Jump to content

BlindGuardian

Members
  • Content Count

    88
  • Joined

  • Last visited

About BlindGuardian

  • Rank
    Member
  1. I think I've resolved my issue. In a recent WoW patch maybe 2 or 3 patches ago they introduced dual core support or better optimizations in their code. I've never had an issue prior because it was never using both cores aggresively if at all. If I set affinity to just core 0 it seems to be much more stable and I do not see a loss in performance. Either 2 things going on, one of my cores is weaker and the overclock is interfering which doesn't make sense since it passes prime. Or their code is has issues with dual core and they are oblivious or not advertising since it works 75% of the time without crashing. Other people have the same symptoms with the WoW crashing, it's just that when it happens to my DFI board I get the weird symptoms. Lately however I've let it just hang there for a 30-60 seconds and then press and hold down the power button to shut off, this usually let's me just repower back up instead of getting blinking LEDs everywhere, perhaps I was somehow triggering a sleep mode pressing and holding the power?
  2. No the video card is not OC'd and I upgraded to the latest drivers the first time this occurance took place. I'm not questioning my systems overall stability as it has passed 12+ hours of prime on both cores in tandem. Just this one application and then the subsequent results with the machine refusing to boot thereafter and the flashing leds wherever there is an led on the board including nics. I appreciate your help, I've definately thought about these things and value the 2nd opinion but I'm not sure there is an answer, I was just wondering given the scenario what anyone thought. Memory errors would be persistent not just this one application and definately not Prime Stable. This happened to me last night after 3-4+ hours of straight gameplay(it's been 1 day since this happened before), then after this occured same symptoms as described above I was able to get it back up and running and continued playing for 3 more hours without issue. I just don't know how to explain this, logically a faulty component would cause this to occur more frequently when being stressed I assume, which makes no sense for the gaps in time this symptom appears and then working perfectly for hours after the fact. Again I would assume overheating or PSU but I don't see anything to indicate either especially when PRIME generates load/heat/psu strain and is stable. Looking from outside in you can't take my word for it and everything must have a logical explanation but all things considered stated as truth what would you try next if you thought the PSU/MEMORY/VIDEO CARD was not an issue? Just confirming this doesn't make any sense is good enough for me at this point. I'll chalk it up to newly introduced software bugs and hopefuly they get fixed in a later update.
  3. I have not done a memtest per se but I did leave it burning in prime on both cores all day while at work with no lockups or adverse side effects. System has been running without issues for almost 2 years now. I have not checked the voltages under load but any other stressful application/game what have you works fine and never have I experienced this with any other software. I recently started playing wow again 2 weeks ago and this is why this is an issue all of a sudden. and it doesn't happen consistently enough to lead me to believe it's hardware related? I honestly don't know, I would assume it's a power supply without anymore information to go on as the would most likely be the cause of hard locks without blue screens and the issue of not starting up but I'm reluctant to assume as much since everything else works fine. Specifically I was hoping there was some explanation for all the blinking leds on every led on the motherboard, not just the diagnostics as to what that is? It can't be good but I can't find any documentation from DFI on this, just the codes for the diagnotics led at the bottom. Something that would definately say it's one thing or another. I'll double check my voltages with a multimeter but was wondering if anyone had an idea of what that is. I think I googled something about being a fluke with the board. Sorry, I'm at a loss here. Just want advice on what to look at specifically. It's most likely a bug in their software as there was a 40 page thread about the lock ups occuring for people with mostly Nvdia and some ATI cards, my only gripe is why the lockups effectively put my system out of commision for 10-15 minutes which could be something else entirely, and of concern.
  4. I've tried searching everywhere and I could not find anything that comes remotely close to my issue. So far the only time this happens is when I play World of Warcraft and it's very sporadic, may happen a couple times or may go a week without occuring. I was going to assume my issue is temps but my system is watercooled and even though temps are somewhat higher then usual they are not over the top. Basically while playing wow the system will freeze solid and I have to power off completely, upon trying turning the power off on the PSU and back on all the LEDS on the motherboard and even the NICs; not just the diagnostic lights at the bottom start flashing orange/yellow on and off and no matter how many times I mash the power button it will not start up. I need to leave the system PSU powered off for 5-10 minutes and then it may start up again. At first I thought it could be a PSU issue but the voltages look consistent without any dips or fluctuations. I know others have complained of this bug in wow but no one can prove it's wow and not hardware issues their support will not confirm/deny anything is wrong just the generic check your system blah blah blah. The blinking on and off orange leds on every section there is an LED on the motherboard is what's got me wondering if there is possibly something wrong aside from buggy game code. My hardware has remained unchanged in my sig. I change the water out of my system regularly but admit it can use a good cleaning/soak in vinegar and new tubing, I am reluctant for this endevor as I'd rather wait till I upgrade and have to put new components in anyway. This system still has plenty of life in it so I may have to conceed to this sooner or later anyway, I just don't think it's the temps that are causing the lockups as it's never been an issue in any other application. Any help or suggestions would be appreciated, this is not an issue that is preventing me from using my machine but I would like to know why it is occuring all the same and what I can do to prevent it. Thanks in advance.
  5. Get an opterown 170 and life is good, run your current ram on a divider. I play f.e.a.r 1600x1200 with all the effects except antialiasing online without a hitch. Same with Oblivion it just runs through areas so fast without any tweaking. Tomb Raider Legend is also very very nice full options 1600x1200.
  6. Just an FYI: There is a review over at firing squad about dual-core performance in Oblivion. It didn't seem to be any more informative then this thread already is. Basically L2 cache for the win, and yes dual cores score better. *This was supposed to go in the other thread about opteron 170*
  7. Mine is 24 hour prime stable on both cores at 280x10 with 5/6 divider(166)
  8. I believe it hovers around the 50% mark is becuase 50+50 = 100% and even though each core is seperate it is still one cpu and using only half. I was also lead to believe that even though software is not optimized for dual core, it will still get a boost becuase it works like raid striping and sends chunks to whatever core is ready at the time unless you remove affinity for both cores and stick to core 0 or core 1. Lastly, the number one reason to get dual core, set your MBM5 and Winamp to 1 core, and play everything else on the other with no ill adverse affects. I also doubt a single processor will; or has as much cache as an opterown and it may clock up to the same speeds but the cache makes a huge difference and always has for any processor single or dual. and don't forget it's the most future proof investment as they are talking about quad cores and reverse hyperthreading. But, please correct me if I am wrong, I am still very much learning about my own processor and this is all stuff I've just picked up while researching.
  9. So far so good , powered up no problems right into my oc'd settings. I just need one more modular molex connector to get everything, that or I will need to pickup some splitters tommorow or something. I wonder if they sell the modular plugs piece meal. Anyway volts look standard, I recall Antec always reporting 4.75 in MBM and it was noted in the review at Driversheaven that the +5 rail was reported under but didn't affect performance or stability so until I see proof otherwise I will write it off as a fluke. Also I don't have the molex and floppy connector plugged into the board for now as I don't have any more available down below. Thank god for the PCI-E plugs or else I'd be in trouble. The one you had was most likely returned bad, I made sure to notice all the other Antecs were shrink wrapped and it seemed odd the NEOs were not, I also saw a return/refurb sticker on one and they wanted $80.00 for it, my time is worth more then $30.00 to go back and return it and fiddle around, so I saw the only one that had shrinkwrap still on the box and purchased it. The place didn't have any OCZ products.
  10. I am getting a new PSU because the current one likes to shutdown the PC when the AC compressor kicks in. I tried a line conditioner and that fixed the problem but it causes the damn router to restart now when the line conditioner adjusts to compensate for the AC unit. Anyway the old 20-pin Antec truepower thats worked flawlessly for years on the same circuit as this AC unit is not affected the slightest by the AC unit or anything else for that matter. I was hopeful the line conditioner would solve my problems, and it did in a way but having a stability on the PC without having stability to the net isn't any better. So I have to get a new PSU and get rid of the line conditioner, one that won't be affected by the AC unit. Well I read all the recommendations of the OCZ PSUs and was thinking about powerstream(What, you don't like rice? Tell me Michael, how could a billion Chinese people be wrong? - The Lost Boys) and noticed it doesn't have active PFC or any of the other features advertised on PC Power & Cooling( was thinking about this one too but $250.00 locally to get one today,right now seemed a little extreme) So since I know that other Antec works like a charm in this environment and the NEO HE got a gold mark from Driversheaven I figured why not. Now that the damage is done and the thing is sitting in my car waiting for me to go home and plug it up. I read all these things about problems with ASUS( I don't care ;P) but there are maybe 4-5 people with DFI boards who say it's great and 3-4 that say it won't even startup period. Now I'm second guessing, it's not a big deal to return it but I just don't understand why no love for Antec Neo in any recommendation threads or anywhere else. In searching, you maybe pull up a page with inconsistency about whether it is decent or not. If like any PC hardware you hear more about the bad then there is good, then by that logic it must be a great PSU? If the OCZ powerstream was modular, I would probably not even bother. I think that sold me on the NEO. Everyone says the modstream sucks. hehe I'll probably end up with a PC P&C after all is said and done, I just wanted to find out why NEO is like the boogyman of PSUs, it will get you, but there's no proof it will, with DFI anyway.
  11. Thanks for the replies. after consistently running some games I figured it was nothing to be concerned about, but this definately helps to remove all doubt.
  12. Hmm, I read the guide on VR-Zone. Is it worth doing a voltage mod? I'm already running 716/895 ( lowered from 720/900 just to be safe, would not test higher then that on core clock) I really can't complain, but I am watercooling so if it's not going to hurt much and get another 80mhz on the core it might be worthwhile. Any opinions? I personally don't think it's worth it for another 30mhz at most.
  13. Hey all I'm wrestling with this ram and it's insane, I am running prime stable albeit some slight quirks like 1 core being delayed by a minute past the 8 hour mark or however long the progression of the delay is to get it to that point as they both start out the same and the first hour or two finish each problem at the same time. Anyway my question is not so much about that but the fact this ram is really finicky and I have to run at 5/6 166 divider. I'd love to go back to 9/10 180 or even 1:1 yeah right. I noticed in G.Skill Techs best settings the LDT/HTT is set to 3x. In the DFI OC guide it says the default is 4x. Is it imperative that I run 3x? I thought this is for the memory controller and nothing to do with the ram, just the fsb you run so you don't go above the 2000 threshold for the memory controller. Does this rely on the ram then? Should I switch it to 3x? at a 5/6 divider I am having no issue with 4x ldt and at higher I noticed instability with it at 4x or 3x. Should I turn CPC off and try to run 280 at 1:1 or will the 1t give me better performance at the lower divider? I've looked at all the guides and am somewhat happy with where I am, I just keep thinking I am doing something wrong and could do better. Any help and tips would be apreciated. Thank you.
  14. I don't know when it started lagging or by how much. I was running custom tests at 600mb of ram each. I started both tests on each core respectively about 11pm last night and this morning one of the cores would lag behind in finishing the calculation by a minute before starting the next problem. It did not start out lagging behind, when initiated both completed each problem at the same time give or take a few milliseconds between pressing ok on each window. Things I've learned that cause problems in prime. 1> Memory starvation, I.E. one instance eating up all the available ram leaving none for the other instance. I figure 600mb x2 would be adequate on a 2 gig system, am I wrong? 2> Core affinity, if the processes aren't set with affinity in taskmanager even if you start each with -A0 or -A1 it will run 50% utilization on one of them and fluctuate between 80-95 on the other core and it will finish much slower. Last night however I made sure that in system resources the cpu usage was 100%( 50 and 50) and both cores were equally at the top of the utilization graph. Lastly this ram won't run prime stable on a 180 divider, which comes out to 254fsb. I have it at a 166 divider now. Is the memory bad? It will prime fine at the optimized defualts for it and the cpu just fine which is (250fsb for this ram). I am using yellow slots, I had mem test errors before on orange with this in m emtest test #5 even at optimized defaults for everything. I flashed to 704-2BT and still had issues with prime stability at 180 divider, I reflashed back to the 2/29 bios and it did pass memtest #5 on the orange. I'm not sure why this is? I'm pretty certain that if I had the ram seated badly or not enough contact, I would have had more serious issues. This doesn't make very much sense. I may try putting it back in the orange tonight, just wanted to hear some thoughts..maybe the ram should be rma'd? I hear that quite a bit in the forums about memory being culprit. I tend to not believe something is defective if it can operate at optimzed defaults as advertised but still surprised 4 mhz over that can make it so testy. Thanks in advance for any advice,tips,etc. *edit* almost forgot, I did play with the timings and the overclocking guide advice, and made some more lax but still did not get results, if ability to boot period with 704-2BT, I am currently running 166 divider with G.Skill Tech's settings and give or take a few more lax timings from the excel spreedsheet compiled here for G.Skill.
  15. Swiftech kit is the way to go, however most places online you can put together your own order and come very close to the swiftech kit prices or maybe cheaper. Dangerden gives you some discounts on when you use their website to put a kit together. If you want a quality system definatley prepare to spend between $200-$400. Like everyone has said the Apex Ultra kit is great, I would have gone with an apex kit but I wanted to many custom things. I wanted a DD Maze4 GPU block (difference between mcw60 is negligible better or worse so I hear), I also wanted a Storm Revision 2 waterblock. I think they give you an apogee. I wanted different tubing, and also different coolant addative and I have a different radiator. So ultimately before you buy anything I recommend lots and lots of research and forum thread reading on the popular extreme/oc hardware sites. You can learn about a lot of snafus to avoid and the right questions to ask as it pertains to your particular situation as you will have a better idea of what you are looking at.
×
×
  • Create New...