Someone PM'd me this question, but I figured others might be curious about it, so I'm responding here...
Quote:
Why would 100 base T top out at 15Mbps? You mean just this device? I'm confused...... |
Okay, I should have been more specific. About 2 years ago, I did some real-world testing of 100bT throughput, with what was (then) very high-end equipment -- industrial servers running dual or Quad Pentium II Xeon processors at 400 MHz, 1 GB ram each, high-performance network hardware, etc. The tests were done with Windows NT (which had the fastest transport stack of any Microsoft OS available at the time.) With the overhead of the operating system, filesharing software, and apps (not many, just the minimum necessary to do a test), I managed to get 18 Mbps actual throughput. Now, I'm willing to bet that the Terrapin (or whatever it's called) portable HD can't sustain anywhere near the kind of throughput that these servers did. But to be generous, I estimated 15 Mbps, IF it has a 100bT port.
Interestingly enough, G3 Macs running OS 8.6 or 9 (I can't remember which) could easily sustain 26 Mbps in the same tests at that time -- their transport stacks were much more efficient. A Cisco rep commented to me that with the (then new) Mac OS networking software, Macs easily provided the highest throughput in their own internal testing of network gear.
Now, before anyone gets all up in arms, let me say that Win2K and XP have improved transport layers, and I'm sure they can sustain higher throughputs, especially on the faster machines available today. But I haven't done any real-world testing with them, or with the new Macs available now.
With apologies for getting way off on a tangent...