Jump to content


  • Content Count

  • Joined

  • Last visited

Everything posted by Vista

  1. Firefly for recent science-fiction/comedy. Can't beat Joss Whedon. The Firefly movie, Serenity, is set to come out this Fall.
  2. For $20 more, these are in stock at Newegg: http://www.newegg.com/app/viewProductDesc....-231-008&depa=0 If you go to the G.SKILL website, you'll see these are tighter than the $280 LE version by a touch. The package I got says: DDR400 2-2-2-5 F1 DDR600 2.5-4-4-8 LA My sig shows I'm running the 2.5-4-4-8 settings (2.5-4-8-4 in the order that the BIOS boot screen shows)
  3. I got mine on March 21, 2005 from Newegg, it used the old packing method and showed some impacts on the bag and one pin, but otherwise was perfect. Fingernail on the pin fixed the only flaw (we're talking about less than 5° out of true for that pin. I would have never noticed it unless I had read all these paranoid shipping damage postings before getting my board).
  4. Most nF4 chipsets seem to run hot but not care, except for the fan noise. Mine does 46-50°C, but load doesn't seem to make much difference. 60°C seems to be about as high as you'd ever want to see it go for an extended period of time. People like AG run much cooler, like 34°C, I assume because they are blowing another fan on that area and have better case vents and fans than average. Some people claim huge temp gains by remounting the chipset fan using AS5 instead of the default pad, but the conservative approach is to never mess with the chipset fan assembly since you'll break it (and probably void your warranty).
  5. From my experience with large server farms and OEM servers, I can attest to what you are saying NEOAethyr. Most pre-built systems from Dell, HP, Gateway, etc. attempt competitive but conservative settings. At best, assemblers burn them in for 24 hours and then ship. I'd much rather trust a system that has been tested and burned in as rigorously as the members of this forum! I thought I was torturing my PC before I joined, now I know the true meaning of pain!
  6. That is a refurbished Socket 754, not compatible with the DFI Ultra-D Socket 939 A64 chips! Be careful at this point, you need A64 939 for the SLI or Ultra-D. Refurbished usually means someone already hand-picked that chip as being a dud for OC at the 15% restocking fee (or they meant to get the 939 socket version). Some people are getting good results with this chip, if you want the retail box and stock HSF: http://www.newegg.com/app/ViewProductDesc....-103-514&depa=0 Or if you want to wait awhile, the Venice cores will be out soon and might be even better.
  7. Please click on "User CP" at the top-left of the screen and Edit Signature with your system specs!
  8. I've only tried "petroleum-based lapping fluids" in the past on metal working projects not up to the finish standards of block lapping for OC. But I can understand how it would be better than just plain water or soap&water based on this link: http://www.techpowerup.com/articles/cooling/39 This link doesn't mention WD-40, but I have a feeling it would be as good or better than "low smell kerosene". I have to completely agree with you though, plain water or the dishsoap method is far less of a smelly mess! Of course, you're going to have to use a really good solvent to make sure the cutting fluid residue is dissolved and evaporates. But that holds for tap water or soap&water as well.
  9. I agree, I could interpret the chart either way, but the footnote about slots 2/4 kind of nailed it for me.. use the orange slots! Besides, it works
  10. My $0.02: Unless you buy a matched set of video cards today to plug into your SLI, you are better off buying the Ultra-D. Same board. If you attempt to buy a second video card at a later date and hope to plug it in, it probably won't be compatible with your first card. You might be able to get the video card BIOS in synch, then again maybe not. It would be cheaper to just buy a new SLI mobo when you have enough cash/really need 2x GPUs, but by then they will have single PCI cards with 2 GPUs and everyone will be talking about 4x GPUs in two slots. I agree with AG about the drives, even though I have the Raptors and love them. You could save some money and have quieter, cooler-running drives with comparable performance in real world terms. It sounds like the downgraded FX-53 Newcastles or Venice cores might be a better OC candidate than the 3500+ Winchester, but I like the Winchester fine and in theory it might run a tad cooler. About the case, I can say that the Lian-Li I have has very good mechanical engineering and construction quality, but the airflow isn't so great for the PWM and chipset without some modding. My feeling is you don't really need the Lite-On unless you are intent on using their testing software. I'd rather have 2x of the NEC's if I was going to bulk-burn. I like the G.SKILL memory in my sig, but everyone has their pick. TCCD runs at lower volts and is cooler, yet still has good latency and reasonable OC limits. The BFG 6800 GT comes overclocked to 370mHz, while the 6800 Ultra is 400. Mine OC's to auto 410 no problem. On the other hand, it's more GPU than I really need, and the stock fans on the 6800 series are screamers. The 6600 is a lot cheaper, quieter and still faster than any game needs for a year or two.
  11. This is one of the best guides I've seen that makes sense to me: http://www.overclockersclub.com/guides/hea...appingguide.php You might get somewhat better results with WD-40 or kerosene instead of soap or water as a cutting fluid/lubricant. Yes there are some kits available, but Home Depot and AutoZone can fix you up cheaper and faster.
  12. Yes, the chart on Page 21 has 10 rows. Row 9 applies to using single-sided DDR400 in either single or dual-channel modes. Row 6 applies if you are using double-sided DDR400 (I don't think you can get 2T with double-sided DIMMs in dual-channel).
  13. RGone has probably the best "lets just make sure it's done and then do it again" attitude about clearing and flashing. Really good CMOS clearing and multiple BIOS flashing seem to make a difference for some people while others never have a problem. My theory is this is just variations in EPROMs, voltages, and memory. Motherboards and flash software attempt to simulate what a real dedicated EPROM programming station does, but it's sort of a hack. What where your temps like while you had it working again?
  14. It does sound possible that you fried the CPU fan controller on the mobo I don't know what the total fan amps DFI expected for the mobo, but I suspect it was just enough to be safe with a stock HSF + maybe a second little fan (total under 5 watts or 0.5 amps, and that might be it.) From the other threads on this topic I'd feel safest using the Molex 4-pin connectors off the PSU to power fans and use the mobo headers only for RPM monitoring.
  15. After downloading the latest version of the manual (lputnf4 847505101.pdf) it looks like they mean that line 9 applies to your question if you are using single-sided DIMMs. So yes, 2/4 will work at DDR400 with 1T for sure. 1/3 will work at DDR400 with 2T, maybe not at 1T. My theory is that 2/4 have less latency because they are physically closer to the CPU and thus have slightly shorter paths. I'm sure the engineers would have moved them further from the CPU if they could have.
  16. Did the fan ever work properly when you first assembled the system? It sounds like you have at least 4-5 fans in there besides the PSU (1 for the CPU and 3-4 for the case). Have you tried plugging them into 4-pin Molex connectors instead of the mobo?
  17. Yup, I made that same mistake early on, trying to turn off stuff in BIOS I thought I'd never use. The key is to make one change at a time so you can isolate the issue and back out if Windows won't load!
  18. Here you go AG, I was wondering the same thing and it turned out to be already answered in a FAQ on your website http://www.dfi.com.tw/Upload/FAQ/US/MAC-IP.htm
  19. I really like the Vantec cables. I replaced my floppy drive cable with one from Fry's for about $3. The cables are thin, very flexible, and have a nice braided copper sheath around all the wires under the clear plastic tube to reduce EMI inside the case. The headers are nicely done as well. Depending on your case aesthetics you might like the natural copper color or prefer the silver braid version. They also make IDE cables, so I'm going to replace that next with the same type.
  20. http://www.logisyscomputer.com/viewsku.asp...&DID=FRONTPANEL Hahahaha
  21. Realize that the Delta fan is only quiet compared to the Vantec Tornado (i.e., it's the second loudest fan in history). However, it does move a lot of air for its size. You can control the RPM directly through BIOS if you plug the fan into the connector labeled "CPU Fan" on the motherboard. Edit: As ExRoadie points out and RGone states back in this thread http://www.dfi-street.com/forum/showthread.php?t=8068 the mobo fan connectors are best used for RPM monitoring, not powering fans!
  22. Go further down where it says "Boot Other Device". Disable that perhaps?
  23. In theory, partitioning your drive can give you some performance gains. In practice using NTFS, you might never notice a difference. I tend to make 3 partitions: a 2-3GB C drive that is only for the boot loader and the page file, a 10-30GB D drive for XP/essential programs, and an E drive using up the remaining space for games/other programs.
  24. G.SKILL doesn't use heatspreaders, for one. However, they do have some stickers on them that can't help but be an insulator, yet you'll void your warranty if you peel them off. Go figure
  25. Alas, passing Prime95 for 8 hours once doesn't guarantee it will do it a second time. At first I thought my 260x10 OC was stable since it passed Memtest, OCCT, and 8+ hours Prime95 Blend overnight. I ran Prime95 Blend over 9 hours the second night. Then last night Prime failed after only 3.25 hours! Room temperatures were cooler if anything last night, so that wasn't a factor. So I've dropped down to 250x10 and I'm in the process of retesting. I tend to agree with Poisonsnak that OCCT might be a better CPU stresser than Prime95, as Prime95 is mostly doing floating point work and wasn't really designed to be a stress test tool. I notice my CPU usage is higher with OCCT than Prime95. Edit: Right now I'm trying out StressCPU. It does indeed seem to get my CPU 1-2°C hotter than OCCT or Prime95 Large FFT, so as advertised it looks like a good max CPU heat test.
  • Create New...