ASUS Blue Cave AC2600 Dual-Band Wireless Router Reviewed

[sfx2000]

Updated review is up. Ranking improved only one position. But the router can handle gigabit wire speed (941 Mbps).

Some of the WiFi routing stuff was also off - haven't had a chance to go back and check the review just yet...

 

[Oscarrandle]

I've checked with ASUS that MUMIMO and AiMesh will be supported on Blue Cave via FW upgrade soon, so it can connect with my old 66U to form Mesh Network in the future, I think it's great. The Blue Cave in my home is still working fine for around one month without any line drop issues, I love it and I think it's a good buy.

 

[Oscarrandle]

Better performance since re-test, but maybe the next firmware update will show improvement. Not up to Asus standards. Beautiful design, but not functional. No mu-MIMO? or AI-mesh? Maybe watch and wait on this product!

ASUS says MU-MIMO and AiMesh will be supported soon via FW upgrade, looks like INTEL platform can still be supported.

 

[thiggins]

ASUS says MU-MIMO and AiMesh will be supported soon via FW upgrade, looks like INTEL platform can still be supported.

ASUS has not disputed those points in the review...

 

[thiggins]

ASUS has confirmed there is a bug that disables NAT acceleration by default, even though the control shows Auto, i.e. on.

They are issuing firmware to fix this.

 

[Dave in NM]

It

WEIRD.

Wait why does the UI show utilization for CPU Core 3? I thought the CPU was just 2 cores.

it has 3 cores

 

[microking4u]

Would love to see OpenVPN performance added to the test bench, as ISP service monitoring is something we should all be worried about! Anyone test weather there is any encryption acceleration on these new Intel Chips? Which routers perform the best currently with OpenVPN? Thanks!

 

[sfx2000]

Would love to see OpenVPN performance added to the test bench, as ISP service monitoring is something we should all be worried about! Anyone test weather there is any encryption acceleration on these new Intel Chips? Which routers perform the best currently with OpenVPN? Thanks!

I wouldn't expect great OpenVPN performance, as MIPS has not been a strong player there compared to ARM or x86 with regards to OpenSSL.

The GRX350/GSX550 are rebranded Lantiq chips - good at routing...

The Intel GRX750 is AMD64, but that's another discussion perhaps.

 

[CrystalLattice]

Would love to see OpenVPN performance added to the test bench, as ISP service monitoring is something we should all be worried about! Anyone test weather there is any encryption acceleration on these new Intel Chips? Which routers perform the best currently with OpenVPN? Thanks!

From Intel product brief: "Hardware acceleration for VPN/IPsec, TLS, storage, etc" https://www.intel.com/content/www/u...cted-home/anywan-soc-grx350-grx550-brief.html

 

[^Tripper^]

In a nutshell; something my wife would prefer but I’d rather skip.

 

[rrgrimm]

Living in a townhome, I have 30+ neighboring WIFI networks, some of which are right on the other side of the walls. This is the first router I have tried in the last several years that doesn't drop connections in this environment.

 

[WSL]

I am running a Blue Cave wireless router. I setup a bunch of Virtual Servers and am using DynDNS as DDNS. All works from WAN (outside), but NONE works within LAN (inside). This seems so strange as most people seem to have the opposite problem where as they work within the LAN but not from WAN. But mine is the opposite. Anyone has any clue?

 

[Internet Man]

I am running a Blue Cave wireless router. I setup a bunch of Virtual Servers and am using DynDNS as DDNS. All works from WAN (outside), but NONE works within LAN (inside). This seems so strange as most people seem to have the opposite problem where as they work within the LAN but not from WAN. But mine is the opposite. Anyone has any clue?

This is actually a common problem since most consumer grade routers don't support NAT loopback.

https://www.snbforums.com/threads/accessing-dynamic-dns-entities-from-inside-the-lan.13370/

 

[WSL]

Thank you! Yes I finally got the term 'NAT Loopback' from my google searches. So there is no easy way around it? The Blue Cave replaced my Merlin-RT-AC66U, which I was worried that it could die soon. Maybe I should just stick to the RT-AC66U for now.
WiFi is actually not important to me. I just need a fast (have a Gbps connection) WAN to LAN Router.

 

[Internet Man]

Thank you! Yes I finally got the term 'NAT Loopback' from my google searches. So there is no easy way around it? The Blue Cave replaced my Merlin-RT-AC66U, which I was worried that it could die soon. Maybe I should just stick to the RT-AC66U for now.
WiFi is actually not important to me. I just need a fast (have a Gbps connection) WAN to LAN Router.

I'd just go with a wired router like the Ubiquiti EdgeRouter Lite/X

 

[RMerlin]

OpenSSL benchmark. This router is definitely not gonna break any performance record in that area...

admin@BLUE_CAVE:/tmp/home/root# openssl speed -elapsed -evp aes-128-cbc
You have chosen to measure elapsed time instead of user CPU time.
Doing aes-128-cbc for 3s on 16 size blocks: 1549473 aes-128-cbc's in 3.00s
Doing aes-128-cbc for 3s on 64 size blocks: 411710 aes-128-cbc's in 3.00s
Doing aes-128-cbc for 3s on 256 size blocks: 116751 aes-128-cbc's in 3.00s
Doing aes-128-cbc for 3s on 1024 size blocks: 29612 aes-128-cbc's in 3.00s
Doing aes-128-cbc for 3s on 8192 size blocks: 3690 aes-128-cbc's in 3.00s
OpenSSL 1.0.2j  26 Sep 2016
built on: reproducible build, date unspecified
options:bn(64,32) rc4(ptr,char) des(idx,cisc,2,long) aes(partial) idea(int) blowfish(ptr) 
compiler: mips-openwrt-linux-uclibc-gcc -I. -I.. -I../include  -fPIC -DOPENSSL_PIC -DOPENSSL_THREADS -D_REENTRANT -DDSO_DLFCN -DHAVE_DLFCN_H -ffunction-sections -fdata-sections -DL_ENDIAN -DTERMIO -Os -DCONFIG_LANTIQ -DDEBUG_NOISY -DDEBUG_RCTEST -pipe -funit-at-a-time -Wno-pointer-sign -mips32r2 -mno-branch-likely -mtune=1004kc -DLINUX30 -DLINUX26 -DLINUX_KERNEL_VERSION=199272 --sysroot=/media/ASUSWRT/asuswrt/release/src-lantiq/tools/toolchain-mips_mips32_gcc-4.8-linaro_uClibc-0.9.33.2 -DOPENSSL_SMALL_FOOTPRINT -fomit-frame-pointer -Wall
The 'numbers' are in 1000s of bytes per second processed.
type             16 bytes     64 bytes    256 bytes   1024 bytes   8192 bytes
aes-128-cbc       8263.86k     8783.15k     9962.75k    10107.56k    10076.16k

Would love to see OpenVPN performance added to the test bench

It doesn't support OpenVPN. Considering the OpenSSL performance of that CPU, I would expect performance closer to an old RT-N66U than to any modern router.

 

[RMerlin]

The weirdest part of it: this is a triple core CPU... Confirmed both by /proc/cpuinfo and also by "uptime" showing processes on CPUs 0, 1 and 2.

Hardware-wise, this is a really weird platform. Three CPU cores, and 512 MB of RAM that barely gets used (I would have expected 256 MB to match the rest of the hardware on this router). Not sure what could use that extra RAM, beside caching for SMB sharing.

 

[Easy Rhino]

the boot log from the Chinese site spat out a linux message that the x
3rd core couldnt be used. maybe its special purpose? because even intel doesnt claim three cores

 

[RMerlin]

the boot log from the Chinese site spat out a linux message that the x
3rd core couldnt be used. maybe its special purpose? because even intel doesnt claim three cores

No, as I said, I see processes allocated to all three cores here:

Mem: 81936K used, 368016K free, 0K shrd, 440K buff, 23432K cached
CPU:  4.9% usr  4.3% sys  0.0% nic 90.4% idle  0.0% io  0.0% irq  0.2% sirq
Load average: 3.42 3.50 3.19 2/130 2411
  PID  PPID USER     STAT   VSZ %MEM CPU %CPU COMMAND
 1070     1 admin    S     8000  1.7   0  0.7 nt_center
 1048     1 admin    S     4516  1.0   0  0.4 protect_srv
  193     2 admin    SW       0  0.0   2  0.2 [kworker/2:1]
 1121     1 admin    S     6808  1.5   0  0.2 httpd -i br0
    9     2 admin    SW       0  0.0   0  0.1 [rcu_sched]
 2242 29354 admin    R     1876  0.4   1  0.0 top
  114     2 admin    SW       0  0.0   1  0.0 [kworker/1:1]
 1552     2 admin    DW       0  0.0   0  0.0 [mtlk_wlan0]
 1047     1 admin    S     9896  2.2   1  0.0 nt_monitor
 1334     1 admin    S     7764  1.7   0  0.0 mastiff
 7338  7323 admin    S     7760  1.7   1  0.0 /tmp/drvhlpr_wlan0 -p /tmp/wlan_wave/drvhlpr_wlan0.conf
 8160  8145 admin    S     7760  1.7   0  0.0 /tmp/drvhlpr_wlan2 -p /tmp/wlan_wave/drvhlpr_wlan2.conf
 1018     1 admin    S     7108  1.5   0  0.0 wave_monitor
 1313     1 admin    S     6904  1.5   1  0.0 networkmap --bootwait
 1016     1 admin    S     6772  1.5   2  0.0 /sbin/wanduck

Maybe that log was from an earlier, buggy release.

 

[Internet Man]

No, as I said, I see processes allocated to all three cores here:

Mem: 81936K used, 368016K free, 0K shrd, 440K buff, 23432K cached
CPU:  4.9% usr  4.3% sys  0.0% nic 90.4% idle  0.0% io  0.0% irq  0.2% sirq
Load average: 3.42 3.50 3.19 2/130 2411
  PID  PPID USER     STAT   VSZ %MEM CPU %CPU COMMAND
 1070     1 admin    S     8000  1.7   0  0.7 nt_center
 1048     1 admin    S     4516  1.0   0  0.4 protect_srv
  193     2 admin    SW       0  0.0   2  0.2 [kworker/2:1]
 1121     1 admin    S     6808  1.5   0  0.2 httpd -i br0
    9     2 admin    SW       0  0.0   0  0.1 [rcu_sched]
 2242 29354 admin    R     1876  0.4   1  0.0 top
  114     2 admin    SW       0  0.0   1  0.0 [kworker/1:1]
 1552     2 admin    DW       0  0.0   0  0.0 [mtlk_wlan0]
 1047     1 admin    S     9896  2.2   1  0.0 nt_monitor
 1334     1 admin    S     7764  1.7   0  0.0 mastiff
 7338  7323 admin    S     7760  1.7   1  0.0 /tmp/drvhlpr_wlan0 -p /tmp/wlan_wave/drvhlpr_wlan0.conf
 8160  8145 admin    S     7760  1.7   0  0.0 /tmp/drvhlpr_wlan2 -p /tmp/wlan_wave/drvhlpr_wlan2.conf
 1018     1 admin    S     7108  1.5   0  0.0 wave_monitor
 1313     1 admin    S     6904  1.5   1  0.0 networkmap --bootwait
 1016     1 admin    S     6772  1.5   2  0.0 /sbin/wanduck

Maybe that log was from an earlier, buggy release.

There was some discussion of the CPU specs in the Phicomm K3C thread. A site in Chinese said, "Network SoC: Intel AnyWAN GRX350 PXB4395EL1600 800MHz Dual Core Three-Thread"

At first I assumed that this referred to the offloading which Intel claims that each radio does but it really seems like the third core could be a result of Hyper-Threading.

Would you be able to confirm the clock speed of the CPU? Intel claims 1600 MHz but Phicomm support and the Phicomm UI show 800 MHz.