Jump to content


  • Content Count

  • Joined

  • Last visited

About arthurleung

  • Rank
  1. SVC-800-Cu. With the airflow coming from my HDD's 120cm fan it is just slightly warmer than my hand and its passive! Although I lose 2 SATA ports (SIL3114) and 4 USB ports (headers are blocked). Work with 6600GT SLI too
  2. Just use windows x64...I had been using it since August with 4x1GB sticks at 510-3-4-4-8-2T. Drivers are fine (except webcam and printer) and games were fine. I tried running 32bit windows on it and it never got the full 4GB working regardless of setting. BIOS got 4G regonized, so do memtest but not windows
  3. Five minutes ago, when I changed one of my SATA disk to GPT disk and rebooted (to install new windows) it stuck at POST. I had exectly the same problem when I moved 4x 160G Hitachi SATA HDDs in GPT instead of MBR (2 are PATA but with converter) The disk booted fine on SIL3114 SATA mode and on PATA (two of the disks). BUT once there is any of the GPT disk on the NF4 SATA it will not get past BIOS. It happens across Hitachi, Maxtor and Seagate. But it works on my A8N-SLI (Both premium and deluxe). Using 704-2BTA Bios. Am I the only one affected by this? P.S. Index showing my thread as NF4 Expert?
  4. My Adata DDR566 was rated at 285-3-4-4-8, If I want no error in Memtest / Prime95 I either run it 250-3-4-4-8 or 200-2-2-2-5. There is a few bits of memory at 925.8M which keep f***ing up. Its always there in Memtest Test6 and occur at any speed between DDR520 and DDR620, regardless of timing being 2.5-4-4-7 to 3-5-5-10 and regardless of voltage 2.75V to 2.95V. I also tried DDR440 2-3-3-6 and it still show up. I guess its just bad luck... The previous sweet spot on my Asus was 500-3-4-4-8 which had no error in Prime95, couldn't get memtest to run due to me being floppy/cdrom free on that rig) The kit doesn't past memtest nor prime95 at speed >DDR520, I'm currently stress testing them (just moved back from Asus to DFI) with WinRAR over network and see if it throws up any CRC error. So far there is no error at all, so I guess if I don't touch that faulty bit of memory it won't do any harm... Now if I can find something to locate and lock that memory range...
  5. See memtest gives you the approximate location of the faulty ram, is there a way that you can lock / hidden / reserve that broken address so that the os / all programs cannot access it, therefore not causing error despite the memory is somehow faulty?
  6. Try to disable NCQ in your Nvidia driver setting for your system drive. It cut my bootup time by halves.
  7. I was playing game just now and the system suddenly hangs. Upon reboot I got disk boot failure. Went into the BIOS and found out that my 2 SATA disks were gone. I did the normal procedure, changing cables, changing port, I even went and moved the hdds to external sata boxes just to put the PSU out of equation. Connected the disks to another computer and they worked fine. Tried an IDE hdd and the BIOS would regonize. I'm posting to see if there is any other solutions before I had to spend a lot of money to RMA the board Thanks.
  8. Do you know what chip is your VS? I'm using UCCC chips with the following settings: 200-E-3-4-8-4-10-12-3-3-2-3-3120-A-E D-0-7-3-8.0-N-5.5-256-D-16-7-D(4) at 255Mhz(DDR510)
  9. Shouldn't you be using the 4V jumper if you need more than 3.2V?
  10. I could do DDR680 passing memtest but only actually boot into windows stable at DDR600. Windows is simply more ram stressing than memtest. If you can't even boot windows, you need a lot slower for prime95 testing
  11. It could be the ram doesn't like the ram slot (which makes it run a few mhz less in order to get stable), Or the voltage reduced by like 0.005V due to 2 sticks or ram, or the Memory controller is getting more load, etc. Lots of reasons can cause the Dual Channel to be less stable than Single Channel. What happens usually is that, because there are 2 sticks of ram, the signal gets weaker or take longer (in nanosecond scale) to send. So there might be chance that the signal is lost or didn't arrive in time. +0.1V or pushing a few Mhz down usually solve the issue, but if it doesn't, try disable CPC (with performance impact)
  12. It really depends on if you just want to do a quick test or a full test. If you are doing a quick test, I would say do test #5 (c135 + enter + 0) for 5 times then test #6 (c136 + enter + 0) for 5 times. You test 1 stick if you want to test the maximum frequency each stick would go, and you test 2 sticks if you want to test the maximum frequency each stick would go in DUAL CHANNEL mode. You should always test both sticks in. You could push your HTT/Ram a bit faster, or use better timing. Then continue to test. If you got any error, back off to the last setting, then performa a full test (for 5~20 times). That would be the quickest way to find maximum ram stable speed.
  13. I usually set to 8ns and 5.5ns, 9ns will cause me not able to boot.
  14. I modded my heatsink, 53'C fully passive cooling lol. 40-45 is pretty good already, except the noise (of the original heatsink)
  15. Pushed the VDIMM up to 2.85V, it is now stable in Prime at DDR600 now. But system would not boot at 2.95V. 18A Probably just cut it because I got a 6600GT only.
  • Create New...