The SX41's VPN setup and capabilities are identical to the VP41's as far as I could see, so hop over to that review if you need the details. The main difference is that you can define only two tunnels vs. the VP41's 70. I like having the Connect and Disconnect buttons, and the Summary pop-up window is handy too. It's also nice that NetBIOS broadcast is supported so that Windows network browsing, i.e. Network Neighborhood or My Network Places works without the hassle of having to Find computers on the other end of the tunnel.
The key difference between the SX41 and VP41 is that the SX doesn't use an IPsec co-processor, similar to most other consumer-priced VPN endpoint routers. VPN throughput for no-coprocessor products has typically measured about 700kbps, which is what I expcected for the SX41. But Linksys' choice of the Conexant CPU looks like a good one, since as the Performance table below shows, the SX's tunnel throughput was a surprising 2.4Mbps... almost 3.5 times that of other non-coprocessor routers! This definitely makes it viable for SOHO VPN applications, since the throughput exceeds the speed of most broadband connections, and therefore won't be a bottleneck.
Since the results were so good, I thought I'd run some other tests. I used Chariot to set up a test that ran the TCP throughput script simultaneously through the VPN tunnel and through the normal non-VPN LAN-WAN direction. Four computers were used for the test - two for the VPN tunnel and two for the LAN-WAN connection. Figure 5 shows that the VPN tunnel throughput dropped to about 1.9Mbps (lower plot line), while the data through the non-VPN path chugged along at a respectable 2.7Mbps (upper plot line). Not bad at all!
Dropped tunnels seem to be a common complaint for some of these inexpensive endpoint routers, so I thought I'd beat on the SX a little to see how it did. I again used Chariot, but this time set up a single pair that used the Chariot FTPget script, started it up, and let it run for an hour. This script simulates logging into an FTP server, downloading a 100,000 Byte file, then logging off...over and over again. While that was running, I carried on my normal Internet activities using another computer that was connected to another SX LAN port. The result was that the tunnel connection never skipped a beat, and operation through my non-VPN computer was trouble-free.
I also ran a similar experiment using one SX41 connected to a VP41. Throughput was, of course, limited to the lower 2.4Mbps of the SX41, but I didn't have any problems with this extended test, either.
While my experiments won't promise you a trouble-free VPN experience, I have to say that I was pretty impressed!
IPsec VPN Performance Test Results
|Test Description||Transfer Rate (Mbps)
[1 MByte data size]
|Response Time (msec)
[10 iterations 100 Byte data size]
[10s @ 500kbps]
|Actual throughput (kbps)||Lost data (%)|
|Local to Remote||2.4||4 (avg)
|Remote to Local||2.5||5 (avg)
|Firmware Version||1.43 May 22 2002|