Jump to content

Fresh Daemon

  • Content Count

  • Joined

  • Last visited

About Fresh Daemon

  • Rank
    New Member
  1. Neither is accurate. The onboard power measuring circuitry is cheap and crappy and MBM5 just queries it. Plug a multimeter into a Molex plug, black into black, red into red, and that'll give you your true 12v reading. Within 5% either way is acceptable.
  2. 45-50C is perfectly acceptable for the chipset. It's not that. What is the last thing you do before it restarts? Try backing off on your overclock. I thought I had a stable overclock (48hr Prime stable) and then I installed BF2 - guess again. Crashed after 10 minutes in the game, backed off the OC, ran perfectly. Try making a second hard drive partition, install another XP and see it if runs perfectly. This will eliminate a software issue without necessitating a full reinstall.
  3. Get a live Linux CD (Knoppix, Slax, etc) and boot it. See if you can read the disk at all. If you can see the files odds are you just need to repair your Windows install. If not, looks like reformat time.
  4. The prongs exist to make sure there's constant contact between the plugs and the casing. IIRC it's a measure for RFI shielding. Seeing as how there are people who run their computers in wooden cases I doubt it's important.
  5. It depends on the resolution and the game. Here's an article: Firingsquad Oblivion tests. In games such as FEAR, Oblivion, Quake 4, Doom 3 etc. it is mostly GPU limited at 1280x1024 and above, especially with AA/AF. For instance, Quake 4 at 1600x1200 4xAA 8xAF shows exactly an 0.8 fps difference between an FX-57 and a lowly 3000+. Below 1280x1024, CPU does make a difference, as well as in a few games. HL2 is CPU-limited at much higher resolutions than the above-mentioned games. An FX-57 shows a sizeable edge over a 3500+ even at 1600x1200, however, in both cases it's well over 100FPS and you'd never be able to tell. This is why I'm running dual 7800GTX 512's and a Venice 3000+.
  6. I see you're in Canada. Try this: NCIX. It works on my Ultra-D.
  7. What voltage? Does it go higher if you relax the timings? To be honest, I wouldn't get too upset over it. I know it seems like you're missing out on performance, but you really aren't. Tight timings are better than high speed, but neither has a huge impact on much.
  8. Try the 6.23-2 BIOS. AG says that it's best for broadest RAM compatibility. See if you can get it running, then perhaps another version might give better results. I don't think A3 is "bad". I've heard people bitching about various board revisions but I've yet to see anyone actually test these theories. As far as I know, about the worst you could get regarding board revs would be one of the Ultra-D's where they epoxied over the bridges, so it can't be unlocked.
  9. Finished testing the RAM. It ran at 250 with stock settings, RyderOCZ said it was Mosel or Infineon, so that's pretty good at 2.5-3-3-6. At 2.5-3-2-5 it doesn't tolerate much above 200 and boosting voltage doesn't help a lot. I should benchmark [email protected] and [email protected] and see which is better. Next is the CPU. 2250MHz is SuperPI stable at stock voltage. Still working on it...
  10. That's true, SLI changes the rules a bit. The only way to be sure is to benchmark. Don't bother with 3DMark, just take the games you'd like to play at the resolutions/eye candy you'd like to play them at and run timedemos or FRAPS, measure the performance increase your overclock is getting. If it makes 3-5FPS difference, meh. If it makes 20FPS, then you've got a discussion.
  11. LOL, I wish! I lowered the multiplier. The CPU was running at 2GHz during that test. I just wanted to see what the HTT ceiling was.
  12. What speed are you running your RAM at? My RAM gets flaky if it's run much below 170MHz. Try and keep it close to 200MHz if you can. Check your RAM timings. My DFI set them tighter than spec and as a result, it hit a ceiling around 250HTT. Once I'd manually set them to the specs for that RAM, it was plain sailing all the way up to 400HTT and 250MHz on the RAM.
  13. DFI says that all the board revisions are physically identical, the only difference is the shipping BIOS version. Have you tried flashing the BIOS? No way to tell which would be the best, but try 6.23-2 for a start.
  14. It depends on what you're doing. If you're a gamer (and assuming by your SLI 7800GTX's, you are) then be aware that modern games are bottlenecked at the GPU, especially at very high res and eye candy, which is why SLI gives good results. Therefore, overclocking past a certain point doesn't really get any extra FPS. So I would leave it at a sane vcore, keep the temps down and the life of the chip up. A64's aren't bandwidth starved and don't suffer much from higher latency. Once you've got PC3200 with half-decent timings, anything else isn't likely to make even 5% performance difference in most apps. The exception to this seems to be compression/decompression apps, which like loads of bandwidth for some reason and will see decent performance gains. Hope that helps!
  15. I would say so. I tightened my memory to 2.3-3-2-5, and I could get 285HTT with 200 actual RAM speed, or 295HTT with 177 actual RAM speed, but nothing higher. Once I relaxed the timings to the memory factory settings (2.5-3-3-6), I could get HTT up to 400 and the RAM speed to 249. This is stable. I'm going to see if I can clock the RAM any higher, but yes, I would see if you can relax the timings.
  • Create New...