KM400 or nforce1?

[optimistic]

I'm looking for a decent performing budget Socket A board to pair up with a Throughbred XP1900 chip and Crucial PC2100 memory.

The nforce 1 board is $64 and the KM400 board is $69.

Which of the two would give me better performance? How about onboard video performance? I assume the nforce1 board doesn't have usb2.0 while the newer KM400 does.

Or should I save up and put out for a nforce2 board? Keep in mind though I will be using my current memory, but will be willing to make the jump if the peformance gains is worth it.

 

[Boonesmi]

i would take the km400 over the nf1 in a heartbeat
especially since you want to reuse your memory (the via will be much more likely to work with cheap/generic ram)

ive not seen any comparisons between the new video on the km400 and the nf1, but i would imagine they are pretty similar

also the km400 will support 333mhz bus cpu's, while the a7n266-vm will not

 

[optimistic]

They lowerd the price on this ASUS KM400 board to $64.16. It uses better onboard sound too!

 

[touchmyichi]

get an epox 8rda for 80 bucks, its worth the extra 10 dollars lol. Or if you are really on a budget, then get a newegg refurbished one for around 50. Nforce 2 is great.

 

[Boonesmi]

the 8rda is a great board, but it cost more then $80 once you add in the cost of video card, and possibly new ram and powersupply

 

[mechBgon]

Originally posted by: optimistic
I'm looking for a decent performing budget Socket A board to pair up with a Throughbred XP1900 chip and Crucial PC2100 memory.

The nforce 1 board is $64 and the KM400 board is $69.

Which of the two would give me better performance? How about onboard video performance? I assume the nforce1 board doesn't have usb2.0 while the newer KM400 does.

Or should I save up and put out for a nforce2 board? Keep in mind though I will be using my current memory, but will be willing to make the jump if the peformance gains is worth it.

Are you planning to use the onboard video, and if so, will you be doing any gaming? Also, do you happen to have a digital speaker system that could take advantage of a S/PDIF connection (if so, which type, optical or coaxial)?

 

[optimistic]

No digital speakers. It would be nice to have onboard video incase this machine ever gets demoted to light web surfing duties. I have a GF3 ti200 that will be going into this system. The key features in the motherboard I'm looking for is that it needs to be matx, cheap, & take PC 2100. Matx because I have an matx case.

 

[DAPUNISHER]

Given those requisites I'd get the A7N266-VM. It's mature, stable, has G2 MX video and will do the interweb and other light duties superbly once you take the G3 out of it. It plays nice with a variety of different hardware too. Mech has dozens of 'em

 

[Praxis]

Here are some benchies using Mad Onion's 3DMark2001SE and ran the default standard battery of tests. Here's the result with an Athlon XP1700+ CPU & DirectX9b:

Biostar M7VIG (early version, not the Pro, KM266 chipset, Savage video):
626

Asus A7N266-VM (early version, not AA, if there is a difference, Geforce2 MX Nvidia, drivers downloaded from MS Update):
1763

Asus A7V8X-MX (at default speed, drivers that came with the board):
1105

For reference, a Shuttle SN41G2 reference on the Mad Onion site (Nforce2 IGP, Geforce4 MX):
~3600

Clearly the Nforce2 kicks if you want a gaming rig with integrated video. Then again the cheapest Nforce2 IPG board out there is about $23 more expensive than my board and might balk at using the cheapo RAM I've got in this KM400 box (I tried it on a Nforce2 FIC AU13 & it would spin the fan, but wouldn't post).

What's more, I kicked up the FSB with the mobo jumper to 166 FSB and increased the DDR speed on my cheap PC2100 to PC2700 (333 MHz) in the BIOS and the board seems perfectly stable and happy. I'm at 39C on a cool day with a $6 GC68 fan. I haven't run Prime95 or Memtest86 all night on it yet, but it seems perfectly content squeezing out SETI units as I type. Curiously, just increasing the CPU speed had no effect whatsoever on my 3DMarks, but when I boosted the memory speed my 3DMarks increased quite substantially 1341, about a 21% jump.

I think the KM400 board might be a better choice than the A7N266-VM if USB 2.0 is a feature you'll ever use, particularly since you won't be using the onboard video (at least at first) & the KM400 onboard video is very adequate for non-gaming purposes. Also, the KM400 has AGP 8X support, while the A7N266-VM only has AGP 4X support, if you intend to upgrade the board in the future to the latest & greatest video. Moreover, if you CPU is a T-Bred it might well be easily overclockable to 166 FSB, which would give you a substantial performance boost, and even if it isn't, 99% of Crucial PC2100 will run at PC2700, which isn't an option that you have with the A7N266-VM. The onboard audio on both boards is excellent in my opinion.

Of course a Nforce2 IGP board would have much better onboard video, so if you can afford it and it will tolerate your RAM, that would be an even better option.

 

[Praxis]

Alright, I take back most of the nice things I had to say about the Asus A7V8X-MX. Everything seemed ducky at 166 FSB for my XP1700+ running my cheapo PC2100 Centron Spectek at PC2700. I left the thing crunching SETI units (I'd burned through 5 last night at something like 2 hours per unit at XP2200+ level and the CPU didn't seem to break 42C). I came back home tonight and fired up VNC to kick up the modem and download a few more work units and the machine wasn't on my network. So I went up to the attic and it was stone cold, hadn't even rebooted.

I fiddled with the power and it wouldn't start. Fine, I said to myself, it was crappy memory and wouldn't even light up a Nforce2 board I had. I reset the BIOS to restore the by SPD memory settings & I noticed that I really wasn't that happy with the way the board was laid out. There were two jumpers right near the battery, the CLR CMOS that you expect there and an odd USBPWR56 to wake up the computer from S3 sleep mode with the front USB port. The jumpers were in each others way and it was hard to grab the right one and move it. What's more, the manual was wrong about which way the CLRTC jumper was supposed to be to clear the CMOS and the default. Pretty sloppy. Also, I didn't like the way the panel connectors were laid out. Some companies (like Epox) figured out long ago the bunching those suckers in two parallel rows was a PITA and now (on the boards I have) put them in a single long row that is easy to access. At least the CPU slot is well located for installing big honking 80 mm fans without getting in the way of the memory.

In any case, suffice it to say resetting the BIOS and dialing the dip switches down to default 133 FSB didn't solve my problem. The system would either reset itself while booting or once in W2K it would reboot as soon as I tried to dial in to the internet. Removing the modem and CDRW didn't help. I replaced the Spectek with a stick of Crucial, it still kept resetting. I'm afraid I fried something when the board must have tried to reset itself numerous times after it figured out that the DDR couldn't handle PC2700 speed (after running fine all night long). I have an SPI 400 Watt PS, but tomorrow I may swap in another one and see if that's the gotcha. The other alternative is that maybe something is shorting out on the cheapo case that I have. If the new PS doesn't solve the problem try taking the mobo out of the case and see how it is working. If that doesn't fly it may be time to RMA this board. But first I'll try flashing the latest version of the BIOS, 1002, and see if that helps at all. As usual, the Asus web site is wildly slow and unresponsive, not to mention disorganized. I found and downloaded the updated W2K VIA Unichrome reference drivers (5 MB plus with dial-up) before I could even locate the BIOS for the Asus board (and I was just on the site yesterday).

 

[Praxis]

Update: I revived the board, I believe with a new BIOS, ver. 1004 (that's an extrodinary rate of BIOS introduction, indicating that they were trying to deal with some serious underlying problems). And a couple of days ago Asus finally released a PDF of Insert 1358, which finally correctly identifies the proper orientation for the Clear CMOS jumpers (the manual that I downloaded on August 1st still had the wrong information). The machine seems to be working well and I pushed it up to 166 FSB and PC2700 memory speeds again (with my crummy Spectek PC2100 PNY memory) and it seems fine, but I don't really trust it, so I'm not leaving it on cranking SETI units when I'm not around to monitor it. My guess is that I was just exposed to some teething problems of a new chipset, but it was not pleasant and quite a pain compared to my experience with a tried and true Epox KT400 board I set up at the same time for similar money with similar part (PS, memory, CPU, overclocked the same).

That said, my Asus A7N266-VM also did some acting up, finally most dealt with by deleting ACPI in safe mode and redetecting the hardware, & I'm not feeling too amiably towards it either, though it is now fine. Change my vote to a Nforce2 with IGP.

 

[mechBgon]

Thanks for the front-line report!

My A7N266-VM, and the 24 that I have at work (so far), aren't running into any problems with Win2000, ACPI or anything else either... they just run like champs. What was the nature of the issue with yours there?

Speaking as the guy who builds/maintains the systems where I work, I :heart: A7N266-VM's When the dust settles on this fiscal year, we should have 42 of them in the fleet. Now, I wonder if I can make a custom Win2000 CD that has not only SP4 slipstreamed into it, but also the nForce 2.03 drivers...

 

[Praxis]

My A7N266-VM, and the 24 that I have at work (so far), aren't running into any problems with Win2000, ACPI or anything else either... they just run like champs. What was the nature of the issue with yours there?

Well, the problem may have been caused by the hard drive getting corrupted when the Seagate Barricade started dying. I RMAed it and copied the data over to the new drive, but wasn't able to rescue the install, so I reinstall W2K over the top of the old install, which shouldn't be a problem, though best practice would have been to format and start over again. The other possible cause was that I applied all the patches from Windows Update catalog at once without rebooting between patches without running the qchain.exe, which I gather is not recommended. I think that may had had the effect of not letting me properly unload the BIOS when shutting down or rebooting or trying to hibernate. Anyway, the machine now hibernates and shuts down, though it does take rather a long time (~2 minutes). But it isn't worth my time to reinstall, especially since it runs SuSE 8.2 most of the time.

 

[mechBgon]

Regarding the slow shutdown, you might try this:

• Start > Run > gpedit.msc
• In Group Policy Editor, drill down to Computer Configuration > Administrative Templates > System > Logon
• Set "Maximum retries to unload and update user profile" to "10" (for 10 seconds) and reboot

(it's a Win2000 problem that surfaces after a certain patch is applied)

 

[Praxis]

Thanks, mechBgon, it worked like a charm, but I may increase the timeout to 15 seconds just to give the computer a little extra time if it is actually doing anything useful.

 

[bambam]

I would agree with Praxis . My experience with Asus KM400 board has been somewhat disappointing .( It is much slower than I expected and I'm having problems getting DirectSound to work . ) I was hoping that this board would be good all-in-one value sytem solution but I can't say that now .

 

[Vette73]

Originally posted by: bambam
I would agree with Praxis . My experience with Asus KM400 board has been somewhat disappointing .( It is much slower than I expected and I'm having problems getting DirectSound to work . ) I was hoping that this board would be good all-in-one value sytem solution but I can't say that now .

Check for a new Bios and drivers. The KM400 is NOT a updated KM266. The KM400 has a new video card and memory controller, etc...