Does any of the BIOSes work right with LDT/"HT Frequency" below 3x?
Printable View
Does any of the BIOSes work right with LDT/"HT Frequency" below 3x?
Seem this board doesnt like my BH-5. I cant overclock it at 1:1 divider at all.
Got my board fixed yesterday...
I hope they use 10nf smd resistors on their boards... Certainly looks the same...
C66 & C65 were the ones missing/knocked off... Boots up fine. Just getting to the part where I rebuild my rig again...
Dammit!!! Can't seem to break the 30k barrier in 01... 29,880 is the best I can do... Haven't turned the volts up yet...
neokenzo,
Mosy Neo2's don't OC well with BH5 @ 1:1 without more than 3v... You should know that...
The best I can do with no OCZ Booster & BH5 is only 217MHz (XMS3500) @ 2-3-3-7.. Using 2-2-2-5 @ 2.85v, I can only do 210MHz...MemTest stable...
Cheers,
Minnyboy
Was wondering what timings you guys using OCZ 3200 EL Rev 2 were using with the Neo2??? Would you mind posting your results and vdimm? Thanks
i'm currently running my Neo2 with my OCZ Plat R2 at these settings:
Speed: 275
Timings: 2.5-3-3-7 1T
Voltage: 2.85v
It's good for 277HTT with those settings, memtest stable :)
---dens
newb question: why do ppl say xxxHTT? is it not the same thing as xxxFSB? just another way of wording it? would appreciate it if somoene could help explain this for meh :p:Quote:
Originally Posted by DeNs
Could you please check if neo2 can handle such configuration:
winchester 8x300
memory 2,5-3-3-8(X) 1T changed with A64Tweaker to 183 mem divider ( it should be around 275 2,5-3-3-8(x) 1T
and to prove it run spi 16mb
TIA
yeah seems reasonable,Quote:
Originally Posted by wampir
very possible :)
dont have them tho'
so i cant test it for ya
Works fine. All in A64 tweaker work fine with my Neo2 when you have booted with higher divider; for example 166->183 does not work but 200->183 works just fine.Quote:
Originally Posted by wampir
i have a problem with my neo 2:
The mainboard dont recognize my Soundblaster audigy 2 zs.I' tried every pci slot with bios 1.3, 1.36 bitpower and 1.4, but nothing.
Some one can help me?
Hey all,
I recently tried assembling these and ran into the dreaded cold-temp glitch.
For me the temperature would always read 40C, WinXP would load and freeze or reboot.
I could boot into the bios or run Memtest86 without issue.
I went back on air-cooling and all is fine (2.6ghz, 10.0x260 at 1.55v Vcore).
This was with the 1.41B modified bios (http://www.lejabeach.com/MSIK8N/k8nneo2.html).
I have been contemplating pulling the CPU and RMA'ing it, verses trying the 1.5X bios.
Anyone actually have any luck with this other bios?
What I mean to ask is: Has anyone having this type of problem, updated to this bios and have the problem go away?
Karl
pci/agp are clocked i presume?Quote:
Originally Posted by daryl
i havent seen this problem before so
i cant help you out here, sorry
its a cpu related problemQuote:
Originally Posted by Karl Norris
some winnies like cold others dont
Hi you Nerds out there, whats your opinion to the following OC run..:
Sorry for long post..
assuming there is a rig consisting the following:
Neo2 Platinum Korea CS Bios 136 Mod...
Winchester 3000+ 0440 SPAW
2 X TakeMS 512 MB DDR400 3200 2.5,3,3,8 1T
6600GT , Chieftec 360W PSU with 12V/17A / combined 340W..
along all tests the Coretemp stayed below 55° at load..
HTT @220, Ram 1:1 /220Mhz HT 3x,,2,5.3.3.8 ,,1T,FFoff, agressive etc, all disabled. AGP @67, SpreadSpect. off --> stable..
Same but @230 Mhz , Ram 1:1 --> crash at boot --> ok think Ram is at its end (TakeMs are nown that they usually give up @225 Mhz).
So: Ram Divider to 166:
HTT @ 270 Mhz, RamDiv.166 = 220Mhz, 1,45 V Vcore , other settings as above --> everything fine stresstest etc.., looks stable.
HTT@ 280Mhz other set. as above --> crash, Ram is holding back again..
HTT@280Mhz, Ram133 = 180 Mhz, other settings as above --> stable.. remember still running @Vcore 1,45 Volts
HTT@285Mhz, Ram133, other settings see above --> BSOD: irq notl essor equal uhmm well. maybe the voltage?
Same as above but with 1.6V VCore: --> crash@boot
:confused: voltage seems to make no difference..
What is holding back now? The Rams are @ 186 Mhz or something, so they are underclocked (should work at least @200 Mhz).
Next test:
HTT@285 , Ram Divider 100, 150 Mhz :brick: Vcore back to 1,5 V,
--> stable.. but the Memoryclock.. :eek:
HTT@290, RAM still @150Mhz --> stable
HTT@300, RAM l around 160Mhz --> stable
Htt@310 " " " ,Vcore 1,65V, Vdimm2,85 Voltage doesnt matter --> BSOD
I would highly appreciate listen to your thoughts:
What do you think, is the MCU of the A64 at the end? Or is it the cheap Ram which drops.
If the memclock is around or below ramspecs, is there something which nevertheless loads the memory and can cause failure at high Htts?
The ram works fine @220Mhz but only at low httclocks, the higher the htt the lower the Ramclock must be (far below the specs)..
why?
Next question in mind: i read all the posts of these
:slobber: godmode rigs with high Vcores and that most of them achive better results if they feed more voltage, i personally found that voltage (other than size) does not matter. I reach best result with near stockvoltage, rising the Vcore makes no difference. Why that? Am i missing something?
great forum ..
thx
Hey Jack,
Thanks. I've come to believe this too from reading through piles of threads :D.
I'm going to RMA the cpu and try again. If it's still intolerant, I'll run it on air-cooled for the short term.
Karl
wow my neo2 died.. has anyone else had problems like this before with their board? :(
http://www.xtremesystems.org/forums/...ad.php?t=51411
Did you change the original chipset cooling ??? It's very noisy :mad: 5000rpm . Can change it for self made heatsink ( cut heatsink from celeron 466) is there a problem to run with it and overclock ??
The Swiftech NB Cooler is VERY nice and fits great... may interfere with large GPU, though....
Ive purchased the neo2 after reading this thread :)
I've hadnt had the chance to overclock because want to have updated all the drivers etc...
But when i connect my new floppy drive ... nothing happends.
I can see the floppy drive in "my computer" but when I dubble click it , it tells me to put a medium or something else in the drive ?
But there is already a floppy in the drive ?
Do i need to install some drivers or etc ...
Anyone had this prob before ?
stupid question, but...
I can't find my actual pci/agp bus speed anywhere. Even sisoft sandra, which on my other computers reports "agp 4x (4X66 MHZ)", reports only "8x". Is there any other utility that will show pci/agp bus speed? cpuz doesn't, corecell doesn't, speedfan doesn't, everest doesn't. I'd just like to prove that I'm not accidentally upping the speed of the pci bus.
A64 3000+ winchester at 2300mhz, voltage at "auto"
230 FSB
x4 htt mult
2x512 crucial ballistix pc3200 at 230mhz, 2.5 3 3 8, voltage at "auto"
thermalright xp30 with 90mm fan running approx 3100rpm
I still haven't hit the "wall" on OC (corecell rebooted at around 260mhz with those mem timings and air cooling) but even at 230 mhz prime95 reports an error after about 6 hours so I've left it at 230 for now.
Another weird thing - When I manually set a cpu core voltage in the bios, the corecell vcore adjustment works in REVERSE! I damn near fried my cpu. It was manually set to 1.55 in bios, and when I opened corecell it showed 18% over. I dropped it to 0% over in corecell and the voltage jumped to over 2v and reported cpu temps went from 125F to over 160F almost immediately. Almost crapped myself before putting the overvolt setting back to 18% which dropped temps back to normal. It does this in both shipping version and latest version of corecell (1.6.9 and 1.7.something). I'm using 1.4 bios which came installed.
Anyhow, I really need a pointer to a utility that will report pci bus speed. Thanks in advance.
flenser
Yeah, get and install ClockGen from cpuid.com. Make sure you get the one for NF3 boards. You don't have to use it to OC, or anything, if you don't want to, but it will show you you're AGP/PCI speeds.
Also, it is my understanding that you should in NO WAY use CoreCell to overclock. Just use it for informational purposes and, even that, with a grain of salt.
Thanks. At 230 fsb I'm getting a bluescreen reboot with the pci less than equal to error. Dropping to 220 fsb cures that. I'm not sure if it's cpu or memory that is causing it. I have no irq conflicts and the only card installed is an audigy2.
Hi flenser,Quote:
Originally Posted by flenser
IRQ not less or equal it showed i think. Not PCI...
This BSOD often means a problem with memory / adressing, i don´t think its your pci Bus clock which causes the problem. Have a look at ram.
Next: i had also this wierd voltage readings, mine showed sometimes 3,2 Volts Vcore.. in corecell.. this went away with "update" to 1.36mod bios. I dont think that these Voltage readings are for real, its just a fault in measure / reading..also the temperatre readings were way to high..
But the subzero boot problem is resolved? I haven't understand it yet ...... :confused:
I think right now it is a hit or miss with the sub-zero temp problem. As for me, I am taking a gamble and just ordered a retail 3500+ winnie, in hopes for a new stepping so that I can try with my Mach I. I will definately post, when I get everything set up and tested.Quote:
Originally Posted by vkbms
Right, it was IRQ not PCI, my mistake typing faster than I was thinking.
Memory checks good 8 hours in memtest86 so I'm not sure. The computer only BSODs like that when running this one game... Unfortunately it's the only game I ever play so I'm hosed I think. I thought if memtest was good then it wasn't the memory that was the problem, but I guess not. I've tried running the mem voltage up and down but that doesn't seem to do anything. Maybe I should up the memory timings...
For what it's worth, my AGP/PCI speed is 67/33.5, right where I set it so that's probably not the problem. I thought it might be the vid card but I ran rithdrbl for a few hours and that didn't crash it, ran both rithdrbl and prime95 for a while, and that didn't crash it. It's just that one game that dies at 230 fsb.
I'll probably drop it to 220, open a cream soda, and try to be happy with a crummy 10% OC. I don't use corecell for anything but an occasional test, but since the auto detect bit took it up past 250 fsb before it rebooted, I really thought I'd get more than 220 FSB out of it. The crappy thing is that I really won't be able to tell if it's the memory or the cpu without doing several more days of testing, and even then if it's like now, where everything tests great except for that one game, I'm not sure I'll ever determine what the problem is.