Like every other website on the planet, SmallNetBuilder uses cookies. Our cookies track login status, but we only allow admins to log in anyway, so those don't apply to you. Any other cookies you pick up during your visit come from advertisers, which we don't control.
If you continue to use the site, you agree to tolerate our use of cookies. Thank you!

Wi-Fi Router Charts

Click for Wi-Fi Router Charts

Mesh System Charts

Click for Wi-Fi Mesh System Charts

Stress Testing

Since its storage features and performance are central to the WNDR4700's value proposition, I decided to first look for interaction between routing and file transfer performance.

The test setup used our USB 3.0 drive formatted in NTFS attached to the router. I connected our standard NAS testbed computer to the CENTRIA and drag-and-dropped our standard ripped-DVD folder to and from the shared drive.

With the router otherwise idle, the Win 7 filecopy progress window reported around 30 MB/s writing to the CENTRIA's drive and around 36 MB/s reading.

I then connected two other computers, one to a LAN port and one on the WAN side and ran unlimited rate uplink traffic between them using an IxChariot throughput script. The IxChariot plot below shows uplink throughput being affected while the filecopy is running, evidenced by the dips and variation in uplink throughput.

Stress test - Filecopy to USB drive and ulimited router uplink
Stress test - Filecopy to USB drive and ulimited router uplink

While the uplink traffic was running, I checked the filecopy in progress and found that throughput dropped to around 3 MB/s from 30 MB/s for write and around 7 MB/s from 36 MB/s for read.

Since few of us have 500 Mbps internet connections and even fewer continuously load that connection, I also tried a more realistic test. I repeated the same tests, but limited uplink throughput to 5, 50 and 100 Mbps in multiple runs. In each case, the uplink throughput plot looked like the one below and the Windows filecopy progress window reported no change in throughput. So the takeaway from this test is that you won't see any interaction between routing and storage performance in real world use.

Stress test - Filecopy to USB drive and 100 Mbps router uplink traffic
Stress test - Filecopy to USB drive and 100 Mbps router uplink traffic

I tried a similar test to look for interaction between wireless and storage performance. I ran steady uplink traffic from a three-stream N client in 40 MHz mode on the 5 GHz radio and then ran the filecopy test. Frankly, given the higher natural variation in wireless performance vs. wired routing performance, I couldn't see any significant interaction between the two.

Finally, I checked for interaction between the router and two radios, with interesting results. The setup was two notebooks associated with the router, the three-stream N client on 5 GHz and a two-stream on 2.4. Both radios were set to 40 MHz bandwidth mode and running uplink for highest throughput. The routing portion of the test used the same test setup as used on the storage test above.

The plot below starts the 5 GHz stream first, then the 2.4 GHz stream, then the router. The initial change in 5 GHz traffic is previously-observed behavior, i.e. that it takes awhile for maximum link rate to kick in. When the 2.4 GHz traffic starts at the 30 second mark, it also starts out low and doesn't cause 5 GHz throughput to suffer. But as the 2.4 GHz throughput picks up, 5 GHz throughput drops.

The next significant change occurs when 5 GHz traffic stops at around 45 seconds in. Routing throughput picks up significantly, but 2.4 GHz throughput remains unchanged. Finally, the 2.4 GHz traffic is holding back routing throughput, since normal LAN > WAN throughput is around 500 Mbps.

Stress test - simultaneous dual wireless and wired routing
Stress test - simultaneous dual wireless and wired routing

I ran another test using different timings that more clearly shows the interaction between 5 and 2.4 GHz radios. In the plot below, 5 GHz throughput clearly drops when 2.4 GHz traffic starts 10 seconds in, then inversely mirrors 2.4 GHz throughput variation until its traffic stops. Note again that 2.4 GHz throughput is unaffected when routing traffic starts at the 30 second mark.

Stress test - simultaneous dual wireless and wired routing
Stress test - simultaneous dual wireless and wired routing

The upshot of all of this is that, like all other high-end routers I've tested, the WNDR4700's wired routing and wireless performance affect each other, as well as its storage performance to a lesser extent.

I'll emphasize that most users will never see these effects, since the test traffic I use is constant and rate unlimited and far above the loads that a home or even small business user would impose on the product. Of all the effects, I think the interaction between the two radios would be of the most concern, since it seems to kick in regardless of wired routing load. But again, unless you are able to put 100 Mbps+ of constant wireless traffic on a radio, which would take multiple three-stream N clients hammering away at strong signal levels, there is little cause for real concern.

Closing Thoughts

For people looking for a wireless router with best-in-class storage performance, it looks like NETGEAR has come up with one with the CENTRIA. Whether you use the internal drive or attached USB 3.0, as long as you format the drive NTFS, you'll get 30 MB/s writes and 60 MB/s reads—better than some entry-level single-drive standalone NASes can provide.

Unfortunately, the CENTRIA's best wireless performance is obtained running uplink rather than down, which will do little to help those looking for a better alternative for wireless HD streaming.

Support Us!

If you like what we do and want to thank us, just buy something on Amazon. We'll get a small commission on anything you buy. Thanks!

Don't Miss These

  • 1
  • 2