Jump to content

David A. Leon

Members
  • Content Count

    48
  • Joined

  • Last visited

About David A. Leon

  • Rank
    New Member
  1. Yep, I would say that is a definet possibility. I think I will flash up to the latest BIOS rev and see what happens. My temps never hit 37 degrees Cel under the old Zalman 9500. They are higher now under the Tuniq. I may just have a super-sample Opteron, or it may be because I run with my case open, or it may be because I built this system in December and it is still cool-to-cold here in April. However, I suspect it will get a lot hotter when the ambient room temps reach 100F in Jully-August.
  2. Okay, I found the problem and the solution. The fan Axel on my Zalman 9500 is bent and warped, causing the fan to jam against the copper blades. The fan is designed to quit rotating if it encounters resistance. Hence the Zalman 9500 turned into a completely passive heatsink. It's incredible that such a heavily overclocked chip kept going given a passively cooled sink. This is a testamony to AMD and Zalman. In any case, I better not heap too much praise on Zalman because I believe I have detected a serious design flaw here. The metal supporting the fan is cheap tin; weak and flimsy; easy to bend into and out of position; far too weak to ensure that the fan will stay in its correct place if bumped hard. Further, the fan is fully exposed to impacts. I replaced this unit with a Tuniq Tower, and I am getting good results. The Tuniq tower seems to be optimized for silence over ultra-cooling (the oposite of Zalman 9500) but it seems to keep the processor locked at about 40 degrees, come what may, and I can't hear a thing. It is totally silent, even during hard rendering and H.264 encoding. There is no chance this beast can suffer the same problem as the Zalman as the 120mm fan is housed inside the body of the sink. The fan is protected by the blades around it. I'll let you know if there are any problems with the Tuniq, but it looks solid.
  3. Up to the present moment just about all NForce boards ship with CMedia ALC850 audio chip modules. While the audio quality from this chip is not bad, it is not particularly thrilling either. I would like to place an official request for the latest and greatest CMedia chip: The Oxygen CMI8788. I was just reading about this puppy here: http://www.cmedia.com.tw/product/product_pci.htm It seems that this chip was designed to compete head-to-head with Creative Lab's X-FI and perhapse even win. I would love to see this on DFI's crop of NForce-5 AM2 motherboards. It would add a lot of value to the products. I don't need to tell you or anyone else about the popularity of digital music these days. Most kids don't have a stereo anymore. They just have computers. Why not give the market what it realy wants and stick a finger in Asus's eye?
  4. At the start of the year I built a DFI NF4 Expert system powered by an Opteron 175. The Opteron is cooled by a Zalman 9500. At first I got astounding results from this combo. The chip ran at around 27 to 31 cel, despite being overclocked to 2.75GHz (and over-charged to 1.525 volts). Stability was perfect, passing all the stress tests with flying colors. Just recently (in the past two days) I have noticed that my Zalman sink has been getting hot for the first time ever. During marathon renders (11 hours long) the temperature spiked up to 75 Cel, which is outrageously hot. It never did this in the past. I could marathon render for 2 days with temperature holding around 33 degrees the whole time. The weather hasn't gotten any warmer around here either. It has been raining like hell in L.A. for the past two weeks with temps in the 50f-60f degree range. I am at a loss to explain why the cooling has broken down, or why my Opteron is super-heating. According to one doubtful story, Zalman's base plate will warp with excessive heat, causing a break in surface contact and then super-heating. According to another doubtful story, Opteron's don't like to be overclocked to this speed, so the wear out faster and begin to super-heat. Anyone have any NFO to share on this question?
  5. Where did you get this 12/7/2005 BIOS? I just checked the DFI website and the top BIOS was dated 2005/11/02? If you have a link I would appreciate it. http://us.dfi.com.tw/Support/Download/bios..._FLAG=A&SITE=NA
  6. You are so right! It is the Denmark core, not the Italy core. I'll have to change my signature. You guys are confirming what I thought was true. We do seem to have problems with the temperature readings on the Expert. I guess we will have to kick it around philosophically, but I think the high reading is more favorable than the low reading. Given a high reading, the fans may pump a little harder, but there's no harm in that. If the reading is too low, the fans won't pump hard enough. There could be some problems there.
  7. I put my Opteron 175 stock cooler on my sister's Athlon 64 3500+ venice core. It works great. Her max reading so far is just 37 cel under maximum stress load.
  8. The official dual core Opteron thread at Anandtech.com seems to indicate that 1.6 is the maximum voltage anyone has used successfully. I am cranking at 1.525v myself with spectacular results. My rig is perfectly stable at 2.75GHz. Somewhere between 1.5v and 1.6v is the promised land of maximum overdrive.
  9. Howdy, Just thought I would tell everybody that I am mostly thrilled with the DFI NF4 Expert and the Opteron 175, and the Corsair TwinX 2048 3500LL, and the Zalman 9500, and the OCZ PowerStrem 600 Watt. I have been able to achieve 100% stability at 2.75GHz (FSB=250, CPU multipliper=11 , CPU voltage=1.525v, Ram Frequency=500MHz, 3-4-4-8, 2.88v. All the benchmarks run to completition and I did a 10 hour load test last night with the Following: (1) Prime95 (2) HotCPU Tester Pro (3) CPU Stability Test 5.0 by Juoni Vuorio (4) SuperPI All of them were run together, simoltaneously for 10 hours. The system was fine this morning. All tests shutdown cleanly, and reported no errors for the duration. Now for the preposterous part: My CPU temp is reading 24 Celcius right now, which is equal to 75 degrees Fahrenheit. During the load test, my temperature never rose above 38 degrees Celcius, which is equal to 100 degrees Fahrenheit. Now, it is winter time, I am using an ultra-high capacity heat sink, this is a great board, a great processor. It is not particularly warm in my house, but I do live in Southern California, so the temperature is not particularly cold either. It might be 64 degrees minimum inside my room. It might be a little higher. Can this be real? The temperature readings just seem too damn low for such a high overclock and such a heavy load test. I trying to confirm or refute these temperature settings by various means (thermometers and my hands) and it sure doesn't seem like the air coming off the sink is very hot. The board doesn't feel hot either. Does anybody know of heat sensor inaccuracies with the Expert board? Is this a problem that anyone has reported. If I had to guess, I woudl guess that my sensor is inaccurate and I am 10 degrees low.
  10. Thanks for that tip. Any ideas of how to enhance stability at higher speeds?
  11. The Audigy 2ZS is pretty passe. I wouldn't bother with it. I had one in my dual Xeon rig before building my first X2. The audio quality provided by the default sound chip is extremely good, and quite comperable to the Audigy 2zs. The one problem I had was a little snap crackle, pop and hiss during overclocking. This eventually spurred me to buy an X-FI, which generates no noise at all during any opperation. If you are going to spring for a new audio solution, go all the way and get an X-FI. This kicks the crap out of the Audigy 2ZS. Before X-FI arrived, many experts were confidently predicting the death of Creative Labs due to the fact that Audigy 2zs was no longer had any technological advantages, and was quite expensive. X-FI is the product which restores the technological lead to Creative Labs.
  12. Ah ha! That's what I thought. I had this problem once before and an interpretation of the beeps indicated that I had an illegal ram arrangement. It turned out I was putting two sticks on two different color slots. Very stupid mistake. I slapped the hell out of myself. I couldn't believe I did it. I knew the rule with two sticks was simple: "Pick a color, just one color, any color".
  13. Howdy. Just built my new system last night. It is similar in most respects to my last rig but I rejiggered some of the parts to ensure better overclockability. So far, I am only having marginally better luck than last time. These are my settings: FSB=250Mhz HT Multiplier=4 CPU Multiplier=11 Memory divisor =1/1 (100%) Loosened latency=2.5-3-3-7 Ram voltage=2.88v, CPU Voltage=1.45v At the moment I have system almost running stable at 2.75GHz. I have been able to get into Windows XP without much problems. Some benchmarks run to completion successfully. Some, like FutureMark's, bomb out. Some apps run well. Some lockup or crash. What pisses me off is this: I have tried several times to lock my CPU voltage at 1.5v (I have a hunch that with just a tad more voltage everything will be just fine) unfortunately, each time I boot up, nTune reports the voltage has been down-shifted to 1.45v. How can I prevent this from happening? Why does this board seem to have a mind of its own regarding CPU voltage? P.S. My PSU is the OCZ PowerStream 600Watt.
  14. Alright... I have decided to bow down to the experts here at DFI-Street.com on the off-chance that you might be right about this and Tomshardware.com might have an unfortunate economic interest in Seasonic. The order for the OCZ PowerStream 600 has been placed... (yikes!) I'm feeling like Grand Moff Tarkin (Peter Cushing) in the original Star Wars when he turns to Darth Vadar and says "I'm taking an aweful risk Vadar!"
  15. Ah-ha, I see... That's bad. Seasonic has admited to the problem and posted recommendations. That's damning evidence. I don't want any piece of those problems. I'm still leery of going with OCZ. Can anybody tell me why the OCZ Powerstream 600 blewup during the Tomshardware benchmark test? http://www.tomshardware.com/howto/20050711...esstest-20.html What do you guys think of the Enermax Noisetaker EG701AX-VE(W)? http://www.tomshardware.com/howto/20050711...esstest-14.html This one performed very well on the stress test, although it has a bit too much ripple for my blood.
×
×
  • Create New...