Looks alright. Heatsinks are rubbish though.
Looks alright. Heatsinks are rubbish though.
Fatal1ty P67 Professional
L1.84http://www.asrock.com/mb/beta.asp?Mo...20ProfessionalQuote:
Modify EZ OC table.
Hey guys. Skimmed through the thread, read about the last 12 pages or so. I might be picking up an i7 so I wanted a little feedback...
There's really two options, Extreme4 or Extreme6. 4 seems to have all the bugs ironed out. Any reason to get Extreme6 (and spend an extra $40)?
One other question too, for highest OC, stick to air or move to water? (I can't really get good feedback on the WCing forum :shrug:)
EDIT:
First question was basically answered by Ket in PM! Thanks! :up:
Any opinions about cooling?
You have PM :p:
ED - For cooling something like a modded CoolIT ECO or a Coolermaster Hyper 212 Plus will get you good temps and probably enough for 4.6-4.8GHz 24/7. If you get a chip thats a hothead though like my 2500k you will need to put it under water.
I guess the best idea is to run it on the stock intel cooler (which looks fairly high end, amazing they didn't do that style before!) and see how my temps are.
Then descide where to go from there since I'll still have my main rig running.
Ok, I'm having the same problem as I did a couple weeks ago with the system not posting. I have tried the cmos clear using the proper method, and I have disconnected every piece of hardware. Now, when I turn on the pc, if I have the ram in slots 1 and 3, I get error code 2E. If the ram is in slots 2 and 4, I get error code 30. I have tried other ram, and it does the same thing. Is it time for an RMA?
was a little tired of my Asrock xtreme6/bios
so I'm working on a new motherboard, here's also a lot to learn. use watercooling
but for 24 / 7 use, I will not go into more a 5GHz with Asrock, so you can keep the problems away
but I've been happy with my Asrock, just it would be slightly more stable at high oc
This thread is about Asrock boards. SS from other boards are irrelivent, and not helpful.
For that kind of buck and load - yeah. It can clock better!
It ended up being the cpu. I had it oc. to 4.8 with offset of .100 v, and pll disabled and at Level 4. Everything else was left on auto. Temps maxed at 64 degrees. Did I push it too far? Or just a bad chip?Quote:
Originally Posted by mean6
Only 64c? Was that while running IBT with AVX? My 2500k hits about 78c @ 4.6GHz running that friggin stability test. I never saw any benefits when raising VTT to help keep high memory clocks stable, but VTT has helped me keep a pretty high OC at low Vcore. Its a recent development for my chip, VTT used to do nothing at all when I last tested it.
VTT @ 1.1v... needed 1.36v for 4.6GHz
VTT @ 1.25v... 1.32v for 4.6GHz.
Good news people! I've just got some info off of one of my Asrock contacts regarding the USB3 BSOD issues. R&D have managed to duplicate some of the problems and are looking further into the matter, some devices have also been sent to Etron for them to look at the problems as well.
Slowly but surely, progress is being made toward the cure :up:
Good job and thanks :)
All I've done is thoroughly test, bring bugs to Asrocks attention and provide detailed info to Asrock. While I aknowledge what I've done is important Asrock deserve equal kudos for doing their all to fix the problems and not simply trying to brush them under the rug :up:
Well, the new 2500k chip I got is one of the costa rica batches, and it has to be up there with the worst of chips. Anything over 4.2 requires PLL enabled, and will not come out of sleep with anything over stock. :(
Thats super pants... even makes my hothead 2500k look good.
What's bad about PLL enabled apart from the sleep thing? Btw my Costa Rica 2500K would do 4.4ghz @ ~1.19V under load (1.24V LLC lvl4) with PLL enabled
My old chip would do 4.8 with PLL disabled @ 1.34V
Hey Ket, I know you have been asking about whether the IHS on the new SandyBridge are soldered. Since I had this dead cpu sitting here, I figured why not. Well, I popped it off, and the answer to your question is..........(drum roll), YES they are. :)
What about the L1.85 beta BIOS? Someone probe it?
i had my costa rica up to 5.0 ghz..updated my bios to 1.61 and now i cant get it to boot at 5 ghz.....stuck at 4.7..i can only guess im missing a setting somewhere in my bios
never mind i fixed it....running at 5gz...processor is costa rica (as if that matters) batch number is 3052a185
What settings are you using? I can't get mine stable at 4.8Quote:
Originally Posted by pythonse
Please share your error/fix! :)Quote:
Originally Posted by pythonse
Can we get pics? :DQuote:
Originally Posted by mean6
Ket, here's my $200+ contribution to the forum. As you can see in the pic, the cpu die is still attached to the IHS. :DQuote:
Originally Posted by Ket
change as little as possible in bios..no need for most of the setting to be modified to get it to 5 ghz and stableQuote:
Originally Posted by mean6
so off the top of my head i have
cpu ratio at 50
turbo boost 255,255 and leave the last one at 1
then i set my ram at 1866
upped the voltage to 1.6
upped the voltage on my cpu to 1.4
saved it to profile and booted up with 1.61 bios intalled, with my 570 gtx im getting over 33,000 on 3dmark 06
i run core temp,precision pro,and asrock extreme untility on the desktop to monitor settings,temps....ill bet ur stable at 4.7 with just upping ur cpu ratio in bios to 47? and nothing else changed? what ram do u have?
what happened to ur chip? u burn it up?Quote:
Originally Posted by mean6
That pic should go in some kind of hall of fame its pretty awesome :D Guess that settles my last hope for cooling my hothead 2500k down though, its just a hot mofo. Hmm... maybe I could run the chip just for like 2 mins though with no cooling on it at all... enough to let any solder spread over the CPU die that maybe isn't being covered properly.Quote:
Originally Posted by mean6
so he fried it?
Dunno, but that has to be the first naked 2500k pic on the interwebz :up:
Ya, not sure how. I was using recommended voltages, and temps where really good. Just went to turn my computer on in the morning and it wouldn't post.
Ouch. I guess that chip must of had a weak IMC or something to die like that.
why didnt u send it back and have it replaced?Quote:
Originally Posted by mean6
Sometimes destruction is better therapy :yepp: ;)
The guy working at the store was telling me that intel can now detect whether a cpu has been oc. and will deny warranty. Any truth to that? I should have asked that before I decapitated it, but as Ket said, "sometimes destruction is therapy"!!!!!Quote:
Originally Posted by pythonse
Intel probably can tell if its been OCd, but the guy in the store was full of :banana::banana::banana::banana:. Intel sell the K series full well knowing the chip is unlocked and will be clocked by OCers and price the chip accordingly to cover any RMAs. In a nutshell, unlocked chip, intel know its going to get OCd, yet the chip is still covered by a 3 year warrenty. Dude in the store is clueless or just couldnt be bothered with processing the return.
So, this new processor batch# 3049A355 absolutely sucks! Last night I had it running 4.6 @ 1.36v. Played C.O.D all night. No problems. This morning I went to fire up the computer, and just got the blinking cursor. So, after playing with it for over an hour, I can get it stable at 4.2 using 1.32 v. But anything higher, I get the blinking cursor. Now is it a poor cpu, or is my mobo killing these chips? All other voltages have be let at auto.
I noted even when I set 1.32v with fixed vcore there would be spikes of 1.4v+ according to realtemp.. or was it coretemp? one of those anyway and temps would spike with those voltage spikes but CPU-Z and AXTU remained at a constant voltage of 1.32v so theres no direct evidence to suggest the vcore spikes are due to the board. If I had to guess I'd say if the spikes are legit and not a sensor anomaly the spikes are being caused by something to do with turbo voltage.
I'm hearing more and more often about the blinking cursor issue. The same scenario. The processor is stable as rock in operating system but next boot and you get blinking death. I can remember one particular case when replacing the mobo did the trick. But don't know which board was replaced by which. Maybe someone experienced that issue, resolved the problem and could share a solution?
Just wondering, if I should trade the chip for another, or the board?
I would return the chip. 1155 boards have more or less zero to do with how a SB chip OCs, its either a good chip or its not ;)
Still playing with this chip. Anybody try 1.62A bios. I tried it out and found it to be really buggy. The offset voltages seem to be way off compared to the numbers I got with 1.61. It also has the spread spectrum bug as in previous versions. In the end, I went back to 1.61
Just an FYI, this won't do any good. Melting point of solder is ~180C. Your chip will reach Tj and shut off before any solder melts. You could use really short bursts of heat with a blow torch to do the job, although that still risks the possibility of burning up your chip.Quote:
Originally Posted by Ket
The thing is: the chip seems to be perfectly stable (in the cases I have read about) but after powering down next boot ends with blinking cursor. That could be motherboard issue. Maybe some setting/bios combinations.Quote:
Originally Posted by Ket
I didn't think 1.62A was that great either but it didn't have the spread spectrum bug and neither has any UEFI since 1.6A I would suspect something else is going on there.Quote:
Originally Posted by mean6
I just noticed that when I had cpuz open the bus speed shown was 98 rather than 100. I went into the bios, changed spread spectrum to disabled, and it wouldn't boot. I remember those exact symptoms from earlier bios'.
I've tested UEFI 1.6A, 1.61, 1.62 and currently testing 1.64A, none of them have the spread spectrum bug on my Extreme4. Downgrade to UEFI 1.3 (DOS version), load UEFI defaults then save and exit, then re-update to whatever UEFI you are using.
I am getting some weird stuff happening. I spent all afternoon, overclocking and trying to stabilize my setup. I had it prime stable for a couple of hours. I shut the system down, and left it for about 45 mins, turned it on and got the blinking cursor again. So, just for kicks, I loaded uefi defaults, rebooted, then loaded my saved profile, and it booted straight into windows. I think there is something screwed with this board. Might explain why my other chip went on me. Any ideas????
Use stock settings then see if you still get the blinking cursor. If you don't, seems unlikely it would be the board.
I did the recommended way of flash you said, and now it has the double boot issue. The only way I can get into windows now is if I boot via the bios. Not a cold boot.
Very strange. Maybe the board has developed a fault then as UEFIs 1.6A through 1.64A deffinately don't have the spread spectrum double POST bug.
I went back to defaults, and haven't had the problem since. So maybe it is cpu related. Just doesn't make sense. I had it running at 4.8 all afternoon playing games, and running prime. Now the best oc I can manage without any of those bugs so far seems to be 4.3 :( Should I take my chances with another chip?
It would seem the logical step as you haven't had any problems @ stock. It may be worth testing your memory as well with HCI memtest.
I'm just wondering... am I gonna notice a difference between this cpu running at 4.3 from my last chip that did 4.8? Probably not. The only kicker is that my previous system was an X58 set up using an i7 920 running @ 3.8. Not really worth the upgrade, if I look at it that way. :(
New 1155 systems are storming fast, but theres no way you should be happy with 4.3GHz even a "locked" 2500 could do better than that.
The new beta bios for the fatal1ty board fixes the bug of oveclocked cpu reverting to stock clocks when disabling all the power saving features, ie speedstep, etc when it enters windows.
Ok, to explain the situation I am in now. My system will power up fine at stock settings, and up to 4.3 oc no problems. Sleep mode even functions noramally. Here's where I am kinda confused, I can go into the bios and set it to 4.8 oc, with 1.4v, using offset and level 3, save and exit, and it will boot into windows no problem. It will hold up through 2 hours of Prime, and a long session of C.O.D. The problem is, if I shut it down, then power up again later, all I get is a blinking cursor. If I reboot, enter bios, then click save and exit, it will boot into windows no problem. So, is this a cpu issue or mobo issue? Just want to know, cuz I am planning on exchanging something, either mobo, or processor. Thanks for the help.
probably is the Cpu. Got the same problem here @48multi
try 49 how is it doing. Sometimes there's faulty multi that yyou should skip!
tried 44-50, it's the same thing. After cold boot I get the blinking cursorQuote:
Originally Posted by BranFB
Internal PLL Overvoltage is Enabled ?
2 hours of prime don't make a stable system.
These sandys are peculiar things - especially with the offset voltage. Most of the time the CPU isn't getting pushed (or if it is your settings maybe stealth down clocking).
So what appears 'good' to you may be falling way short after 2 hours prime? and some gaming.
Ideally run IBT with AVX - just to check your Gflops - not stability.
Then Prime 95 blend (lastest version) on your final settings for at least 12 hours.
BTW all of the above should be done at stock mem timing at 1333 or 1600 if your memory supports that.
IF your overclock your memory, then you need to run Prime 95 large FFTs with your final overclock to check that too.
Sorry but this is OC'ing 101.
You've killed 1 cpu already.. just saying.
1.70 Extreme4.
Quote:
Improve mouse movement in BIOS setup.
I'm little disappointed...
HCI memtest is far superior for testing memory.Quote:
Originally Posted by banwell
Thats it? Improve mouse movent in the UEFI? Disapointing. 1.64 > 1.70 is a big jump just for a mouse movement tweak..... Seems more likely 1.7 has all of the fixes of the betas as well, but somebody will need to test it to verify that.Quote:
Originally Posted by burebista
If the CPU were busted, it wouldn't boot at all.Quote:
Originally Posted by mean6
it has NOTHING to do with him not running prime for 2 hours. I can make prime crash in the first MINUTE, from low vcore, but my CPU boots *every* time. The fact that he can enter windows and game for hours tells me that the CPU is working perfectly. These things either work, or they don't work.
The blinking cursor you have is a symptom that others have when PLL overvoltage either needs to be enabled, or when you're outside of the frequency range where PLL overvoltage works. It sounds to me like your board is having issues setting PLL overvoltage on a cold boot. Have you tried toggling this setting, and seeing what happens if you enable it, or if you disable it, and then try to cold boot?
Of course there are other reasons for a hanging cursor like that also...
1.70 loaded.
you guys are screwin around with too many settings to achieve ur oc rating....i have 1.61 bios, no dual boot,i have minimal settings tweaked to get to 5 ghz...set everything back to default uefi settings and turbo up to 4.6 and see if its stable...if not something is wrong
Quote:
Originally Posted by banwell
I have been overclocking for years, and have never killed a cpu. This is a first. Even this one, I wasn't even pushing to far. 4.8 @ 1.36v, with PLL disabled. I don't think is much at all. As for being stable cuz I can play games and run prime, all I meant by that, was that the settings that I had are letting me do my everyday things with no problems. The memory I have runs at 2133 mhz, and has not been touched, just running it's default settings. My point is that I think I just have a bad overclocking chip.
Yes, I have tried switching between PLL enabled and disabled. Anything over 43X requires PLL in order to get me passed the blinking cursor, but only when I boot to windows via the bios. Cold boot with anything over 43X gives the blinking cursor regardless what the settings are.Quote:
Originally Posted by Falkentyne
I have tried using the turbo overclock and nothing past 4.2 will go into windows.Quote:
Originally Posted by pythonse
so u had one bad chip already? and now problems with this one also after u replaced the bad one? if this is true u could have a bad board....ur power supply is a good one i take it?
And what voltage are you having to feed the memory & memory controller to run at 2133?
There is little to be gained running the memory past 1600 - what ever they are rated at it or not. I am thinking that this is your main issue. High rated and 'juiced' memory with a moderate overclock.
Personally I'd rather have looser memory and higher cpu overclock.
That's what I am trying to get to the bottom of. I know, ever since I got the B3 replacement, things didn't seem as smooth as the previous version. But the system did seem to run ok for the last 2 months. No signs of overheating or anything. Just went to turn the computer on one morning and it wouldn't post. I popped out the cpu, and noticed a little "burn mark" on the bottom of the chip which I was able to rub off. Tried it on another board, and it wouldn't work. I am using a 750W PC Power & Cooling power supply.Quote:
Originally Posted by pythonse
Just using the rated voltage of 1.65v I should really setup my signature.Quote:
Originally Posted by banwell
^^man if that's the case I feel for you. Sound like you've got a defective board. I'd be submitting a RMA asap and claiming for the 2 cpus its bummed.
Signature done.
Now, I am kinda in the same situation as I am with the cpu, they both work fine with stock settings. Where does the board fail?Quote:
Originally Posted by banwell
Any sort of burn mark on CPU should be enough IMO.
well trust me when i tell u if u have the same problem after replacing the cpu..its something else. look at everyone else here who has the same hardware. other than swapping out parts...i dont know what else to suggest as i dont think it could be a corrupt windows install. u have 7 64 bit? if u do i had to reinstall mine once due to it not booting after a bios update. im just trying to shoot from the hip here with suggestions
to help...sorry ur having to go through this i know how frustrating it can be
You could always just exchange he board and CPU to be done with it, you'd just have to live with the fact you'll never know what the fault really was :p:
The thing is, my old chip had no problems overclocking. It's this cpu I have now that I am have a hard time with. I am running windows 7 x64. I can boot into windows no problem, it's once I try oc. past 4300mhz, is where I run into problems. That's why it's hard for me to say whether it's the board, or the chip?Quote:
Originally Posted by pythonse
Since you probably have the best relation with ASrock, maybe you could help. Before I knew it was the chip originally, I thought the board was at fault. I contacted Asrock via email and told them my issues, and mentioned that I have owned the board for 2 months. They were very quick responding which is a plus, and gave me some troubleshooting to try. None were successful, I notified them of this, and they told me to contact the vendor for a replacement. I was unsure of this but, I contacted the place I purchased it from, and they won't replace anything that is over 30 days, and told me to try and contact ASrock again. It seems like I am just running circles. What would you recommend?Quote:
Originally Posted by Ket
AFAIK the vendor is primarily responsible for replacing the board for the first year of warrenty. I'd get back onto them.
yes...its not like they eat the cost of the board as they send it back to asrock..total bull crap! where did u buy this board?
www.canadacomputers.comQuote:
Originally Posted by pythonse
Never did get my system stable past 4.5ghz :(
Turns out that previous "success" was short lived.
I also noticed that I happen to have the same memory as Mean6. I *NEVER* got it stable @ 2133 (advertised speed and timings). Funny thing is that the G.Skill folks on the forum blame the motherboard. :rolleyes:
u can de-tune the ram to run at 1600 and u have xmp enabled? i would put the ram at 1600 or loosen the timings on the ram..u have all four slots or 2? if only 2 slots which ones u have filled?
What slots are you using? I found that the white slots worked best for me.Quote:
Originally Posted by blacksun1942
i suggest backing off the timings...also running them at 1600 or 1866..you dont need 2133 ram
Sure you do. 4.5GHz and 2133 vs 1600 is the difference of 45secs per pass in IBT and 36secs. For my purposes thats a huge time saving when it all mounts up.
im running gskill 1600mhz at 1866 speed with cl8 timings and im at 5 ghz and can push it further if needed...u do not need,meaning have to have...2133 ram.
Somebody told me the exact opposite (might have been in this thread); I'm using the blue slots, haven't tried the white ones. Guess it wouldn't hurt to try... :shrug:Quote:
Originally Posted by mean6
What are your BIOS settings, if you don't mind...? Extreme6 or 4? 2500 or 2600 cpu?Quote:
Originally Posted by pythonse
I would *LOVE* to get to 5ghz. I feel like my cooler is overkill for the paltry temps I'm getting...
I've got a 2600K with the Extreme4 board and I seem to be having some issues that are similar to what I'm hearing other people say here. When I first started overclocking this rig, I could get up to the 4.4 automatic turbo overclock in the UEFI to boot into Windows, but 4.6 just gave me the black screen with blinking cursor. I could get 4.5 to load windows by setting the multiplier myself, but 4.6 wouldn't boot Windows no matter what I tried.
Then, I came across a post by blacksun1942 that linked to some foreign site that had some settings that blacksun1942 said allowed him to get into Windows at 4.8. The link is http://ts2peca.no-ip.info/?p=604
When I used those same settings I was suddenly able to get into Windows at 4.8 as well, but any kind of stress test would either blue screen or freeze up the system. I wasn't willing to give it any more Vcore for stability so I backed the voltage down to 1.35v and was trying to get 4.6 stable in Prime95. I actually had it stable with small FFTs overnight for 8 hours without a single error. I then rebooted the system because I was going to try one notch down with the voltage so I could get better temps and decided to shut the system down completely to make sure it would cold boot properly because prior to using those settings I could never get 4.8 to boot Windows before. After shutting off the system completely and turning it back on, sure enough it was back to the black screen with a blinking cursor. Keep in mind I had changed nothing at this point from when it was 8 hours stable in Prime95 small FFT, but now it would not even boot Windows. It got past POST, but no further. This makes no sense to me. Since then I've tried every combination of settings I can think of and I can't get into Windows at anything higher than 4.5 even if I use the settings from that site again.
I'm no expert about this kind of thing, but it doesn't seem like a CPU limit to me. If it was the CPU, then why would it fail to load Windows at the exact settings where Prime95 had run for 8 hours previously without errors? I've considered doing an RMA of the board, but I'm hesitant because I don't know if it's actually the board or the CPU. I was hoping a future BIOS version might help but I just flashed 1.70 earlier this morning and nothing has changed.
At this point I'm not sure what to do. I've got 4.5 stable at 1.3v and the system is certainly fast enough, but the fact that I had it stable at a faster speed that I can no longer boot into is really bothering me.
extreme 4...2500k...this ram....http://www.newegg.com/Product/Produc...82E16820231445
bios setting are left alone..i mean i didnt change anything but these settings
cpu ratio=50
turbo boost=255,255
cpu voltage1.38-1.4
llc=3
ram=xmp,1866,1.6 volts
bios 1.61..no dual boot
reboot and everything is stable
i run core temp.precision pro and extreme tuning
I asked one of my contacts about your predicament with canadacomputers to see if theres anything that can be done to help you out. I'll let you know :up:Quote:
Originally Posted by mean6
By any chance, did you have anything plugged into the front USB2 ports? I've had this happen to me literally once or twice and both times I had a USB pen drive in the front USB2 ports, took the device out, Ctrl+Alt+Del restarted and no problem. My guess would be the system tried to read from the USB device thinking it was the OS drive for some reason.Quote:
Originally Posted by VonZippa
I've had a Logitech Rumblepad 2 gamepad plugged into the front ports on and off, but it hasn't made any difference. I haven't left any pen drives plugged in or anything like that.Quote:
Originally Posted by Ket
I'm afraid I'm at a total loss on that one then :confused:
Well, at least that makes two of us. :p:Quote:
Originally Posted by Ket