Jump to content


  • Content Count

  • Joined

  • Last visited

Everything posted by oralpain

  1. A Corsair VX450 will be more than enough for that setup. Goes for around $70. Generally, any recent, truly reputable brand PSU with 400w or more should suffice.
  2. I have two DFI Blood Iron's; one in each my primary and secondary system. They were a good buy upon release at $140, they are a great buy at $108. The memory performance and tweaking options of this board are second to none. OCing in general is extremely good. The layout is above average. I've tried RAID with them just enough to confirm that it works fine. I haven't tried using the last two SATA ports for it however. Like pretty much every P35 board in existance it has a single port (two device) JMicron PATA controller. I've been able to get my PC6400 CL4 Ballistix (not tracer, but they use the same chips) to a stable 580MHz (DDR2 1160) @ 5-5-5-15, with tight subtimings on this board, 24/7 stable. I currently run 500MHz (DDR2 1000) 4-4-4-12. With active cooling on the ram and a lot of voltage 600MHz is possible, but I wouldn't want to leave it that way in the long term.
  3. Only glaring weakness I see for gaming is the video card. What are the specs on that Fortron PSU? Fortron has somev ery good models, so you may not need a new PSU at all. A quality 500w is more than enough for a fast dual core and a single 8800. I've been running, in what is now my secondary system, a 420 watt Enermax (29A +12v) for over a year with an overclocked dual core opteron (now an oced C2D), and an OCed G80 8800GTS. I've had zero power related problems.
  4. What is the heatsink attached to exactly. if the component is somwwhat durable, putting it in the freezer for a while may very well work. If not, submerging it in pure acetone (the kind you buy at a hardware store) for several days may do the trick.
  5. I have to agree, at least if you are building a new system. You can save about 10-20 dollars by going AMD if you are positive you will not OC and are willing to risk the possibilities of not ever having a worthwhile upgrade path. I've got a 70 dollar Intel chip (Pentium E2140 @ 3.2GHz, cheap air cooling) in a DFI P35 board that currently goes for $105 and the setup is faster than any dual-core AMD you can get without sub-ambient cooling. As for AMD chipsets and overclocking, I haven't really seen a problem with them. A good board with an AMD chipset will take the CPU to it's limit before the board craps out. NVIDIA chipsets may very well OC better on the HT bus, but this means pretty much nothing in terms of performance. If you already have a decent AM2 CPU, go ahead and buy an inexpensive AM2 board. if you've already got an AM2 baord and need a faster CPU, go a head and buy a black edition. However, if you are building from scratch, there is virtually no reason to side with AMD at this time.
  6. 9800 line looks to be not much more than G92 GPUs with a few revisions. 8900 would have been a more apt name.
  7. Not a bad OC on that card, so far. It's probably an A3 rev and should have more headroom than that to play with. Use a program that can alter shader speed separate from the ROPs (core). A BIOS flash won't do squat to help you OC more. The memory can be vmodded with a pencil, but the gains before heat becomes unconrolable with the stock cooler are very small. Altering the GPU voltage is the only way you are going to get substantially more out of the core, and that requires a hard mod (soldiering). The only reason a flash has helped some GTs is that some cards were not set to use the highest voltage availabe through the bios (1.1 2d/1.15 3d for the 8800GT). Any voltage over 1.1/1.15 requires a hard mod for that card. As far as I am aware, no 8800GTS was set to use anything but the highest available voltage. I am positive they would not have set a lower voltage on a pre OCed card. The reference clock on all of these cards is 27MHz. ROP, shader, and memory clocks are derived from this through multipliers. Best I can figure: The ROP clock can be altered with a granularity of one-third and one-half multipliers. The Shader clock can only be altered in increments of 2 whole multipliers (i.e. 54Mhz steps). The memory clock can be changed in increments one-third multipliers (9MHz steps).
  8. Still looking for this? I have piles of PC2100 (128MB-512MB).
  9. AMD is pretty much a better all round processor than intel, at this time. A the same price point, intel is having trouble even things it's traditionally dominated AMD in, esp when it comes to the dual cores. http://reviews.cnet.com/4520-10442_7-6389077-1.html?tag=lnav Yonah, and the the other next generation intel chips do look like they will be more competative though. Funny thing is, by the time most people finnaly realise than AMD has had a very firm performance lead for the the last 2.5-3 years, intel may be making a comeback. So, the ignorant fanyboys will aways be a step behind...
  10. I probably not, as everythign I've seen leads me to believe DDR2 and DDR3 are pin compatable. You might need a new cpu for DDR3 (when moving from ddr2) but I really doubt you will need a new motherboard. I'm pretty sure M2 is supposed to be 940 pins. Socket F seems like it would need extra for the additional hypertransport links. Well, we'll see.
  11. If I had a choice between an opteron 180 and an X2 4800+ (both socket 939), for vert simialr prices, I would get the opteron. I'm inclined to think it would be more likely to overclock better, and is otherwise the same chip. I'd be using my own heatsink anyway, and if the chip lasts the first week in my hands, it's probbaly going to last as long as I need it to.
  12. There is also a possible issue about BTX putting the memory too far from the processor for AMD chips.
  13. If it's got an integrated memory controler, it's a K8.
  14. Generally yes, but it depends on system. Even 3dmark 2001 is video card bottlenecked on a TNT2 with any recent CPU. Likewise 3dmark 2005, is CPU bottlenecked on a 3.9Ghz FX-57, and SLI 7800GTX 512MBs cooled with phase change and overclocked to 800+Mhz core. However, you are right, for most reasonable systems, 3dmark 2001 is mostly CPU/memory limited, and 2005 is mostly video card limited, with 3dmark 2003 being in between. Any time. Just don't take my word as gospel. Do some research for yourself.
  15. My AGP X800pro non-vivo will NOT mod with the bios flash, even if I close the bridge properly. I'm fairly certain there are resistors else where on the card that are also blocking the extra pipes. ATiTool shows a difference in the resistor positions between the vivos and the non-vivos. To connect the bridge I used the copper conductive paint from a rear window defogger repair kit, a needle (to scrape off the coating on the two contacts) and a tooth pic (to paint the line). Note that you have to go around the laser cut or fill it in with soemthign non conductive before putting the conductive material over it. I have used the same method previously to unlock Athlon XPs and a few other things, and I am positive I did it correctly on the X800, but the card was still not modable. I also tried it on a PCI-E x800pro vivo (wich also has the laser cut), it didn't work there either. The AGP x800pro vivo cards I've seen do not have the cut at all and a bios flash is all that is needed to mod them. Most are failed X800XTs to the AGP vivos tend to OC very poorly. I have an artic silencer on one (properly applied AS5 is the TIM) and at 100% fanspeed the highest 100% stable core clock is 520MHz at 1.4v. And this is not just a bad card, it's the best one out of 4 tries.
  16. Ignore the plain CPU temp on everest. Just look at diode. Also, "CPU diode" on everest is not the on die temp, its the heatspreader temp (tCASE), atleast with the DFI mobos I've used. 28C is quite normal for an idle venice tCASE temp.
  17. I had the original true control 550, was ok, but really lacked in the 12v rail (24 amps, ok but not good for 550 watt psu). The SLI one looks like it should be a bit better.
  18. I'm very suspicious of that Crosfire DFI board. It appears to have some major problems.
  19. Bad sectors are useually, but not always, a physical problem with the drive. I say try the low level format/zero write and use the drive. If you get more bad sectors soon after, the drive is dieing and you should replace it asap, unless you don't mind permanently losing data.
  20. Alown, that is not how it works. The multiplier does not divide anything into frames. It's far from pointless to have a higher multiplier. Current (super scalar, pipelined, out of order) CPUs could not possibly process all the data they would beable to recive a 1:1 (cpu:bus) memory and/or system bus, in the vast majority of situations. If you double a modern CPU's data buses and halve the multiplier, you will probably see a nice performance increase. Howerver, if you go further and tripple it while cutting multiplier three fold you would see much less of a boost. Efficency drops at each stage. You will eventually hit a wall where the CPU just cannot process data anyfaster, so will stop asking for it till it needs more. This is why, with my new S939 dual channel setup, I get only about 50-60% more memory bandwidth than I did with the same memory and settings in a single channel config. Max theoretical bandwidth did double, but my CPU is not fast enough to process all the information it recieves in many cases. I am not normally limited by memory bandwidth even with a CPU clock 12 times as fast as the memory clock. I actually have more *usable* memory bandwidth running the memory at 194MHz with the CPU at 2.71GHZ than I do with the memory running at 250MHz and the CPU running at 1.5GHz, for example. CPUs do alot of stuff with a given instruction and peice of data. They do not need massive relative bandwidth to reach their limits. It takes alot longer than 1 clock cycle to process most anything. http://arstechnica.com/paedia/c/cpu/part-2/cpu2-1.html 3dmark tends to be video card limited. This is where the term bottleneck comes in. If one component is faster at getting a given task done than other components, you have a bottleneck. It's going to vary from program to program and task to task. When recording video my hardrive is the bottleneck. When encoding video, my CPU is the bottleneck. When playing BF2, my video card is the bottleneck.
  21. K is for kryptonite (intel was "superman"), just a code name they came up with over a decade ago. Still used, offically or not, to denote cpu generations. K7s where the entire pre 64-bit athlon line from the original 500MHz to the XP 3200+ (slot A and socket 462). K8s are Athlon 64s/Opterons (S940, 754, and 939).
  22. The reason for that is the increased cache latency with the 600 series. If latency was kept the same, there would have been more of an improvement.
  23. There are several games that will use more than 256M of video memory, at the right settings. The newest cards from ATi and NVIDIA are getting fast enough to really use these settings, so 512M is not worthless, definatly not "necissary" for msot at this point, but not worthless. Anyway, the massively higher clocks are responsible for much more performance than the increased ammount of memory.
  24. In most cases yes, but I still would not call this too precise (no such thing IMO). It's good to have a safety margin.
  • Create New...