Jump to content

Smallies

Members
  • Content Count

    55
  • Joined

  • Last visited

Everything posted by Smallies

  1. aaaaah.... i see, that makes plenty of sense. I've been trying to figure this out for quite some time
  2. I've noticed something quite peculiar with my corsair TwinX2048-3200C2 memory: when I run at 200mhz stock speed with all timings manually set to 2-3-3-6-1T (as well as ALL other memory timings, such as read preamble etc.), the system is rock stable and works fine. Memory in orange slots. If I run at 237mhz FSB but use a memory divider of 9/10, which results in exactly 200mhz on the memory, the system refuses to run memtest86 / boot into windows with the exact same memory timings - keep in mind, every single memory timing is manually set, nothing is set to auto, so the memory should in effect be running at the exact same timings as when the FSB was set to 200mhz. If i drop to CAS2.5, and change nothing else, it works like a champ. I find this odd considering with the memory divider in place, the memory is running at the exact same timing. Anyone have any ideas as to why this is happening?
  3. Mmm.. take out your audigy ZS, use teh onboard sound. See how it works. I'd bet a bazillion dollars you have your vid cards on teh same IRQ as your sound card and its causing stuttering... I too have an audigy, and it is using the same IRQ as my vid cards, but doesnt stutter. Thats not to say that it couldn't though.. only my vidcards and soundcard are sharing the same IRQ, but i've built systems in the past that had the ethernet, sound, vidcard, serial port and a billion other things all on one IRQ. Swap PCI slots for your soundcard as well. For reference, my audigy is on the bottom pci slot. If that doesn't work, take EVERYTHING out of that computer except 1 vidcard, 1 stick of ram, the cpu, and the mobo itself lol. Run 1 hard drive and 1 DVD-rom/Cd-Rom as well. Turn off all your network/on-board sound. Do a fresh install of WinXP, install SP2. Install NOTHING else, no anti virus etc. etc... Write down every single IRQ/IO address etc. that is listed in device manager by each device. now install a game, and play. If it doesn't stutter, ad 1 component at a time back into your system until it stutters again. Once it does, remove the last piece of hardware. See if it still stutters. If it doesn't, problem solved. If it does, reference your IRQ/IO list you wrote down. See if anything changed (since it stutters even with the offending piece of hardware removed). That should be a good starting point for now P.S. im building a system for my 8 year old cousin (lol, luckiest little . ever. what 8 year old has a 7800GT and a 2.6ghz AMD64 lol, love that little guy - yes, he lives with me) - but point is, DFI boards are not that bad at all for non-overclockers. The nice thing about DFI boards is even if something doesn't work, there are so many settings that you can change to eventually get it to work. Some mobo's out there if it doesn't work, you're screwed, because it's all set "automatically" - sometimes never allowing you to specify the CORRECT setting yourself. Just a thought, but good luck mate. P.S. 2 MAKE SURE you don't have programs like motherboard monitor etc. installed... they poll the system every second causing this stutter issue.
  4. something interesting to think about... adding a heatsink to the PWM IC may make it cooler.. but the sensor isn't even directly on the PWM IC. Surprising that it actually drops the temps, since the sensor really is just an "area" sensor, not specifically for the PWM IC like say a CPU sensor.
  5. Well, i was having an issue with my PWM IC temp hitting 73C after I installed my X2 4400+ and OC'd to 1.52v @ 2585mhz. I ended up putting a 80mm low speed fan blowing directly onto the mosfets to the right of the CPU and now the PWM IC is running at 54C under load. 38C just futzing around right now in windows. Point is, I can OC a lot easier now, ive prime95'd for 4 hours on both cores (one small FFT, one large FFT) and ran 3dmark etc., played games for hours, its stable as can be so far. I've got some OCZ BGA ramsinks laying around, i'm thinking of just putting them on any exposed chips on the board lol... anyone know any particular problem chips that generate a lot of heat? In particular, does the actual PWM IC (not the mosfets) generate a lot of heat? I searched and found one guy saying it does, but i'd like to know if anyone else has heard that. I am thinking the reason why my PWM IC temp has went down so much is the airflow is merely hitting the temperature sensor, not actually cooling the PWM IC (pulse width modulation integrated circuit). So thats why I'm thinking of putting the ramsinks on the actual chip above the 4-pin power connector (the long rectangular one).
  6. well i jimmy rigged a very low speed 80mm fan sitting above the PWM IC's, its generating enough airflow to keep the PWM IC below 56C now with both processors at full load. I imagine as the room gets hotter it'll rise to around 60C, but thats a far cry from the 73C it was yesterday. Now to see if thats whats keeping me below 2.6ghz.... this thing boots at stock voltage at 2.8ghz so i know i can get more out of her than 2585 heh...
  7. positive there's no air bubbles, i moved the block back and forth on the chip to ensure there would be none, just as arcticsilver.com recommends i also fully tightened down the waterblock correctly now, it was a bit loose, but still no difference really.
  8. nah they're clear of all wires etc... just after running prime95 for an hour the PWM IC reached 73C.... veeeeery friggin hot...
  9. believe it or not, ym case has great airflow. CM Stacker, 2 120mm intake at the front, 1 120mm exhaust at the rear (all panaflo's, low speed, around 67cfm i believe). 2 x 80mm exhaust at the rear 1 x 80mm exhaust at the top of the case. 1 x 120mm low low speed blowing on the 7800GTX's and chipset (lowers chipset temps like 15C, same deal for gtx's) all the fans are low speed though, but still, it generates enough airflow i believe.
  10. Ok, got a X2 4400+ here, and I'm doing some OC'ing, my cpu voltage is at 1.55v (1.52v in ITE). @ 2.6ghz, the PWM IC is at 69C, cpu is at 42-44C. Chipset is 46C. Should I be worried with the PWM IC so high? My old venice 1.8 @ 2.6 never ever exceeded like 40-45. I do realize this is sucking a lot more voltage, but still... Anyone in the know out there who could enlighten me?
  11. Hey all, I just popped a X2 4400+ in my NF4 SLI-D, and I checked out the temps and it idles at 30-31C, running at stock voltage (reported as 1.31v in ITE/bios, seems kinda low but its set to auto). My old venice 1.8ghz @ 2.6ghz (288x9) @ 1.675v used to idle around 27C. Maybe im not taking into consideration there's two cores on this puppy, but it seems like it should of been idling even lower, no? Im thinking I didn't properly mounted my swiftech waterblock when I installed the 4400+, but I haven't been able to find many idle temps of 4400+'s with watercooling to compare against. I'm thinking I didn't tighten the screws down enough, they're just barely hand-tight right now. P.S. the venice temps were with the same watercooling setup. P.S.2 I did search, sorry if I missed any results that quickly answered my question! Thanks in advance guys
  12. My corsair 2gb works great... IF I use 7/04 bios. I also used the Ram Guy webpage on corsairs site to find out what modules my ram had, they are infineon chips. Perhaps your sticks have different chips and thats why Im having great results and others are not. That said, if I ran ANY overclocks whatsoever with any bios besides 7/04 the system was unstable as . unless I ran 3.0v - even at stock 200mhz. But with 7/04, im running 7/10 divider to get 202.3mhz on the ram, and the cpu is at 2.6ghz (from 1.8). Works great, and I haven't even bothered to push it harder yet, I bet I have a lot more headroom
  13. i have my stacker installed with a SLI-D... ??? works great.
  14. Ok I have the exact sticks you are talking about, and they are not entirely "plug and play" With the stock 1/25 bios, it will boot and run fine, as will the 3/10 bios, but you won't be able to run it with any overclocking whatsoever. If you do NO oc'ing on the processor, it will run like a champ though. But even if you're just OC'ing the processor and leaving the ram at stock 200mhz, the dividers DO NOT work with this ram. Only way to get it to work even with dividers is to raise voltage to 3.0v on the ram. That said, I've tried a bunch of bios's and none of them really changed this scenario besides the 7/04 bios. I'm using 7/04 now, and running 2.8v on the voltage (stock for these sticks) with a divider to set the ram @ 200mhz. I also am running 2.5-3-3-6-7 which is their stock specified speed. Runs like a champ. My processor is also at 2.6ghz from 1.8ghz (venice 3200+). In short, get 7/04. This ram works PERFECT with 7/04 but not as good with any other bios's. And BE SURE to read up on the 7/04 CPU voltage bug, it could ruin your processor if you're not careful (very rare, but possible). There is a modified 7/04 floating around that supposedly fixes this though. Good luck
  15. Ok, with my setup below I had my chipset running 48C idle and 57C under load. I added a very low speed 120mm fan (like 20-30cfm) that blows directly onto the chipset between my two 7800 GTX's and now I get 44C idle, 48C load. With a higher CFM fan (delta 120mm, pushing about 70 cfm) my idle temps went down to 41C. I didn't test under load.
  16. try 623 (or 7/04 if you're feeling frisky; best bios yet imo as long as you are aware of the voltage bug)
  17. to answer the question, YES two SLI cards does create a heck of a lot more heat. My cpu idles at 27C (watercooled), my PWM IC is around 40C, and my chipset idles at 48C. Under load, things shoot up to 33C, 48C, 57C. I'm adding extra cooling in a few days to compensate.. I have 7800 GTX SLI btw.
  18. Well, I understood that his was the rational suggestion, however I was expecting the AS5 to make some difference, and I started the thread in the hopes that some would have some novel ideas (such as the HeatSLink)
  19. Well I just held a 120mm Delta between the two vidcards and my idle dropped from 49C to 42C. So its definitely an issue with the 7800's creating an assload of heat and dropping it dead smack on the chipset... The question now becomes how do I mount a fan (albeit, a quieter one) so the heatsink gets ample airflow across it.
  20. well I left the house for 2 hours and let it idle, it went down to 44C. About to play some games now and see what happens..
  21. Chip voltage is stock, 1.5v. In addition, my vid cards are back to stock speed as well, i forgot to mention that. XFX GeForce 7800 GTX's running at 450/1250 (stock).
  22. Ok, I left the square foam pad. I obviously removed the center thermal pad with a razor blade and then cleaned it with acetone then applied a small dab of AS5, smeared it with a plastic bag, wiped the excess off, then put a paper-thin application of AS5 to the nvidia chipset with a credit card
×
×
  • Create New...