blackdragon57

I sent the EA Support Staff a message in regards to Need for Speed Carbon & SLI. It seems as though the demo doesn't support the technology. While this may not seem to raise too many eyebrows, their reply will: " Response (Leo V.) 11/06/2006 07:37 PM Hi, Thanks for writing us at Electronic Arts Technical Support. I'm sorry for the inconvenience this has caused you. Unfortunately, as the demo is not a finished product or an official release, we are unable to offer technical support. Issues we have noticed and received from other customers have been dually noted, and these issues should be fixed in the final release of the game. As for the SLI support with our games, there's no final note on this as the games were never tested with the SLI technology. Should you require further assistance with this game or other Electronic Arts games in the future, please visit our website and review our extensive Self Help knowledgebase at (http://support.ea.com/). Thank you, Leo V EA Technical Support " I would like to draw special attention to the part stating, "...there's no final note on this as the games were never tested with the SLI technology."

I am currently using a Linksys router on my setup. I used both nVidia LAN and Marvell. Both ports work perfectly. I'm sure this wouldn't be the case if I installed that nVidia network managing software...

Many reach a wall around the same bus/htt speed (220 mhz). Many have solved this by increasing the vcore voltage total to be between 1.35v and 1.40v. However, if you do this, please monitor your temperatures carefully. Happy OC'ing!

I never seen the benchmarks comparing the two, but from what I understand, current video cards cannot produce enough data to saturate the PCIe bus anyways. So, theoretically, no. Again, I haven't tried it, nor I have I seen benchmarks.

You need to set the jumpers. The first slot (16x) is meant to contain one individual PCIe video card; the second slot for the video card was intended to be used for SLi, but probably can be used individually. Example: SLi Jumpers Normal PCIe Video Card Slot 1 = 16x Bandwidth PCIe Video Card Slot 2 = 2x Bandwidth (8x slower than the first slot) Jumpers Set to SLI Mode PCIe Video Card Slot 1 = 8x Bandwidth PCIe Video Card Slot 2 = 8x Bandwidth (4x faster than before)

Unfortunately, that wasn't the case. I did adjust the LDT/FSB to x4 (and is currently that way). Since all other functions were set to Auto at the time, the motherboard automatically adjusted the speed to equal 833 mhz. I do know exactly how strange that is. Again, temperatures are all within normal ranges on any test. The highest recorded temperatures come from the GPU's that reach nearly 78 C. Next is the chipset at 55 C. Processor only reaches 41 C. All these temps are recorded in a 26.7 C environment. BSOD message would not occur. PC restarts regardless of that setting. At the moment, the system has been completely stable while running 3DMark05 at full settings (over 10 passes on every test consecutively). I also manually set my memory timings to 3-4-4-8 (2T setting) a few days ago, instead of having it on Auto (no errors in several hours of Memtest86+). I noticed near-complete stability in 3D applications when the two things happened: (1) The processor was set to 833 mhz for an unknown reason by the motherboard, (2) Dual-Core Optimizer intalled (but driver has always been installed). I normally judge complete stability by all of the following: (a) 8 hours of Memtest86+ without error; (B) 12 hours of Prime95 without error; © 8 hours 3DMark05 on full settings; and (d) one full week of gaming on the machine.

That is most-definitely correct. However, I never did install the Dual-Core optimizer. I just did that (and increased the voltage on the chipset); now I found more stability. We'll see how long it lasts. This thread can probably still considered closed. It takes SLI problems too long to show (usually a week or so). I took a look at the X1900XTX and the performance is similiar in BF2; however, other games show more difference between 7800GTX 256MB SLi and the X1900XTX. It took me some time to decide between the two. On one hand, I could have much more stability and over-clocking capability; then on the other, I have SLi... a technology that sounds nice, and is nice -- only when it wants to work, lol.

Wow. Unfortunately this isn't solved, lol. I really thought it was, until I noticed that my CPU speed dropped to 893 mhz when I dropped the LDT/FSB to 4. Naturally, I believed that setting the CPU/FSB multiplier to 12x and manually setting the HTT to 200 would solve the problem... This, again, causes the game crashed. Which told me that I could possibly blame the CPU for once. Therefore, I tried 24 hours of Prime95 and not a single error occured. After 30 minutes or so in 3DMark05, the computer restarts and SLI is disabled for various reasons. Only once did I get an onscreen error when I returned to windows ("Windows has just recovered from a serious error, blah, blah.") When I sent the error report, it pointed at nVidia's drivers (84.21) or other Windows XP updates. Why don't I believe that the drivers are a problem? Mostly because 3D applications were working perfectly when my processor was running at 833 mhz, lol. Yeah, still not acceptable. I am now left back at my first conclusion, the PCIe or HyperTransport bus just isn't capable of handling that much data flow (though I could try increasing the voltage on the chipset -- my current load temps are 50 C for the chipset). **I did try increasing the voltage on the processor to a total of 1.4V, but no luck. It just raised the temps without any gain in 3D stability**

LMAO. Okay, that wasn't correct -- or at least not for my fan. After I ran 3DMark05 for 1 minute, the entire system shut down. It took a few reboots to get everything back online. At first I believed that my video cards messed up all over again, but no. This Vantec fan splits the main motherboard power cable. The power is drawn from the PSU, and one yellow cable (to give the speed information) is sent to the motherboard via one motherboard fan slot. It's messed up. I nearly cooked my processor (temperatures reaching around 85 degrees C). Well, after that fiasco, I think I'll buy a new fan only interacts with the motherboard, lol. **Good news, it's back down to 29 degrees C with the case cover off, lol. That was too close for me.**

K. Thanks for the info. Everything is working great now.

In the Genie subset of the BIOS menu.

Good call magpie. It was the HyperTransport bus stalling the system (which is why I would see hear problems with all devices on the PCI bus). The only two types of cards installed on that bus are the video and sound. I reduced the HTT down to 4x and found much more stability. I guess this thread can be closed with Magpie's solution, "drop you htt to x4." Thanks everyone for all the support!

Per the BIOS: CPU Fan Fully On at > 50 degrees C CPU Fan Fully Off at < 25 degrees C CPU Core Temperature = 34 degrees C CPU Fan Speed = 3055 RPM The max speed of my fan is 3100 RPM. I'm sure the motherboard is reading it a bit wrong, or the fan really isn't capable of a true 3100 RPM. It is a Vantec fan that includes a 4-pin (power supply) & 3-pin (motherboard) connectors; both are connected. I could attempt to disconnect it from the power supply, so it would only draw power from the motherboard. Though, I do recall my CPU fan working properly on my previous LanParty Expert board with the same setup. This isn't a huge problem... just more of an annoyance.

Any idea why a CPU fan would always be running at 100% all the time, despite temperatures of 32-36 degrees C on average? When I turn on the PC, the CPU fan sounds normal, but after a couple minutes, it winds up to full and stays that way forever (as long as I don't shut it down )

I'm just curious at this point about one thing. Has anyone got SLI to work on a DFI LanParty Expert board with a dual-core processor?