacked, and this drops bandwidth.


BadAxe 2 is ok, but it's slot arrangement doesn't suit my needs. I have one tho...and have no issues getting over 400FSB. Mind you I have no issues doing such on P5W(68MOAG, 450FSB no problem) or XBX1 too(over 50% using windows, 417FSB, rebootable. more is only prevented due to clockgen), and lots of others DO have issues reaching such speeds.

It's really weird, because I have bought many parts...spent about 15k already this year on Core2Duo cpu's, boards, and DDR2. Very few of the issues I see related on forums are ones that I have had. All my e6600's(23 now) reach 3.6ghz. all of my E4300's(15) cannot do 400FSB, but 333FSB for 3ghz, @ stock volts and cooler each can do without problems. So either I got EXTREMELY lucky on all the parts I have(which i HIGHLY doubt), or something else is amiss here. I got a couple of quads when they first came out...ran into same FSB issues most had...but when it comes to Core2Duo clocking, I've yet to get a really bad chip. Mind you I don't mind pumping 1.6v for 3.6ghz, if needed.

---

Thanks...nice info

Have you had any e6600 die from 1.6v...there seem to be sporadic reports of slow death/degradation...I not really happy using 1.52v on my e6600(temps are fine)

BTW, why do you think the ram is dying..not sure I understand why?

I will definitely watch this and send in for RMA if need be....problem is I returned my cheapo ram(it wasnot so cheap $85 for soemthing I would not use)

15k on parts...what do you do with all the stuff....

Quote:

---

Originally Posted by marauder16

ok, can anyone tell me how to disable EIST in this board's BIOS? I can't find a setting that would look like it does that, I ran through the whole BIOS a couple of times.

---

look at this screen in Genie Bios..scroll down to bold
CPU Vid Control: 1.4625v
CPU VID Special ADD: Auto
DDR2 Ram 1.8 Voltage: 2.20v
CPU VTT 1.2V Voltage: 1.23v
NB Core 1.2V Voltage: 1.42v
NB PLL 1.8V Voltage: 1.93v
NB PLL 1.2V Voltage: 1.21v
NB PCI-E 1.2V Voltage: 1.22v
NB Core 1.2V convert from: 1.97v
SB Core 1.2V Voltage: 1.23v
Clockgen Voltage: 2.92v
GTL REF Voltage control: Enabled
CPU Core 1/2 GTL REF Volt: 48
CPU Core 3/4 GTL REF Volt: 48[
North Bridge GTL REF Volt: 128
CPU PLL Setting Overide: Enabled
PLL Override loop count: AUTO
Conroe P-State Multiplier: x9
C1E Function: Disabled

Have you tried to disable the C1E function?

it is disabled, but the EIST isn't, that guy from CPU-Z told be that after looking at the registry dump of mine.
don't know where else to look..

EDIT: for some reason, that line in the bios is set to disable and can't be changed, anyone know why is that?

Quote:

---

Originally Posted by nealh

Thanks...nice info

Have you had any e6600 die from 1.6v...there seem to be sporadic reports of slow death/degradation...I not really happy using 1.52v on my e6600(temps are fine)

BTW, why do you think the ram is dying..not sure I understand why?

I will definitely watch this and send in for RMA if need be....problem is I returned my cheapo ram(it wasnot so cheap $85 for soemthing I would not use)

15k on parts...what do you do with all the stuff....

---

LoL high-end rigs topping 4k is not uncommon...15k is not much in the scheme of things. Filling an office with new machines took care of most of the hardware. Friends that are gamers but not into overclocking or anything like that took care of the rest. I have 3 Conroe rigs of my own...and about to sell them off to buy new ones.

in Regard to 1.6v killing a cpu, no, not happened yet, but I HAVE seen cpu's loose thier legs and not go further than 3.6ghz after attemping for more @ 1.6v. Each of these needed 1.5v for 3.6v, and would do 3750 @ 1.6v, but after extended burn-in(aircooling, 70c plus temps) they no longer could do 3700, even.


I think ram is dying because BOM of these sticks has changed over time, and you have issues that seem mostly ram-related. You only revise hardware if there is an issue.:fact: I have had many sets of this IC and pcb combo die, although i really think it's a small SMD on the pcb, and not the pcb itself that is the problematic part.

It's hard to say yes, it's dying, or no, it's not, without having them in the same board that they first started in. Very easy to see if they are still capable this way, and this is why I have many rigs. One to test ram, one to test cpu, one to use. Friends will come over to game, so it's all worthwhile for me.

LoL. best part is that you are under water, and concerned about 1.5v...when I'm talking air! Anyway, it's hard to find conclusive evidence of problems, or lack thereof, without proper testing. It's impossible with just one set of hardware, and the more you add to the mix, the easier it is to fnd correlations between everything. this is why i have very many issues with many of the comments in this thread, as they do not look at the whole picture, and merely focus on a small part, which, while useful, is not accurate.

Quote:

---

Originally Posted by cadaveca

LoL high-end rigs topping 4k is not uncommon...15k is not much in the scheme of things. Filling an office with new machines took care of most of the hardware. Friends that are gamers but not into overclocking or anything like that took care of the rest. I have 3 Conroe rigs of my own...and about to sell them off to buy new ones.

in Regard to 1.6v killing a cpu, no, not happened yet, but I HAVE seen cpu's loose thier legs and not go further than 3.6ghz after attemping for more @ 1.6v. Each of these needed 1.5v for 3.6v, and would do 3750 @ 1.6v, but after extended burn-in(aircooling, 70c plus temps) they no longer could do 3700, even.


I think ram is dying because BOM of these sticks has changed over time, and you have issues that seem mostly ram-related. You only revise hardware if there is an issue.:fact: I have had many sets of this IC and pcb combo die, although i really think it's a small SMD on the pcb, and not the pcb itself that is the problematic part.

It's hard to say yes, it's dying, or no, it's not, without having them in the same board that they first started in. Very easy to see if they are still capable this way, and this is why I have many rigs. One to test ram, one to test cpu, one to use. Friends will come over to game, so it's all worthwhile for me.

LoL. best part is that you are under water, and concerned about 1.5v...when I'm talking air! Anyway, it's hard to find conclusive evidence of problems, or lack thereof, without proper testing. It's impossible with just one set of hardware, and the more you add to the mix, the easier it is to fnd correlations between everything. this is why i have very many issues with many of the comments in this thread, as they do not look at the whole picture, and merely focus on a small part, which, while useful, is not accurate.

---

I agree today it is not hard to hit thousands for a high end rig....

I agree with your comments and wish I could afford the time and wife factor to have several test rigs..but this would never be justifiable expense to her...

Thanks for the interesting insights and observations...

BTW, excuse my ignorance what is BOM

BOM is bill of materials....the parts that go into making a card, whether mobo, cpu, mem, vga...the list of SMD's and such that make up the component. Usually in revisions of a product, this changes. If it's a pcb change, then typically you get a whole new product name. This is why ASUS, gigabyte, etc have many boards that are the same, but different name, slightly different functionality. What differs between the products is the BOM(like Striker and P5N32-E, but not). this means not the main parts(at least to me), but hte supporting parts. Not memory IC's, nor gpu, but all the other parts, capacitors and such.

FYI:
new bios
http://www.dfi.com.tw/Support/Downlo...FLAG=A&SITE=US

at long last.. a bios.. :)

will try this later.. :)

Quote:

---

Originally Posted by nealh

FYI:
new bios
http://www.dfi.com.tw/Support/Downlo...FLAG=A&SITE=US

---

Small hint: Do a byte to byte compare ...

:(

I think the same but readme not the same. I've not taste the beta bios so i can't say if he rise vddr to 1.95V so maybe based on beta bios but with little (or big) changes.

Quote:

---

Originally Posted by Pete@X

Small hint: Do a byte to byte compare ...

:(

---

are you suggesting they posted an official bios that does not do what they list in the readme...:eek:

the support on this board has been blah so far...so no surprise to me if this is the case

Hmm not getting this at all, can't seem to get stable at 3.6 at all.

Processor is L628B Which I thought was a good stepping, so I think it maybe the board.

Having to run in yellow dimms being as ornage refuse to run dual channel and the board also makes a whine noise when under load. Done all this from the very start. Maybe time to RMA.

Edit: I can get 3.2 with no voltage increase, after that then it seems voltage has to be used.

Quote:

---

Originally Posted by Mekrel

Hmm not getting this at all, can't seem to get stable at 3.6 at all.

Processor is L628B Which I thought was a good stepping, so I think it maybe the board.

Having to run in yellow dimms being as ornage refuse to run dual channel and the board also makes a whine noise when under load. Done all this from the very start. Maybe time to RMA.

Edit: I can get 3.2 with no voltage increase, after that then it seems voltage has to be used.

---

not all e6600 of any stepping will hit 3600..and I find going from 3420(1.4625) to 3507 needs 1.52v..about .06...big jump

you will need to boost vcore and as you increase the chip may need a fair bit more vcore to hit that speed

also as you increase fsb the NB will need more volts

I would suggest setting vcore to at least 1.5v and see what you get ..but increase NB to at 1.42v to help reduce the chance the chipset is a factor
Genie Bios

>Performance options

Memory Clock Frequency: DDR-533
Control CPU Frequency: Manual
x CPU Frequency (Mhz): 390
NB FSB Strapping: auto
PCI-E Overclock: 100
Software Reset Clockgen: Enabled
Memory Overclocking: Enabled
Overclock Frequency(MHz): 400

>NorthBridge ASIC CFG
>NorthBridge Selections
Memory Channels: channel A & B
Memory Power Mangement: Disabled
IGD Clock Gate: Disable
NB Channel Mapping: Auto
Simple Addr Translation: Disable
Chip-Select Swap: Disable
Burst Length: 8
Bank0 Map GFX: 4
Bank1 Map GFX: 5
Bank2 Map GFX: 6
Channel Mapping NB: 5
Channel Mapping GFX: 1
Bank0 Map NB: 5
Bank1 Map NB: 11
Bank2 Map NB: 7

>Memory Timing Menu

CS Output Timing: Delay 1/4 clock
CK Output Timing: AUTO
Dual channel Split Mode: AUTO
ADDR/CMD Driving Pre-AMP: Enabled ..now switched after p95v251 diasbled
ADDR/CMD Driving Strength: 05
Clocking Driving Pre-AMP: Always
Clocking Driving Strength: 05
Address/Command Timing: 2T
Memory Timing Select: Manual
CAS Timing (TCAS): 4
Read RAS-CAS Delay (TRCDR): 4
Write RAS-CAS Delay (TRCDW): 4
RAS Precharge Time (TRP): 4
Cycle Time (TRAS): 18
Row Cycle Time (TRC): 24
Refresh to Active (TRFC): 42
Bank to Bank CMD (TRRD): 3
Write Recovery (TWR): 6
Read-Write Command (TRTW): 3
Write-Write CS (TWTW_CS): 2
Write-Read Command (TWTR): 3
Write-Read CS (TWTR_CS): 2
Read-Read CS (TRTR_CS): 2
Read Data Transfers: 5

Memory Auto-Calibration
Read Data Transfer: Enable
Read Strobe Delay: Disable
Read Strobe Arrival: Enable

>FSB IO Strength

FSB IO Driving COntrol: Default
x 1X NMOS Strength
x 1X PMOS Strength
x 2X NMOS Strength
x 2X PMOS Strength
x 4X NMOS Strength
x 4X PMOS Strength



CPU Vid Control: 1.5v
CPU VID Special ADD: Auto
DDR2 Ram 1.8 Voltage: 2.20v
CPU VTT 1.2V Voltage: 1.23v
NB Core 1.2V Voltage: 1.42v
NB PLL 1.8V Voltage: 2.03v
NB PLL 1.2V Voltage: 1.21v
NB PCI-E 1.2V Voltage: 1.22v
NB Core 1.2V convert from: 2.14v
SB Core 1.2V Voltage: 1.34v
Clockgen Voltage: 3.40v
GTL REF Voltage control: Enabled
CPU Core 1/2 GTL REF Volt: 48
CPU Core 3/4 GTL REF Volt: 48
North Bridge GTL REF Volt: 128
CPU PLL Setting Overide: Enabled
PLL Override loop count: AUTO
Conroe P-State Multiplier: x9
C1E Function: Disabled


the bolded values will help alot in most cases...give extra voltages at first and then you can tweak them

The new bios isn't important uh? There 're only little changes.

Is it possible that the heat from the NB Heatsink could damage the PCIe slot?
The screen shows all blues as greens now when I browse the net or click on a icon since my first benching session. I know it sounds crazy, but it worked fine up until that point. I have tried the card in other computers and it works fine not showing this issue....

Bios showed 75~85 degrees celsius(Depending on how accurate it is) while I was benching. It was a 3 hour period of this. I had two 120mm fans running on it and it felt cool to the touch. I assumed it was a reaction to the board under stress and that it would go back to normal under stock conditions, but the next day it was still showing this issue. Any thoughts?:confused:

three words? r m a

I still cant figure out why the board will not go past 450FSB. Played with just about all the settings I can. The CPU goes to 7x450 with Vcore on AUTO which is pretty darn good anyways. But I figured maybe it just needed more voltage for higher mghz/FSB but I can give the CPU 1.55V and still will not POST correctly after 450FSB. Is it possible this is just the limit of the CPU and it will not go any higher, or is it just the board refusing to go higher? I also tried raising the NBcore up to 1.8 or 1.9V but not for long as this is way too high I just did it to see if it would pass 450FSB and it still refuses to. I am curious to see if this CPU will go higher, i mean it it does 3.15gig at stock voltage so why wouldn't it be able to do another 100mghz?

It should not be the board limiting the FSB. You have to give more voltage to the NB.

Link


I was having the same issues. Once I raised the the NB voltage I was able to max the board out. The highest I could get the board was 530FSB. Just make sure you have proper cooling on the chip set.

[IMG]http://img253.imageshack.us/img253/9...tledfn0.th.png[/IMG]

Quote:

---

Originally Posted by White B

Just make sure you have proper cooling on the chip set.

---

Above 1.53V on the nb aftermarket cooling is a necessity. And the chip isn't going to last with high voltage 24/7.

yeah i added a fan to the NB heatsink max load temps are about 55C. and idle is about 50C. That is NB Core on 1.52V I tried maxing out the NB voltage but it didn't seem to help. i also messed around with all the voltages concerning the NB but ti still refuses to POST correctly past 450FSB. I dont want to go above 1.5V for long periods because from what i have read it could damage the the Motherboard or CPU.

Oh yeah I tried the 2/12 BIOS and it wont boot from a USB keyboard with the new BIOS it just goes to a black screen after the logo displays and stays there. I plug in a PS2 keyboard and it works fine.

anybody know how changing the GTL ref voltage on cpu affect cpu life

I saw a post that seem values of 70-80 are helpful per Tony aka Big Toe(supposedly better than 48/48)
I changed mine to 70/70/128 and was able to lower cpu vtt from 1.23 to 1.21...according to Intel papers cpu vtt of 1.26 is max and on my mobo 1.23 bios=bios reading of 1.26

also Iwas able to load winxp at 3600 at same vcore as 3500 and this usually would just hang on screen after the initial post

Quote:

---

Originally Posted by nealh

anybody know how changing the GTL ref voltage on cpu affect cpu life

I saw a post that seem values of 70-80 are helpful per Tony aka Big Toe(supposedly better than 48/48)
I changed mine to 70/70/128 and was able to lower cpu vtt from 1.23 to 1.21...according to Intel papers cpu vtt of 1.26 is max and on my mobo 1.23 bios=bios reading of 1.26

also Iwas able to load winxp at 3600 at same vcore as 3500 and this usually would just hang on screen after the initial post

---

nice find.. :)

here is the link
http://www.thetechrepository.com/showthread.php?t=76

look at last post

what I am not sure is if you should increase the NB GTL?

not sure how this affects the cpu/mobo life

Quote:

---

Originally Posted by Tony

To start with you have to Enable GTLRef to have the setting take some effect, Oskar has preprogrammed good auto settings but they can be improved on as each processor does respond in different ways to voltage tweaks.

In my original guide Oskar had informed me to tweak up from 0 to 48 for the CPU GTL voltages, we have now found this is not the best option. Thru the work of many end users using dual and quad core CPU's we have found increasing over 48 is best to increase the FSB the cpu will run on. Also please rememer DFI have locked core 0+1 and 2+3 together for adjustment, they did this as this is how they are latched onto the FSB. Here are some points to remember.

1 If using a quad core you should only need to increase GTL for core 2+3, signalling to 0+1 is usually OK at auto unless you are seeing a huge FSB hole.

2 Increase GTL values BEFORE CPU VTT, this is VERY important, only increase VTT if GTL has zero effect or makes things worse.

3 Increase GTL voltages in preference to NB Vcore voltage. While an increase in NB core voltage does have a huge effect on FSB overclocking you do need to keep the NB core voltage as low as you can to ensure the NB has a long life.

Now the final question....What is the optimal settings for the highest FSB?

The answer is every CPU is different, some have found setting between 70 and 80 best on the CPU, some have even run 120+ so it all depends on what the CPU cores need to open up the FSB. be prepared for a lot of tweaking and testing...but as Kris has shown it is possible to gain quite a few MHz if you are willing to put in some time testing.

---

I have increased the GTL 80 80 70 and it's runs fine (Orthos 7h stable 1h less to finish). Vcore is 1,42.

Quote:

---

Originally Posted by GNU

I have increased the GTL 80 80 70 and it's runs fine (Orthos 7h stable 1h less to finish). Vcore is 1,42.

---

so you lowered the NB to 70???

did you lower other mobo voltages?

the gtl values were taken from intel specs using the voltage scale used by oskar in the bios, shrek posted a link to this at dfi club, having used the nearest to spec on dual core and found it worked better. Quad core needs higher gtl ref voltages. In terms of component life using GTL I would not concern myself excesivley, you are far more likely to hurt compnents with vnb and vcore and vdimm than gtl ref . Remember to stick within 75-85 on most dual core cpu's and you should be ok for higher clocks booting up stable.


Pcie is a voltage we routinely abuse on most boards, on this board near staock works best once gtl is tuned..

regards
Raja

Quote:

---

Originally Posted by raju

the gtl values were taken from intel specs using the voltage scale used by oskar in the bios, shrek posted a link to this at dfi club, having used the nearest to spec on dual core and found it worked better.

---

Do you have this link..I missed it at dfi

Quote:

---

Quad core needs higher gtl ref voltages. In terms of component life using GTL I would not concern myself excesivley, you are far more likely to hurt compnents with vnb and vcore and vdimm than gtl ref . Remember to stick within 75-85 on most dual core cpu's and you should be ok for higher clocks booting up stable.

---

Great info

Quote:

---

Pcie is a voltage we routinely abuse on most boards, on this board near staock works best once gtl is tuned..

---

I am finding at 1.36v on NB...if I want 450 ram speed I need to increase from default my NB PLL 1.2v to 1.29 and PCI-e 1.2v voltage to 1.34

Still not sure I am stable but closer

HMM..I may try 85....

Quote:

---

Originally Posted by nealh

so you lowered the NB to 70???

did you lower other mobo voltages?

---

No, other mobo voltages are setting up.

I got another problems. Maybe anyone here could help me to solve it.

http://csd.dficlub.org/forum/showthread.php?t=1952

It's the speed problem with my HDD and I haven'T found a solution.

Quote:

---

Originally Posted by GNU

No, other mobo voltages are setting up.

I got another problems. Maybe anyone here could help me to solve it.

http://csd.dficlub.org/forum/showthread.php?t=1952

It's the speed problem with my HDD and I haven'T found a solution.

---

I d/led HD Tach
Seagate 320 gb ..SATA 1
read 66.3 mb/sec
burst 251 mb/sec
13.5 ms access time

seagate 300gb... SATA 2
read 58.3 mb/sec
burst 133.2 mb/sec
access time 15.6 ms

on ati controller ..non raid...I do have an SATA ODD on sata 3

OK..I reread Kris artice at tech repsoitory...it seems the disabled default values are more in line with intel spec of 2/3 vtt for cpu ref voltage than when you first enable it in bios(48/48/128)

GTL ref voltage should be closer to 75-80 for the cpu..I have no idea on the NB

A little follow-up:

I found for me to run my ram at 451mhz at fsb of 390...I need to increase the bold

running dual P95 v 51 at 8 hrs plus and going
3500...3 x 390
ram is 4-4-4-4-18-24-42-3-6-3-23-2-2-5 2t at 2.05v(52 passes on test 5 memtest 86+ 1.7 with no errors, ran 3D mark05 ok)

CPU Vid Control: 1.51875v
CPU VID Special ADD: Auto
DDR2 Ram 1.8 Voltage: 2.05v
CPU VTT 1.2V Voltage: 1.23v from 1.21
NB Core 1.2V Voltage: 1.36v
NB PLL 1.8V Voltage: 1.93v
NB PLL 1.2V Voltage: 1.38v from 1.22
NB PCI-E 1.2V Voltage: 1.34v from 1.21
NB Core 1.2V convert from: 1.97v
SB Core 1.2V Voltage: 1.23v
Clockgen Voltage: 2.92v
GTL REF Voltage control: disabled this keeps the GTL more in line with Intel specs
CPU Core 1/2 GTL REF Volt: 48
CPU Core 3/4 GTL REF Volt: 48
North Bridge GTL REF Volt: 128
CPU PLL Setting Overide: Enabled
PLL Override loop count: AUTO
Conroe P-State Multiplier: x9
C1E Function: Disabled

I disabled GTL ref voltage as it supposedly is more in line with Intel spec..48/48 is on low side.... 75 or so in spec

If the goal if to keep all in line with Intel specs what it's the point of overclocking? If u think Intel specs must be folow stickly please minus ur vore from 1.51875 to 1.35 if u don't ur cpu to be blows in the second:rolleyes:

Quote:

---

Originally Posted by AbelJemka

If the goal if to keep all in line with Intel specs what it's the point of overclocking? If u think Intel specs must be folow stickly please minus ur vore from 1.51875 to 1.35 if u don't ur cpu to be blows in the second:rolleyes:

---

Did you look at any of the links I posted above....Jeez...I am not advocating using less....or following just intel specs..crap you always are looking to crap on posts

Sorry..you missed my point 48/48..applies less GTL ref voltage..which is what many are using

70-80 values actual are closer to Intel spec...therefore if you want more voltage you need to be higher than 48..which Tony aka Big Toe and freecableguy note in the links above

here just look at this link and you will hopefull see my point
http://www.thetechrepository.com/att...8&d=1170855299

Tony info
http://www.thetechrepository.com/showthread.php?t=76

Kris's article on GTL reference voltages
http://www.thetechrepository.com/showthread.php?t=87

Quote:

---

Originally Posted by nealh

I am serious looking at WCing block but I really hate the idea of rd600 specific block and while I like a MCW30..I am not sure the clip menchanism will keep a good tight fit with my tube routing

---

I got a redundant DD RD600 block if you want it m8y!?
It's as much use as a chocolate Fire-Guard for me now!
PM me if you want it...Pay Pal is cool this end.
Swap it for a DD 680i if you can get one.....None in the UK atm. :(

I have rd600 with gskill memory 2*1=2gb. But it doesn't works with it.I try to use new bios in dfi's web site but it doesn't works too. Before ı buy to rd600 ı dıdn't know gskill memory didn't works with rd600. So what can ı do ? to use gskill memory with rd600.Anyone else has Gskill memory and use rd600 without problem ?? :confused:

Look at Memory Subforum, hope this helps.

I haves and it works perfectly. 2*512 GBHZ 6400
http://valid.x86-secret.com/show_oc?id=170799

Quote:

---

Originally Posted by eXo_Ultimate

I have rd600 with gskill memory 2*1=2gb. But it doesn't works with it.I try to use new bios in dfi's web site but it doesn't works too. Before ı buy to rd600 ı dıdn't know gskill memory didn't works with rd600. So what can ı do ? to use gskill memory with rd600.Anyone else has Gskill memory and use rd600 without problem ?? :confused:

---

Have you tried using just 1 Stick of Ram in Slot 3? Try both of them even if the first stick fails. Reset CMOS & try them both again. If you get the board to Boot this way ...Send it back. It could be faulty.

My First board was DOA, the Second was "Picky" about using my Geil 4300 & I had to do the Ram Dance thing. Rumour had it it was the SPD settings on my Ram Modules! .....Rubbish...as Board number 3 never had to have the Ram Dance on Slot 3 to get it to work...so it WASN't the SPD timings on my Geil Ram. Board number 3 finally rolled over & died a couple of weeks ago.

*edit* btw....the GSkill RAM conforms to JEDEC Standards....so technically, it should Work.

hey all..

have you use a 1333 cpu on the RD600??

can it handle 1333 cpu's???

Yes someone had reached it.

Quote:

---

Originally Posted by erwinz

hey all..

have you use a 1333 cpu on the RD600??

can it handle 1333 cpu's???

---

yeah it can run 1333's afaik...

which is the best chipset block for this board guys? MCW30 or the DD one?

stock cooling, using AS5, gives me 55c loaded @ 1.3v, 389fsb.
DD Maze4 gives me 40c loaded.

I am at a loss on this mobo..my RMA board(sure looked new)...
is 26hrs stable p95..however all of sudden it will no longer run 3dMark05 at settings that were fine 1 week ago....I have increased voltages to no avail..

I wil lreboot on cpu test2...

I have unloaded AV/spyware...reinstalled version 1.2/1.3....

I do not want to rma this again...it took 10 days with Newgg and $18 shipping...damn it

Quote:

---

Originally Posted by cadaveca

stock cooling, using AS5, gives me 55c loaded @ 1.3v, 389fsb.
DD Maze4 gives me 40c loaded.

---

I am not sure what rad you have in your loop..but did you find overall temps go up after adding the NB into the loop? if so how much

BTW do you have any idea what a safe max NB temp is for 24/7...mine is loading at 60/62C...idles at 55C..ambient temps are 24C

Quote:

---

Originally Posted by nealh

I am at a loss on this mobo..my RMA board(sure looked new)...
is 26hrs stable p95..however all of sudden it will no longer run 3dMark05 at settings that were fine 1 week ago....I have increased voltages to no avail..

I wil lreboot on cpu test2...

I have unloaded AV/spyware...reinstalled version 1.2/1.3....

I do not want to rma this again...it took 10 days with Newgg and $18 shipping...damn it

---

Test #2 is almost always cpu instability. Back the processor speed down some and I bet it will pass.

Quote:

---

Originally Posted by Praz

Test #2 is almost always cpu instability. Back the processor speed down some and I bet it will pass.

---

I agree 100%
The settings are used were tested 26hr of P95 and ran 3dmark03/05/06 as well memtest 86 1.70+ before I claimed stability..now 1week later these same settings will not work

I lowered cpu, increased NB and assoc voltages...with no real change

The same cpu speeds were P95 stable on my old now rmaed rd600 as well...

now 3 days later it is failing...so as I see it..either 1.52v degrade my cpu at 3500...or something else is causing the issue..memory, mobo etc

my cpu was at 1.52v FWIW for about 4-6 weeks not running 24/7 under watercooling....temps have been fine

Quote:

---

Originally Posted by nealh

I am not sure what rad you have in your loop..but did you find overall temps go up after adding the NB into the loop? if so how much

BTW do you have any idea what a safe max NB temp is for 24/7...mine is loading at 60/62C...idles at 55C..ambient temps are 24C

---

I don't like over 50c myself. No idea if it added much if anything. lop is cpu-nb, gpu, gpu, pump, rad(BIX3).

Cpu had very little heatdump into loop...gpu's are the big heatdumps, so anything before them is not goingto make much of a difference.

Quote:

---

Originally Posted by cadaveca

I don't like over 50c myself. No idea if it added much if anything. lop is cpu-nb, gpu, gpu, pump, rad(BIX3).

Cpu had very little heatdump into loop...gpu's are the big heatdumps, so anything before them is not goingto make much of a difference.

---

thanks

Just keep in mind that this chipset is 90nm, so you can look back @ 90nm cpu processes @ the same voltage to get a good idea of where you should be. Just so happens that the chipset is about the size of a venice cpu...so I use that as a general comparison. This is also what made me cry foul on the recommendation that chipset voltage stay below 1.5v. They'd have to be some pretty shoddy silicon then, and from what I have seen, I do not beleive that to be true. Mind you, in the hands of TSMC, anything is possible, I suppose.

Quote:

---

Originally Posted by cadaveca

Just keep in mind that this chipset is 90nm, so you can look back @ 90nm cpu processes @ the same voltage to get a good idea of where you should be. Just so happens that the chipset is about the size of a venice cpu...so I use that as a general comparison. This is also what made me cry foul on the recommendation that chipset voltage stay below 1.5v. They'd have to be some pretty shoddy silicon then, and from what I have seen, I do not beleive that to be true. Mind you, in the hands of TSMC, anything is possible, I suppose.

---

hmm..very interesting information

I just can understand this sudden lose of stability on this board..really concerns me it is not poor sample

This cpu was P95 stable on both boards
at 3420@1.4625 with multiple runs of 3Dmark05
at 3500@1.52v with multiple runs of 3D Mark05...

then 3 days later neither speeds at those voltage can run 3Dmark05....I do not buy it...

Its not like I did not run P95 for more than 20hrs 4-5x, I ran 3Dmark05..at dozen times on each board....

I ran a Adware Pro..no issues..need to rerun my AV(NOD32)

very strange

i just bought mine today, and i am having a issue which so far i didn't understand where it comes from.

I get an FF error code on the POST, and i can hear clearly the board making its usual small beep that concludes POST. I can't get any image on screen, though i suspect the board finished its boot.

my specs are:
dfi icfx3200
g.skill pc6400 PK
xfx 7900GT xxx
lc power 550W 34a @ +12v

all other components are disconnected so far.

what does this error code relates to ?

normal operation after boot.

hmmm, yes the 'FF' code should be normal, but if you say you don't get any image on the screen then maybe it's sth with your VGA, did you check the connections and everything? do you get anything on the screen at any time, or does it just stay blank? when do you get that code, I mean are there other codes shown before it (they should be)?

EDIT: I ran to a problem. I needed to use my old 40GB IDE drive to save a backup of sth from my system. the problem is, any way I set the bios (or don't touch anything), any way I plug it in (as single, or as slave with the DVD-rom) it just won't let me boot into windows. It comes to the boot screen and then it loads for a sec or two and then it BSOD's, I tried everything, does anyone have any idea how to make this work? I need to reinstall my windows so I need to backup my data somewhere, please help.

Regarding the new bios, 2/12/07, is it now possible to select the boot order of sata/raid devices on the onboard ati sb raid adapter?

On the original 12/22/06 bios, when using the ati raid and booting a 2x74gb raptor raid0 array the only way for my plextor px716sa to work is to keep the drive unplugged, login to windows, then plug in the sata cable into the drive. Otherwise the adapter does not see my drive correctly in bios and allow me to use it.

Quote:

---

Originally Posted by nealh

I am at a loss on this mobo..my RMA board(sure looked new)...
is 26hrs stable p95..however all of sudden it will no longer run 3dMark05 at settings that were fine 1 week ago....I have increased voltages to no avail..

I wil lreboot on cpu test2...

I have unloaded AV/spyware...reinstalled version 1.2/1.3....

I do not want to rma this again...it took 10 days with Newgg and $18 shipping...damn it

---

You have the same problems like me. My system was a longer time stable but the last days, it gets blackscreen if I playing some games or cpustress applications. Very intersting is: I can playing some hours (4-5hrs) with all games and after 4-5hrs I got the first sights of problems (TFT blinks, freeze and blackscreen) after a short time shutdown (4min) I can do something for some hours. Funny not?

Mine used to do that, when the chipset was getting to hot. Cooled the chipset a bit and no problems after.

Yeah, sounds like an overheating issue. Add some fans :)

Quote:

---

Originally Posted by Yoxxy

Mine used to do that, when the chipset was getting to hot. Cooled the chipset a bit and no problems after.

---

Nope..chipset hits about 60/62 underload and will run P95 blend with chipset between 60-62C for 26hrs

My current cooling is stock HS with 120mm fan on my ram and chipset..HS get warm at best not really hot


This is not temp related on the chipset dont buy it for 1 sec....I was running at 1.36v and under or at 60c with P95 for 20 hrs plus on P95 blend, at 3500:450 and then 2 days later no longer stable

cpu temps at load are about 53-55C..my ambients are 23-24C

I my ram never received more than 2.2v in bios....but I guess it could be the culprit

i don't think so. If chipset is over 60 it started to make the screen flash. I would say keep ti under 55 and you will be fine.

I literally had the exact same problem. Once I got the chipset temperatures under 55 it never did it again.

I have a watercooler and I get every times flashed screen, freeze etc.. That's not fair.