
At a glance | |
Product | Western Digital My Net N900 HD Dual-Band Router (WDBWVK0000NSL-HESN) [Website] |
Summary | Ubicom-based "N900" router with Atheros radios, seven Gigabit LAN ports, USB printer and file sharing. |
Pros | - High routing throughput - Produced the highest wireless throughput seen yet - 20+ MB/s NTFS filecopy write/read |
Cons | - Better wireless throughput uplink vs. down |
Typical Price: $69 Buy From Amazon |
Overview
In our original My Net N900 review, I found a manufacturing problem in my review sample that caused very poor 5 GHz performance. WD said that the problem was confined only to early product that was sent to reviewers and not to stores. But it took WD until just last week to send a new sample for retest.
I reran the 5 GHz tests only and updated the relevant Router Charts wireless benchmarks, replacing the previous data. The summary and analysis with the new 5 GHz data follows. The short story is that the router is capable of very high uplink throughput in 5 GHz, but downlink performance isn't as strong, especially as signal levels drop.
Wireless Performance Overview
The retest Benchmark Summary is shown below. The overall average difference between two and three stream operation still isn't that striking because three-stream N can only provide significantly higher throughput under very strong signal conditions (same room or next-room).

N900 benchmark summary
So that you don't have to flip back and forth between the Benchmark summary in the original review, I've compiled the two sets of results into Table 1 and calculated the difference between them. It's obvious that the two-stream average results are all better and the three-stream all worse.
Test Group | Retest | Original | % Difference |
---|---|---|---|
2 stream 20 MHz Dn | 32.3 | 25.9 | +25 |
2 stream 20 MHz Up | 50.5 | 35.6 | +42 |
2 stream 20 MHz Up/Dn | 85.9 | 62 | +39 |
2 stream 40 MHz Dn | 40.5 | 35.8 | +13 |
2 stream 40 MHz Up | 63 | 40.2 | +57 |
2 stream 40 MHz Up/Dn | 108.5 | 86.2 | +26 |
3 stream 20 MHz Dn | 40.1 | 55.7 | -28 |
3 stream 20 MHz Up | 61.9 | 81.4 | -24 |
3 stream 20 MHz Up/Dn | 92.8 | 100.1 | -7 |
3 stream 40 MHz Dn | 46.3 | 65.3 | -29 |
3 stream 40 MHz Up | 73.7 | 126.3 | -42 |
3 stream 40 MHz Up/Dn | 116.1 | 120.5 | -4 |
Table 1: Retest and Original 5 GHz Benchmark Summary results
Table 2 summarizes the highest wireless throughput measured out of all locations in the original test runs.
Test Group | Max Dn (Mbps) | Max Up (Mbps) | Dn/Up (Mbps) |
---|---|---|---|
5 GHz, 2 stream, 20 MHz | 41 | 51 | 62 |
5 GHz, 3 stream, 20 MHz | 78 | 93 | 100 |
5 GHz, 2 stream, 40 MHz | 58 | 70 | 86 |
5 GHz, 3 stream, 40 MHz | 98 | 167 | 120 |
Table 2: Highest 5 GHz Throughput, original test
Table 3 summarizes the highest wireless throughput measured out of all locations in the retest.
Test Group | Max Dn (Mbps) | Max Up (Mbps) | Dn/Up (Mbps) |
---|---|---|---|
5 GHz, 2 stream, 20 MHz | 56 | 90 | 86 |
5 GHz, 3 stream, 20 MHz | 58 | 103 | 93 |
5 GHz, 2 stream, 40 MHz | 66 | 128 | 108 |
5 GHz, 3 stream, 40 MHz | 61 | 139 | 116 |
Table 3: Highest 5 GHz Throughput, retest
Comparing Tables 2 and 3 shows the same pattern as Table 1, i.e. performance improved for two-stream, but decreased for three-stream, with only a few exceptions.
The overall good news, however, is that I was able to run tests in Location A, C and D in all modes and with both the two and three stream test clients. And I didn't have any of the random disconnects I experienced with the original test sample.