Dave wrote:
> Device Manager properties and various diagnostic apps show the LAN
> adapter as being a Via Rhine ll device, which I assume is integrated
> into the 8237/R SB. Any drivers I've found for the K8T Neo-V, or
> specifically 8237/R southbridge -- Via's site, for example --
> indicate the same thing: Via Rhine ll.
>
> However, page 3 of the K8T Neo-V manual has the following:
>
> LAN:
> * 10/100 Mbps Realtek 8201CL
> * Gigabit Ethernet LAN Realtek 8110SB (optional)
> - Both integrated Fast Ethernet MAC and PHY in one chip
> - Both compliant with PCI ACPI 2.2
> - Both support ACPI Power Management
> (I assume the Gigabit option is not on this mobo.)
>
> The curiosity is the Realtek 8201CL. Does it exist on this board?
> (My wife's machine, Win2k SP4)
>
>
> We have 3.0 Mbps Embarq DSL service and I can only get 1.5 Mbps out of
> the Via Rhine II LAN adapter. I'm using a Via driver -- 3.66.0.451,
> 12-20-06 -- newest I can find. My Epox 4PDA2+ machine (Win2k SP2) with
> the Broadcom netXtreme Lan device provides consistent 3 Mbps rates.
> The 2 machines were tested separately using Embarq's online test,
> confirmed with DU meter.
>
> A speedstream 5260 dsl modem feeding a Netgear DS104 hub feeds the 2
> machines. Embarq allows for the assignment of multiple IPs.
> Connecting my wife's K8T Neo-V puter directly to the modem makes no
> difference.
>
> The question is whether that Realtek 8201CL device actually exists on
> this machine, and if so, how do I enable it? BIOS only provides
> enable/disable of LAN device. Finally, assuming it doesn't exist, why
> can I not get maximum data transfer rates out of the Via Rhine ll
> device?
>
> Am I missing something obvious? Very confusing.
> Dave H
The RealTek 8201CL is a PHY chip:
http://www.realtek.com.tw/products/productsView.aspx?Langid=1&PFid=13&...el=5&Co
The 8237 appears to have an integrated Ethernet MAC, and uses an external PHY.
http://www.via.com.tw/en/downloads/presentations/chipsets/southbridge/...psets_v
About the only thing I can initially suggest, is checking what rate the
PHY has negotiated with whatever it has connected to. For example, your options
could be things like 10 base T, 100 base T, half or full duplex. Sometimes there
is an option in Windows to force the interface to a particular state. But that
doesn't completely fit your problem description - 10 base T would be limited to
1.25MBytes/sec max. You claim 1.5Megabits/sec, which is 188KB/sec.
(MB is megabytes, Mb is megabits).
Even if the MAC wasn't using DMA to transfer the data, the thing should
go faster than 188KB/sec.
One way to test computer performance over Ethernet, is to connect
two computers together. 10/100 interfaces generally don't have MDI/MDIX
(the ability to do the necessary crossover cabling on their own). With
two computers with 10/100 interfaces, a crossover cable would normally
be used (the one I own, has a red connector on one end of the cable, and
a blue connector on the other end, visually suggesting the cable is crossover).
If one of the computers involved in a direct connection is 10/100/1000
type, generally those have MDI/MDIX option. And then the 10/100/1000 end
figures out the required cabling. With such a setup, an ordinary cable,
with connectors being the same color on both ends, would work fine
between the two computers.
http://en.wikipedia.org/wiki/Ethernet_crossover_cable
What I use, is I set up an FTP server on one computer, and an FTP
client on the other. Then I transfer a file from one computer to the
other. (I use FTP to avoid Windows file sharing, which is a performance
variable I don't need.) To take the disk out of the picture, I have
used a RAM disk, but for the purposes of a quick test, you don't really
need that refinement. When I was using a RAM disk on the FTP server side,
I was trying to get more than 40MB/sec from the thing. I was testing two
computers that had gigabit Ethernet, which should have been able to
do 125MB/sec max - but it turns out that Win2K won't go that fast.
WinXP is more capable in that regard.
Paul
FAQ
Profile
Private Messages
Log in
