Jump to content


  • Content Count

  • Joined

  • Last visited

About Maarek_Stele

  • Rank
  1. But unlike those UCCC's, his sticks will top out way sooner. I assume a divider is being used to keep them near 200mhz for starters.
  2. If that's an old CO clawhammer, most probably ~2.4 will be it's limit. Some newer CG can go near 2.6 or above, but sometimes with tons of voltage. Usually they max out at 1.7v, but I wouldn't do that on air 24/7, but up to 1.65v with proper cooling is safe for these .13 parts. Try lowering htt mult to 3x, relaxing max assynch latency and read preamble to something high as 9/7 for starters.
  3. Pretty weird, on my nf250 I've directly swapped a 3200+C0, a 2800+CG and 3700+CG without a serious problem. But sometimes on first boot I got a garbled screen, a reset would solve it forever but on the other hand, I get that when I change pci stuff too. If the cpu is good, it might be memory controller related, maybe you should try to boot it after cmos reset with only one stick.
  4. Sorry for not helping with this post, but you are the first person around that I've seen successfully running a x1600 with this board.
  5. Mine does that too if I leave some usb stuff connected like an ipod nano.
  6. Edited the first post to only deal with nF4 boards.
  7. Not that I'd recommend it to anyone, but talking about small psu's and the good old dfi 250gb: I'm running a 5 year old chieftec 340W (15a 12v) with an a64 [email protected] 1.55v 48h prime stable without a problem. That computer goes trough some serious hours of punishment (image editing, 3ds max etc) weekly. The gfx is only an fx5200 tough.
  8. I guess it should be 237mhz (ddr474), not ddr430. Because if you are using a 9.5 multiplier, the "ddr400" memory divider will use the next upper entire value which is 10.
  9. Works fine with me, but either usb kb and mouse legacy support must be disabled as with previous bios.
  10. I have 2 hammers (co and cg) and a new castle (cg), and 3 754 boards, two dfi nf3s, one abit kv8max3 (via kt800). With every cpu on every dfi, I can do stock speeds with agressive timings on any slot, but when overclocking using the middle one I get less mhz without errors. On my kv8 max, to even be able to use trcd2, I must use 1+3 - but at same settings, it's slightly faster on 1+2 than with 1+3, I never tested it on my dfis, but it's probable that the slightly faster combination will top earlier than the most relaxed one in my case (1+3).
  11. If the problem is the hd taking too much time to start up so it cannot be seen on first boot, the simple solution is to add an ide detection delay, like 2-5s or so. For array screens, the normal solution is to reduce htt multiplier a notch when chipset is the culprit. If it's memory controller related (sometimes it happens when running high htt 1:1, or with newarks even at low htt with some chips), try raising max assynch latency along with preamble time. Default is 7/5, try 8/6 or 9/6.5, 9/7.
  12. I might be wrong, but some time ago people only went above 1.7v with active cooling. Most do that to achieve higher HTT.
  13. http://www.dfi-street.com/forum/showthread.php?t=68177
  14. This is the mobile pad I was talking about. https://shop.msicomputer.com/App/StoreFront...il.aspx?ID=1021 Sorry for posting links to other mb manufacturer, it's just that dfi hasn't a similar product.
  15. There was some brand selling copper heatsinks for the cpu like the alu ones on desktop cpus (as msi or gigabyte). You might look around for it to be able to use that newark with any sink, despite some loss in performance.
  • Create New...