Jump to content

Jangor

Members
  • Content count

    18
  • Joined

  • Last visited

About Jangor

  • Rank
    New Member
  1. pcie connector connected! I read things about corrupted bios and when any videocard is in the upper pcie slot or both ram sticks are in the symptoms are exactly like that. What speaks against it is that everything works with videocard in lower slot and only one stick of ram. I am considering either a long cmos clear or flashing the bios but I would like to hear some opinions first.
  2. Jumper are in sli, x1900xt 256 card is in pcie slot 2, one stick of ram in upper orange slot = stable system with no issues. Same settings with two sticks of ram in both orange slots posts, says bios checksum error, i enter bios and reload optimized defaults, restart and no post. Same settings with x1900xt 256 card in pcie slot 1, jumpers not in sli, one or two sticks in any slot(s) = no way to post, not even after short cmos clear and not even with my x800gto! I am so confused.
  3. If it is the psu, would it become stable in single channel mode?
  4. Had a stable rig, see below, memory and cpu check out well in stuff like dual prime95, superpi, benchmarks for hours etc. Then i chose to upgrade my videocard to an x1900xt 256mb. My old card was running in the second pcie slot so i put the 1900 in there too. System took a lot of resets, cmos clears etc to even post, else freeze sometimes at first ligth blinking or last one on with long beep. Sometimes it would post but freeze during boot. Rare times it would boot and run stable for hours with occasional bsods, sometimes after static on the audio. I thought it might be something like a conflict so I changed the card to the first slot. There it would never post into bios. Thought the card was faulty, took it out and out the old one into first slot where it was running stable before and now ti would not even post at all even after cmos clear. After switching and testing graphics cards, ram slots and ocd/ not ocd system finally i got the new card to run stable in the second slot only but only in single channel (works with both sticks). In conclusion by testing i am pretty sure my components are fine (were working together before well also). So what happened? Just by swapping videocard i can only boot in single channel and the first pcie slot never works (posts after cmos clear, then load failsafes or any bios config that worked before it hangs at last light) . I really tried all bios stuff, the parts are fine and the os and drivers are also fine. Is it that i have to totally reset the board or something like that for it to reassign everything? It seems to me a conflict between graphics card and ram. Any suggestion much appreciated and sorry for the crappy post but this testing took a lot of time and i have work to catch up on.
  5. Which offloadable setting? In the nvidia network controller driver options there are two options: checksum offload and segmentation offload,
  6. Jangor

    Everyday overclock

    OK guys, now things start to get clear - never heard about that! Although now I am wondering why all those options are in the bios to adjust the way C&Q behaves while overclocking... But at least we got that part sorted out - and I was wondering why nobody was playing around with c&q!
  7. Jangor

    Everyday overclock

    most people in this thread are saying that C&Q should be disabled when overclocking, but nobody so far has even hinted at the reasons for that. I took the time to explain how I am using C&Q together with overclocking, explaining what I believe and to my experience is the way it works, why it could be potentially dangerous (if you dont use the dynamic overvolt or the fid limitation, if you find your oc is best with at a lower multi) and how to avoid these problems. Now I would think it fair to not just post again "no C&Q when oc'ing", but to explain why it it is assumed not to be good, should one think so, or at least to critically read the description of the way i use it in my previous post and to comment on it. Sorry if i am pushing the point but i am really wondering why everybody is saying not to use C&Q, when my sys is running stable with oc and C&Q for months, and it actually was not even the main topic of this thread but it is the thing that comes up most as a reply. The main topic was whether there are suggestions how much you can push parts and expect them to last or if you can push them all the way if they keep sort of cool, stable and reasonably low voltages. But to put this issue of c&q that keeps coming up bluntly: do people just repeat a notion about c&q they may not even understand or have experience with, or is there some actual understanding inthe matter of why oc and C&Q should not work together?
  8. Jangor

    Everyday overclock

    @ Mack27 How would cool & quiet kill the cpu faster? S12 600w is working without problems from what i could see in the last couple of stress testing days. @ madu *the way* i am using it is this: i checked that my cpu needs just over 1.45v to be stable at 240fsb x 10. I could just get the vcore to that value if i wanted to just overclock to 240x10=2400mhz. But the vcore and multi get changed by c&q according to load. Usually without oc and fsb at 200, c&q puts the multi to the full 10 and the vcore to 1.35 under load and reduces multi to 5 and vcore to 1v in idle with this cpu. If i oc the fsb to 240 to get the cpu to 240x10=2400mhz and i set the vcore in bios to 1.4v, the next time c&q throttles the multi and vcore to 5x/1v i get 5x240 1200mhz/1v which might work in idle. But when then it senses load and puts the multi back up to 10 it will only put the vcore to 1.35v, which with the fsb of 240 wil give 2400/1.35v which is not stable. So if i instead leave the vcore be, since i can not control the way it is changed by c&q, but use the bios option to always overvolt by 10%, i will be idle at 240x5=1200mhz/1v+10%=1.1v and load 240x10=2400mhz/1.35v+10%=1.48v which is both stable. By the way if you oc with reduced multi, say 9x, you can also specify it in bios as the max multi to ramp up to for c&q. Is there anything wrong or risky with this method? Is it an added stress for the cpu when c&q ramps it up and down? @ blick thanks, that is the kind of info i am looking for. Would it go along with what you say that adding 0.1v to my ram (stock 2.6v) to reach 2.7 to make it be stable at 240 (or 250 if i can get the cpu to that), this combination of slight overvolt and added frequency stress should still not be too risky? Would I need to monitor the temps on the ram to be really sure?
  9. Jangor

    Everyday overclock

    Hey thanks, but the way i am using cool and quiet works really well with ocing. Just to bump the thread and clarify my question: i read various threads about ocing and know the basics and the approach. The question was rather (apart from stability and how to get there) : can I run a ddr400 ram at ddr480 and expect it to last a reasonable time, say a year or will it start making trouble sooner or later? is, say 440, more reasonable? Experience reports would be very apperciated! Anyway thanks a lot!
  10. A little experinece that happened to me. You know how everybody is telling you to have a good psu, at least 480W, 24pin etc.? I was so smart to dish out the cash for nice parts on this rig but thought to myself: now, I still have this quality enermax psu from my old rig, it may have just 465w and a 20pin but hey, it is enermax quality, it will do the job! So I went and built the rig, tested it a while and all was stable! Great! Off to the forum to brag about it! I was so smart! Off we go to work...months pass...all stable...maybe a little BSOD every couple of weeks, no issue. UNTIL... slowly a couple more BSODs! And more! Check the error messages, logs, notifications: error 7, error 51, disk error, paging error, BSODs while defragging, corrupted data on the hd! OMG whats up, it was stable! The hd must be broken, check smart, omg cascadng spin up times atacrc errors, whats up! Finally drive struggles to spin up after BSOD! Must be a failing drive! Swapped hd, new hd also lots of smart errors or dangerous values! Is it the dreaded NF4 data corruption bug? For safe ran a full R/W surface scan on old hd to see what is lost...next day, freeze! Restart, no post! After a lot of part swapping down to one stick of ram that will not post alone or together with another, other one works well! Finally it dawned on me: : : PSU? Can it be? My fabulous but underpowered and 20 pin psu? Swapped psu and even with the one still working stick of ram all went normal! No paging or disk errors, no BSODs, smart values all normal! OMG! In hindsight i should have known, seeing how under load the voltages were sagging a lot. I thought it was ok because speedfan showed them to be still in tolerance but now i know that the values are not accurate anyway and can only be taken as reference: my new psu does not sag or ripple at all (a super tiny little bit maybe, nothing compared to before) under load! Obviously when all was loaded and the disk was thrashing a lot the current collapsed (maybe because of the way i hooked up the psu on the hd) and the drive got no juice. After the BSODs sometimes the drive would not spin up properly, lots of clicking. After a while the it would again. Probably the rail was still hot or overloaded or whatever. Maybe the psu was fine in the beginning but the constant overload made it degrade until it was not enough at all anymore. After replacing ram and psu and all seems working well again I am still wondering if i should still worry about the mobo and the hd. Maybe they got damaged and will fail soon? How can i check? Maybe the hd is more at risk then the mobo since it got trashed by the low juice? But also the ram dying would point to something that might have damaged the mobo too? Suggestions welcome. Else thanks to bear with me with this story! Maybe it can be a warning against saving comparatively little on a psu and putting a lot of other stuff comparatively expensive stuff at risk!
  11. I am trying to oc my system to be that much quicker but still be a daily reliable workhorse for my job. So far I am quite happy with my cpu oc at 2.4 which keeps still quite cool under load on air and keeping c&q enabled so it idles in the comfort zone. (did it by giving it +10% Vcore which dinamically ramps up when the multiplyer increases - great dfi feature - but you probably know) I was just wondering if i can maybe go up to 2.5. What would i have to take care of to be sure that it will not die on me suddently? Vcore up to 1.5 and load temp up to 50°c? What am i risking apart from instability (which i can check with prime95, superpi, occt , memtest)? Shortening lifetime (how much)? And about the ram (2gb set of ocz 3200 platinum -> sig) now it is running stock by memory divider but i had it stable at 250 3-3-2-8 2.7v and 218 2.5-3-2-5 2.7. What would you guys think is a reasonable oc for this ram again not to have it sudden death on me (does that happen for ram?), and what would i have to take care of, considering i dont have a way to measure ram temps? Case has fair ventilation with PWM at 53°C and NF4 at 39°C, both under load and passive sinks. Thanks!
  12. Jangor

    Infinity Ultra vs. Ultra-D

    Guys I don't mean to heat up the discussion and I understand the specs by DFI are there for a reason but I just thought to share with you that i am running on probably the same psu, enermax 465w with a 20pin atx connector and a 20-24pin adapter and it has always been running fine. I dont have the most hungry grafics card and only one hd. Also I am a rookie and it was the first system I built in years, thats probably why I went so careless about it, but still it runs great. (ducks into cover)
  13. Thanks for exploring this topic with me guys!
  14. I am using an older enermax psu with 465 Watt and the old atx 20-pin connector with a 20-24 pin adaptor. I connected this psu to my mainboard at all 4 points, and with my current configuration it is running well, ocd at 2400 and stable under prime, superpi occt, etc, so I was using it with confidence, until i noticed in speedfan that the -12 rail is only at -9! :confused: Is it a bug in speedfan/the bios, or is the PSU the culprit, just too weak/broken/not good with adaptor? And anyway, what is the -12v rail for? Is it important? Why does it run at all so much out of spec!? Thanks for any suggestions, I thought this psu is still good but now I am wondering if I should get a new one.
×