Jump to content

Bullhonkie

Members
  • Content Count

    26
  • Joined

  • Last visited

About Bullhonkie

  • Rank
    New Member
  1. That comment brings back memories of me spending almost an hour trying to squeeze or angle in a 120mm fan in past that assembly. I gave up. How did you manage it? I'd love to know too. I've had no luck trying to fit a normal 25mm thickness fan in that area. Right now I have a Zalman 80x15mm rigged up there. But I'd much rather use a 120mm fan for greater airflow at lower noise levels.
  2. I've got a Opteron 165 CCB1E 0608MPMW, supposedly one of the better steppings that can break 3ghz. I've read this stepping is coldbugged and tends to run hot though, so watercooling is usually needed to extract all of its potential. Unfortunately I run air cooling only and mine doesn't seem quite as good as others that I've seen. For 24/7 usage I run 2.6 at 1.3v. 2.8 takes 1.475v and 2.9 with 1.55v. I'm not really comfortable pushing voltage beyond that, so I doubt 3ghz is practical on air cooling for this chip. I'm happy with 2.6 on a slight undervolt to 1.3v though and as long as you're happy with it, that's what matters right?
  3. I was able to get my 7900GT in the upper slot with a VC-RE. The very bottom of the PCB seems to just barely touch one of the fins on the VC-RE, but it works.
  4. I like what you did with the Zalman NB47J on the chipset. I tried to do something similar but because of the length of my 7900GT and the way that capacitors are positioned on the PCB, I couldn't get pins bent far enough to give it full clearance into the upper PCI-E slot. It's a shame because I would've loved to rig up a thin 80mm fan in that area like you did. Instead I had to go with a VC-RE which is a little noisier than I'd prefer.
  5. That's been my goal with the P180 as well. I think it's the best set up case out of the box to build such a system. Mine is running the stock tricool fans all on medium and further undervolted on a Rheobus. Yate Loon D12SL as intake on the upper HD cage and another strapped to the Ninja HS on the CPU, both doing about 6-7v. Lower cage has a Zalman 80x15mm fan at 5v in front of it to help cool a Cheetah 15k. Overall it's fairly quiet for the hardware in it. I've put together quieter systems but they all had much lower thermal loads to deal with. I sit right next to it and it's lightly audible but never bothers me. At this point the loudest things are the hard drives. It becomes inaudible by 3-4m away depending on the ambient noise level (higher in the daytime).
  6. I saw little improvement (2-3C at best) going from the stock chipset HSF to the VC-RE. It idles at 48C with room temp of 28C. Before I switched it would be 50C+ regularly. However I replaced it more for silence than temps - the fan on the VC-RE only needs to spin about 2800rpm to keep the same temps as the stock one did at 4000rpm. I have a 7900GT with VF-900 on it dumping its heat pretty much right over the chipset. Also I have 2 HDs installed in the upper cage in my P180. All of this results in a big dead spot around the chipset with very little breathing room because of everything surrounding it. I've thought about trying to rig up a fan somewhere in the vicinity to cool the chipset and PWM area, but there's very little room to work with and I cringe at adding more fans since it'll end up introducing more noise.
  7. I got 10899 so far with mine (see sig for link). I'm sure I could break 11k with 1.4v but I'm too chicken to do it. I need my parts to have good longevity because I hand down my used stuff to family so things usually see usage for 3-4 years.
  8. That's completely fine IMO for being under load. My PWM also experiences relatively high temps during load because of the huge dead spot between the Ninja HSF and mobo in a P180 case. PWM idles around 37-38C but can go as high as 59C on load. I've had no stability issues though - just passed 40+ hours of dual Prime in fact.
  9. I was just curious because I've read of some CCB1Es exhibiting cold bug like symptoms on normal air cooling. Not sure what that means exactly though. My core 0 runs like a champ. Was prime stable for the one hour I tried it at 2952. Core 1 starts crapping out on me as soon as it hits 45-46C every time at any speed/voltage. I never hit those temps until I get to 2800/1.475v and above so I don't know how much can be attributed to the temp or just the limit of that core.
  10. http://www.bullhonkie.com/img/desktop/desktop_stress1.jpg I'd say that's pretty much stable. How 'bout you?
  11. Mine behaves pretty much the same way, although it is a 165 CCB1E 0608MPMW. Can do up to 2.655 at stock volts, 2.7 1.4v, 2.792 1.475v. Anything higher and core #1 seems to fail as soon as temps hit 45C or so (is this what's known as the cold bug effect?), which happens pretty quick at 1.5v and above. Watercooling definitely needed to squeeze everything out of this one. For 24/7 usage I'm sticking to 2.6 at 1.3v.
  12. That's what I had been doing with SP2004, but I ran into the temp file error when trying large FFTs. After running two copies of SP2004 from two separate directories, that no longer happens to me. And as an update to the original post, 1.275v ended up erroring after 10 hours of SP2004. Now I'm doing 1.3v and everything is going well so far at 9 hours.
  13. I did 1.3v on mine and can get up to 645 core. I always bump it down 5 for benchmarking and another 10-15 for 24/7 usage. So my startup settings are 620/850 but I've been able to run rthdribl at a max of 645/870 with no visible artifacts. I'm kind of tempted to try 1.4v now but at the same time I'm kind of chicken. A fellow on the Anandtech forums can do 700 core at 1.4v and I think my card has a chance as well. I was able to hit 585ish max stock, so 0.1v bought me 60mhz on the core. If it scales linearly to 1.4v (which is no guarantee) I'd have a good chance at 700.
  14. I've been testing stability with other programs as well as using the system normally and have had zero problems otherwise. Sometimes I'll run rthdribl and one SP2004 to stress both cores along with baking the video card. After I set up SP2004 to run from separate directories, I no longer get the temp file error when doing large FFTs. But since I got it using small FFT more than 10 hours into testing, it still makes me suspicious. I'm finished exploring my upper overclocking limit on this chip for the time being. Now I plan on running 2.5 or 2.6 for 24/7 usage and I'm looking for the lowest stable voltage at those speeds.
  15. I was dual priming small FFT overnight last night to test 2.6ghz at 1.28v and woke up today to a 'FATAL ERROR: Reading from temp file.' I've gotten this error in the past when I run two instances of large FFT. One of them always gives this error within a few seconds of starting the other. I've been using two instances of SP2004 from the same exe in the same directory. Now I've been reading that probably wasn't the right way to do it, which is probably why I always got that error immediately on dual large FFT - even with everything at stock settings. I'm used to seeing rounding errors or illegal sumout when I encounter instability. So I'm not sure if this temp file error was due to any actual instability or just a configuration error - my incompetence of running them from the same directory. I'm going to redo it tonight anyway, running them from two separate directories. But I wanted to see what your experiences were? Did you ever see the temp file error due to instability?
×
×
  • Create New...