The standard router storage test procedure was used to measure file copy throughput for FAT32 and NTFS volumes connected via USB 2.0 & 3.0. All the data is in the Router Charts; I'll be comparing USB 3.0/NTFS results only.
The P's more powerful processor propelled it to the top of both write and read charts. The R8000's 39 MB/s write and 74 MB/s read for the same benchmarks pale in comparison. Note the R8000 was not retested; the results shown were just copied from the original test. I spot-checked NTFS/USB3.0 performance and found it about the same as before, if not a bit lower.
USB 3.0/NTFS storage performance comparison
The R8000P was loaded with V126.96.36.199_1.1.74 firmware and tested with our V10 router test process.
|Test Description||NETGEAR R8000P||
|WAN - LAN Throughput (Mbps)||940||937|
|LAN - WAN Throughput (Mbps)||940||940|
|HTTP Score - WAN to LAN (%)||61||31|
|HTTP Score - LAN to WAN (%)||51||36|
|Bufferbloat Score- Down Avg.||617||478|
|Bufferbloat Score- Down Max.||372||113|
|Bufferbloat Score- Up Avg.||477||448|
|Bufferbloat Score- Up Max.||162||358|
|CTF Score (%)||59||42|
Table 2: Routing throughput
Both routers show gigabit wirespeed routing using our simple iperf3-based test. But the tougher HTTP tests show a significant difference in download and upload average performance.
HTTP Score averages
Looking at a plot of the benchmarks that make up the averages above, I added in the ASUS GT-AC5300 that sits at the top of both charts for comparison. The R8000P and ASUS are the same on WAN-to-LAN at the two largest filesizes tested. While neither NETGEAR equals the ASUS on the two smaller filesizes again on WAN-to-LAN, the gap significantly widens for LAN-to-WAN.
HTTP Score comparison
Plot key file size: [A] 2 KB, [B] 10 KB, [C] 108 KB and [D] 759 KB file
Comparing average Bufferbloat scores shows the P doing a bit better on average download than the 8000. But the table below with actual latency measurements added shows even the ASUS' relatively low uplink score equates to around 4 ms, not much to worry about.
|Test Description||NETGEAR R8000P||NETGEAR R8000||ASUS GT-AC5300|
|Bufferbloat Average Downlink Score||617||478||450|
|Bufferbloat Average Downlink latency (ms)||1.6||2.1||2.2|
|Bufferbloat Average Uplink Score||477||448||227|
|Bufferbloat Average Uplink latency (ms)||2.1||2.2||4.4|
Table 3: Bufferbloat performance
The Cut Through Forwarding tests look for throughput reduction when various router features are used. The P's CTF score of 59 means the router lost around 40% of its normal throughput when, in this case, either of its domain blocking or traffic meter features were enabled. The R8000 also suffered throughput loss when either of these features were used, but more like 60%. Interestingly, enabling QoS on the R8000 did not reduce throughput.
The R8000P is not Wi-Fi Certified. It was loaded with V188.8.131.52_1.1.74 firmware and tested with the Revision 10 wireless test process . The router was reset to factory default, then set to Channel 6 and auto 20/40 MHz bandwidth for 2.4 GHz and Channel 40 and 80 MHz bandwidth for the low band 5 GHz radio. The high band radio was shut off and not tested. WPA2/AES encryption was used for all connections. The Revision 10 process still uses 20 MHz bandwidth for 2.4 GHz tests for throughput vs. range, but uses 40 MHz for peak throughput tests. These settings are enforced by the octoScope Pal test client.
The router body was centered on the test chamber turntable with all antennas vertical as shown in the photo below. The 0° position for the router had the front facing the chamber antennas. Although you see four chamber antennas in the photo, only the center two are used for throughput vs. attenuation testing, which is done with the Pal set to operate as a 2x2 AC device.
NETGEAR R8000P in test chamber
We'll start by comparing average throughput to get an initial feel for how the R8000P measures up. Average 2.4 GHz throughput was in the lower half of the charts for both up and downlink.
2.4 GHz average throughput comparison
The 5 GHz average plots show much poorer downlink than uplink throughput. But 5 GHz downlink was also an R8000 weakness.