A7N8X Deluxe: 3Com LAN or Nvidia

Tycoonbrad

n00b
Joined
Mar 5, 2004
Messages
32
Which built-in ethernet is everyone using on the Asus A7N8X Deluxe and why?

3Com LAN or Nvidia Ethernet
 
I use the one that's on Top...3Com i'm sure.

The other port is for connecting another computer to the internet through it's port.
 
I like to use the Top (3 com) port because it seems to be more reliable. I used to have an Xbox connected to the second port and it would not connect sometimes, where as the 3 Com port never had any issues.
 
I use the 3com because compiling the 3c59x driver is much easier then trying to figure out nvidia's driver bullshit(in linux, that is).
 
I've been using the 3Com one.... I'll use the nVidia one when I'm on a different network, just so i don't have to mess with changing settings...
 
RBall151 said:
I use the one that's on Top...3Com i'm sure.

The other port is for connecting another computer to the internet through it's port.
WTF? Don't post useless info, man, you'll confuse people.

The one on top is 3Com, the one on bottom is nVidia.

The 3Com MAC is connected through PCI and is slower (in throughput) than the nVidia MAC, which is directly connected to HyperTransport. So for normal usage, use the nVidia. Also, the nVidia MAC has less CPU utilization. (Actually, for normal usage, you probably won't notice any difference. :p)

HOWEVER, the 3Com is MDI/MDIX aware—if you don't have a crossover cable, you can still jack another computer directly into it. I usually connect to a network using the nVidia port and leave the 3Com open in case I need to just jack someone in temporarily.
 
The A7N8X board is nforce2, so it's not hypertransport. But the nvidia LAN is built into the MCP. I'm not sure if this is better or worse, though, since it should give lower latency but it could also cause more SB utilization (and maybe heat).

Personally, I use the top one, just because it was easier to reach. ;) I haven't really thought about which is the "better" choice.
 
Karmakaze said:
Personally, I use the top one, just because it was easier to reach. ;) I haven't really thought about which is the "better" choice.

Exactly. The performance difference is very slim, so it really comes down to which one is eaiser to reach. I have all of my school network stuff set up on the nVidia card so i use the 3com when i'm at home.
 
Stiletto One said:
WTF? Don't post useless info, man, you'll confuse people.

The one on top is 3Com, the one on bottom is nVidia.

The 3Com MAC is connected through PCI and is slower (in throughput) than the nVidia MAC, which is directly connected to HyperTransport. So for normal usage, use the nVidia. Also, the nVidia MAC has less CPU utilization. (Actually, for normal usage, you probably won't notice any difference. :p)

HOWEVER, the 3Com is MDI/MDIX aware—if you don't have a crossover cable, you can still jack another computer directly into it. I usually connect to a network using the nVidia port and leave the 3Com open in case I need to just jack someone in temporarily.
So are you saying that the 100mbps data transfer rate of the nvidia port is faster than the 100mbps data transfer rate of the 3com port? Regardless of internal structure, if they can't both transfer data at the same speed then I want my damn money back.
 
jpmkm said:
So are you saying that the 100mbps data transfer rate of the nvidia port is faster than the 100mbps data transfer rate of the 3com port? Regardless of internal structure, if they can't both transfer data at the same speed then I want my damn money back.

Does any 10/100 ethernet product transfer at 100mbps?
 
swoop56 said:
Does any 10/100 ethernet product transfer at 100mbps?

Assuming perfect conditions, yes.

Conditions are never perfect.

Therefore, 100mbps product will never transfer at 100mbps.

It comes much closer when you include the overhead of the protocol that you're using though...
 
you can get to 100% network utilization but not 100mbps. oh well. the 3com is what rogue_jedi uses, b/c of the linux support. however, 2.6 has hacked support for the nvidia lan so if that works he'll prolly switch over 'cause the 3com thing at boot in windows is annoying. me, i get along fine with my gigabit intel csa lan :cool:
 
Karmakaze said:
The A7N8X board is nforce2, so it's not hypertransport. But the nvidia LAN is built into the MCP. I'm not sure if this is better or worse, though, since it should give lower latency but it could also cause more SB utilization (and maybe heat).
Wrong. HyperTransport first hit market in nForce1, and the nVidia Ethernet MAC was on HT back then too.

Anyway, the two are the same speed rating, but the nForce MAC comes out on top in performance. Not by a huge amount, but it's nice to know.

Yes, it'll heat up the southbridge more, but then again, you've probably noticed that most nForce2 boards don't even bother with a southbridge heatsink. So I'd just shrug it off.
 
Stiletto One said:
Anyway, the two are the same speed rating, but the nForce MAC comes out on top in performance. Not by a huge amount, but it's nice to know.
How did you test this? I would be interested in seeing the numbers to see just how much better the nforce one is.
 
I use the bottom one because its easier to reach. Even if there is a slight difference how much is it going to change anything? Maybe a few kb in a download or 3-5 ping in a game?
 
I use the top to connect to my DSL, and bottom to network my laptop with a crossover, and then bridge the internet connection to the laptop.
 
Lazn_Work said:
I am sure he is referencing this review:
http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=1719&p=10

But an update here addresses some of that:
http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=1731&p=4

==>lazn
That is quite interesting but pretty pointless since they say that the 3com's poor performance was due to a bug in a testing version of the chipset. Did they do a benchmark with the final version of the chipset? I might have to run some tests myself(if I can get the nvidia-net interface working).
 
jpmkm said:
That is quite interesting but pretty pointless since they say that the 3com's poor performance was due to a bug in a testing version of the chipset. Did they do a benchmark with the final version of the chipset? I might have to run some tests myself(if I can get the nvidia-net interface working).

I could not find one anywhere. But at the time of the first review, they did not know the 3Com Nic had a bug when used with the MCP-T.

==>Lazn
 
I use the nvidia port for my home network, have used the 3com before, no difference from either one that I could tell. I stay with the nvidia because its on the bottom less cable overhang. No other reason.

MrDux
 
Back
Top