Jump to content


  • Content count

  • Joined

  • Last visited

About Habanero

  • Rank
    New Member
  1. GPU temps are fine, I have a full tower, with two 80mm fans in the rear grill just above the cards, and a 120mm Sunon as a side intake blowing right on the cards. So I dont think its a thermal issue. The scores I posted show that I am getting single card performance in SLI multi-rendering modes except for the Alternate frame setting. I guess I have kind of resolved some issues with the 84.56 drivers, but I am still not sure why I experience the problems I described with ANY other driver that is more current. Maybe I should take this to a driver forum, as I am thinking it doesnt have much to do with my mobo. Thanks for the responses.
  2. Sorry, I thought I made it clear that I have tested each card individually. I have installed one card at a time in the top pci-e slot, and each of them run just fine. I have also tested where both cards are installed, and I plug the monitor into the bottom card, and run single gpu rendering. When I run tests with the option to show that green line (the division of how both cards divide the screen rendering) there is no line, because only one card is rendering the entire screen. All of these "individual" tests work just fine, no problems at all. The thing I cant test with only one card installed are the various SLI Performance Modes, which is where I am having a problem. I also stayed up late last night testing all different kinds of settings. I discovered that with the 84.56 drivers (which are the only drivers I have tested so far that do not produce the colored electricity in 2D and 3D), I can enable all of the SLI Performance Modes without getting that colored electricity, however, here are the following results: Game: Company of Heroes SLI Single GPU = 28 FPS Average (no green line) SLI Multi GPU Split Frame = 29 FPS Average 58 Max 10 Min SLI Multi GPU Auto Select = 27 FPS Average 61 Max 11 Min SLI Multi GPU (This option causes undesirable flashing or flickering of textures on the ground)*** Alternate Frame = 49 FPS Average 101 Max 10 Min SLI Multi GPU (This option caused less undesirable flashing or flickering of textures on the ground, but still present) Alternate frame 2 = 29 FPS Average 64 Max 12 Min SLI Multi GPU SLI Antialiasing (some undesirable flashing or flickering of textures) = 27 Average 61 Max 11 Min So, of course, I would love to run the Alternate Frame mode, without the flashing textures of course, lol. Oh, just ran the video stress test for CS:S (which is the same as HL:S I believe) and I can turn all the audio and video options all the way up EXCEPT for Antialiasing. When I select either 2x or 4x, I get purple coloring all over textures. The SLI Performance Mode under the CS:S settings is set to some custom setting, and the green line that shows the SLI rendering is the bar on the left side of the screen, which is the alternate frame mode I think. Also, for World of Warcraft, there is no green bar, just white lines on the left side. I dont think WoW supports SLI, and it gets turned off on my system, no matter what option I try to enable. ***Another updated edit - I have messed with the Company of Heroes settings, and discovered that the Alternate Frame SLI rendering setting does work if I turn off Shadows. The shadows, either on low, med, or high, cause ground texture flashing. When I turn it off, there are no ground flashing (and no shadows either, lol). This allows me to turn everything else up to the max, except for the Texture detail, which is set to high, but works in Ultra although I get a message about that setting possibly causing problems on my system. So here are my stats, with everything turned up all the way, Texture on High, and Shadows off: FPS Average 49 Max 95 Min 6 I think thats what I am going to use. Shadows would be nice though, ya know.
  3. I have noticed that I have had issues since I have built my system back in April with the SLI feature. I have been playing games, believing I was actually using both my video cards because SLI was "Enabled", however, I now realize that I have been running a Single GPU Rendering, defaulting from a "auto-select" performance mode. I have gone through the instructions found in these forums, and I believe I have my SLI rig set up properly. The problem occurs when I enable a SLI Performance Mode to anything except "auto" (which defaults to single GPU) or Single GPU Rendering. Selecting Split Frame Rendering, Alternate Frame Rendering, or Alternate Frame 2 results in corrupted graphics in 2D as well as 3D immediately, and in every game I have (HL:S, CS:S, DoD:S, WoW, CoH, etc...). The graphical errors look like some kind of rainbow electricity, which moves and shimmers slightly. It is stable (in that I can run 2D and 3D applications without lockups or crashes) and I get about double the frames per second in game benchmarks as compared to the "SLI Enabled"/"Auto-Select" (Single GPU Rendering). I can run either card, in either pci-e slot, using either DVI port, in "SLI Enabled" / Single GPU Rendering WITHOUT any of these graphical errors (and easily stable in 8 hour Naxx raids in World of Warcraft). But as soon as I enable an actual SLI performance mode, my 2D and 3D applications suffer with this problem. I would think that if an individual card had a defect, then it would be made known when I run the card in "SLI Enabled" / "Single GPU Rendering", but I have tried both cards like this, in both pci-e slots, and they operate just fine individually. I have tried about 6 different drivers, and they all produce the exact same results. Any ideas on what may be causing this, and what I can do about it?
  4. OK, I thought I might fool with the bios settings in order to test out a theory. I figured that if the problem wasnt my power supply, because its totally stable once the system is booted, then there might be some kind of CPU voltage or heat issue where it is cutting off for safety or other reasons before it actually boots. I noticed that there are two CPU voltage settings, one is a "boot up" CPU voltage setting, and the other one is a regular system CPU voltage system, which is associated with the +% voltage setting. I am running 1.3v on my "boot up" CPU voltage, and 1.3v +116%, which gives me like 1.52v after the boot sequence. So, I wasnt sure if the "boot up" CPU voltage benefits from the +% or not, but I bumped it up from 1.3v to 1.325v, and my system booted right up after that. I have NOT experienced the cold boot problem since I made this small change, and I have used my system 2 weeks so far, without a single issue with booting. Hope this helps others who may be experiencing this issue.
  5. Thanks for the ideas. I am suprised that DFI has not identified the issue, and provided a solid fix for this. Is it possible that they can get a mobo having this issue, and experiment with it so that they can declare a solid fix. This problem is anoying enough to bring me to post on this site, however, not bad enough to justify tearing my system apart and skipping my World of Warcraft raids.
  6. Hey, My build has been excellent so far, excpet for a booting problem now. Whenever I cold boot it, the system powers up for about 2-3 seconds, then shuts off. I cannot boot until I either unplug the power cord, or reset the power supply rocker switch. This happens about 90% of the time, and its getting old. I contacted my OCZ about the power supply, but they said that its most likely my mobo, and that this mobo is known for this problem. I have searched the site for this issue, and only found that if you clear your cmos, and reboot, you should be fine. I cleared my cmos for at least 1 hour, rebooted with default bios settings, and problem still exists. Oh, and once it does boot up, it runs fine. I can play 10 hours of games without a problem. I only experience an issue with this after I shut down the system, and try to power it up again. Please help.
  7. Habanero

    Boot Problem

    I am having a similar issue with my system. My system is a few months old now. Things have been working great, until about 3 weeks ago. I pressed the power button, and everything turned on, for about 3 seconds, then it all powers off. I repeat this about 2-5x, then it will boot, and run stable for as long as I have my system running. Then, things will work fine for a few days, then I press the power button, and I get the same problem. I was wondering if it was a heat issue with the CPU maybe tripping the high temp bios emergency shutdown option, because it has gotten a lot hotter where I live in the last 3 weeks. If its not this, then I would assume it has something to do with the PSU, because I hardly get to the post screen when it shuts down, so its not Windows, and probably not my other items of hardware. So why would something cause a shut down upon boot, yet run stable for hours and hours without an issue?
  8. Alright, I started having problems with some patches. I was getting errors about corrupted files, and I wasnt able to unzip some zipped patches. I was worried about my system not being stable, and somehow corrupting data. As I did some research, I discovered that others who had the Nvidia firewall had the same issues with downloads. I was running the Nvidia firewall, and liked the features, however, when I disabled it, the downloads worked just fine. The more I researched this firewall, the more I learned of the kind of problems it can cause your system. I was lucky enough to actually get it to uninstall, and I just ran my system for 7 hours without a hiccup in World of Warcraft. Now thats what I am talking about! So now I will be testing for a higher overclock, and I just might be able to access some more mhz because that dang Nivida firewall isnt messing up my system anymore. ***I just changed clock settings to 9x285 (2565 mhz), with memory settings of 3-4-3-8, with my video cards clocked to 634x895(895x2=1790mhz on the memory). With these settings, I scored 11,309 on 3dmark05! Way happy outcome! "UV ON, TWEAK FUN!" lol.
  9. I just updated to the forceware drivers from 84.43 to 84.56. I still get some messed up 2d pictures on websites, but I just ran 7 hours of World of Warcraft without a single problem! Ah, I am so happy now. I hope this continues to be stable!
  10. I am using the most updated WHQL video card drivers, and I am using the chipset drivers that came on my disk. I think they are pretty recent compared to whats available. It has been my experience that the video card drivers update like weekly, but the chipset drivers update like 1x every 3-6 months. Um, sometimes my system performs awesome, and just when I think its stable, it crashes in World of Warcraft. I need to keep messing with settings to figure this out. Still stable in everything else though! Lovin my overclock, and if I can only get the isues resolved with WoW, things would totally rock!
  11. Ok, Things look much more stable during play. However, the only problem I have now is during transitional load screens. I went through 6 of them this evening, and it froze up on 1 of them. Not bad for 6 hours of play. In game was great, no problems at all. What I have messed with is my SLI options. There are some settings that will adjust the way the two video cards renders video. When I try SLIx8 AA, things look amazing, however, I get cut down to about 20-36 frames per second, which is unacceptable. When I try "Alternate frame rendering" things work pretty good, like 35-75 frames per second. I get the same results with "single GPU rendering" so I am using the alternate frame rendering. I just need to figure out what that one little glitch is, when it just hangs during the loading screen. Other than that, my system is really solid! I am wondering if really doesnt have anything to do with my BIOS settings at all, maybe its a video rendering issue! Thanks for all of your help! Maybe I will try turning off the antivirus, to see if that stops the freeze in transitional load screens!
  12. Hey, thanks for the ideas so far: I have changed my preamble to 6.5, and am currently testing it. ***EDIT*** I just ran it for 30 minutes with no problems. I will continue testing. If this setting works, I will be really stoked. Then I will start to tighten up some of the settings one at a time to find performance along with stability. I am running AVG antivirus, however, in the last 9 months of playing WoW, I have always ran it, and never had problems like this. So I am not sure that I would have problems with this on my new build running this antivirus. The SLI issue could be part of the problem. I know on my last system, I ran a XFX 6800nu, and if I had it overclocked too far, I would get the same memory error and kicked to the desktop. I got this error on my new build before I started overclocking my current SLI cards. I will need to look into the SLI issue. I never had a problem with my internet connection. I am running a 3k up 856k down DSL line through Verizon. I am only like 100 miles from the WoW servers, and usually have excellent ping times. Currently, they are running in the 30's to 50's usually. Also, when I have this error occur, nobody else in my guild has the problem, so then I suspect that its my stuff that is contributing to it. When I experience bad lag, it usually corrects itself, and I can get back into playing with no problems, but this error never self corrects, and can only be dealt with by alt tabing out and closing the application. I tried deactivating all my mods, and I dont think thats part of the problem. I never have problems at the character selection screen, or the initial load screen, only when I transition to instances, or when I die, does the load screen stick at times. I notice a significant difference in the frequency and intensity of problems as I adjust my settings, so for now, I believe this is where I need to focus. I just want to make sure that there is a sweet spot in the adjustment of settings so that I can actually achieve the stability I am seeking.
  13. I believe my memory is getting appropriate cooling. I have 2 80mm case fans in the back of my case, 3 80mm fans in the front, and I installed a 120mm Sunon in the side panel that provides 1/2 of its airflow to the video cards, and 1/2 of the airflow to the memory. Also, my Zalman hs/fan blows the exhaust right over my ram, and to my suprise, its not hot air! My temps are running really cool, for my cpu, case, and chipset. I am starting to wonder if my system really is stable, but World of Warcraft just doesnt like my particular overclock settings. I have had a game in the past that didnt like my cpu mhz, because when it detected my system settings, it wasnt a standard mhz, so it didnt know what to do with it. If my memory was heating up and causing errors, I would see the instability in the memory testing, namely memtest86 or prime95, right? Thanks!
  14. I have been tweaking my system for about 7 days now. This has been the most intensive build to date, and includes the possibility for the sweetest overclock. I have found my system to be stable in everything except for World of Warcraft with the first set of settings listed. I did about 2 hours of memtest 86, prime95 for 3 hours, 3dmark05 repeateadly for 1 hour, and CS:S / DoD:S each for 3 hours with all video setting set at max. After all these tests, I played World of Warcraft, and during the 5 hours I played, I experienced 3 program lock-ups during the loading screens (I am not sure what is being stressed during the load screens, I imagine that my memory is being cleared and loaded with the new area content, but I think its mostly hard drive access, anyone know what is being stressed here that would cause instability, so I can pinpoint this), and was kicked to the desktop 3x with memory errors. So, considering this is not stable, I attempted to adjust my settings. I lowered my CPU/FSB to 6, and World of Warcraft still errored. Then I loosened many of the other settings, and those can be found in the 2nd set of numbers. What I found was that as I changed settings, the more unstable World of Warcraft became, the lower the frames per second were. It seemed to get stuck at 36 frames per second, and when it was like that for a while, I would experience a load screen freeze, or it would kick me to the desktop. However, as I continued to make changes, and the errors decreased, my frames were higher, like in the 50's - low 70's, and I experienced a significant decrease in errors, however, what that looked like was not a kick to the desktop, but rather some kind of looping effect, where all other characters were like running in circles, and I could not mount, or talk to NPC's, but I could still run around and some functions were still available, like viewing my bags and settings. I am finding that because there are so many settings and options, I am not sure what to change. What I am thinking is to loosen my settings as much as possible, without going overboard, to find a place of complete stability in World of Warcraft. Once I get that, then I can continue to tighten one setting at a time, and test for stability. Looking at my setup, and my bios settings, what do you guys think about my experience so far, and what I should do from here to get my World of Warcraft stable? Everything except for World of Warcraft stable: HTT/FSB: 275 LDT/FSB: 3 CPU/FSB: 9 PCI: 100 CPU VID Start Up: 1.5 CPU VID: 1.300v CPU VID Special: +121.9% (shows 1.55 in windows) Chipset: 1.7v (shows 1.67 in windows) DRAM VC: 2.96 (shows 2.89 in windows) DRAM: 200 1/01 CPC: Enabled (T1) TCL: 2.5 TRCD: 4 TRAS: 8 TRP: 3 TRC: 8 TRFC: 16 TRRD: 3 TRW: 3 TWTR: 2 TRWT: 3 TREF: 3120 TWCL: Auto DRAM Bank Interleave: Enabled DQS SKEW: Auto DQS Skew Value: 0 DRAM Drive Strength: 7 DRAM DATA Drive Strength: 4 MAC ASYNC Latency: 08.0 Read Preamble: Auto (5.0) DRAM Response: FAST IDLE Cycle Limit: 256 Dynamic Counter: Disable R/W Queue Bypass: 16x Bypass Max: 07x 32 Byte Graduate: Disable(4Burst) Everything stable and World of Warcraft stable for 1 hour so far: HTT/FSB: 280 LDT/FSB: 3 CPU/FSB: 9 PCI: 100 CPU VID Start Up: Auto CPU VID: 1.300v CPU VID Special: +121.9% Chipset: 1.7v DRAM VC: 2.96 DRAM: 200 1/01 CPC: Disabled (T2) TCL: 3 TRCD: 4 TRAS: 10 TRP: 4 TRC: 17 TRFC: 17 TRRD: 4 TRW: 3 TWTR: 2 TRWT: 3 TREF: 4708 (per OCZ website for my ram) TWCL: 1 DRAM Bank Interleave: Enabled DQS SKEW: Decrease DQS Skew Value: 0 DRAM Drive Strength: 7 DRAM DATA Drive Strength: 1 MAC ASYNC Latency: 08.0 Read Preamble: 6.0 DRAM Response: FAST IDLE Cycle Limit: 128 Dynamic Counter: Disable R/W Queue Bypass: 8x Bypass Max: 05x 32 Byte Graduate: Enable(2 burst)
  15. Habanero

    New build...

    OK, I have done some more tweaking, and here is where I am so far. My CPU will run stable at 2655mhz with my ram turned down, and set to DRAM: 100. However, it doesnt like any other divider other than 100 or 200, nothing inbetween. At HTT/FSB: 295, my memory cant handle that at 200, or any other divider other than 100. That kinda sucks. So I altered my approach, and tested what my memory could do at the DRAM: 200. It looks like my limit with current settings is about 275. At 280, bad things happen. So at the 275x9, my CPU is at 2475, which is 200 mhz below what I know my CPU is capable of. A 9.5 multiplier would be really handy right now, lol. So what other options do I have at this point? Also, at 275, my LDT/FSB is at 3, which gives me 825x2=1650. Thats 350 below where I should be. If I used 4, I would be at 2200, which would put me at 200 over the 2000 mark. Any recommendations about this would be useful. I am really happy with the stability of my current settings, however, I want to make sure that I am pushing my limits as best as I can. I am memtest86 stable for 30 minutes, and I am prime95 stable for 30 minutes so far. I scored 10,790 on 3dmark05, which is very respectable, and I have not touched my video card clock speeds yet. Any advice as to where to go from here would be appreciated! ***EDIT*** I just adjusted my HTT/FSB to 280, my DRAM DATA Drive Strength to 1 and my MAC ASYNC Latency to 08.0, and its stable so far! This gives me 2520mhz! UV on, Tweak Fun! HTT/FSB: 280 LDT/FSB: 3 CPU/FSB: 9 PCI: 100 CPU VID Start Up: 1.5 CPU VID: 1.300v CPU VID Special: +112% Chipset: 1.7v DRAM VC: 2.8 DRAM: 200 1/01 CPC: Enabled (T1) TCL: 2.5 TRCD: 4 TRAS: 8 TRP: 4 TRC: 12 TRFC: 16 TRRD: 3 TRW: 3 TWTR: 2 TRWT: 3 TREF: 3120 TWCL: Auto DRAM Bank Interleave: Enabled DQS SKEW: Auto DQS Skew Value: 0 DRAM Drive Strength: 7 DRAM DATA Drive Strength: 1 MAC ASYNC Latency: 08.0 Read Preamble: Auto (5.0) DRAM Response: FAST IDLE Cycle Limit: 256 Dynamic Counter: Disable R/W Queue Bypass: 16x Bypass Max: 07x 32 Byte Graduate: Disable(4Burst)