RT-AC68U reporting 1 ghz CPU

[escape75]

Looking at all the posts it just looks like it's a complete mess

B2 can be 800/1000, A2 can be 800/1000, C1 can be 800/1000
A1 is only 800, E1 only 1000 but who knows, very nice ...

 

[Wutikorn]

I don't know if that makes sure the wireless chips really are BCM4360, maybe they can still be the more advanced BCM43602 and using the same SDK (explaining still all RT-AC68/RT-AC1900 models share the same firmware).
Anyone?

I have AC68U rev C1, but has
kernel: eth1: Broadcom BCM4360 802.11 Wireless Controller 6.37.14.126 (r561982)
kernel: eth2: Broadcom BCM4360 802.11 Wireless Controller 6.37.14.126 (r561982)
exactly the same as blades'.

And it seems outside the USA and Canada only the A2 hardware version is available (even fresh made in 2016).
I am not sure if it is worth to mention CTF and FA support as well.

I bought mine in Thailand, and I got C1 revision. So in asia as well?

 

[wouterv]

Looking at all the posts it just looks like it's a complete mess

B2 can be 800/1000, A2 can be 800/1000, C1 can be 800/1000
A1 is only 800, E1 only 1000 but who knows, very nice ...

Finally, we talk about just a router (with 4 port switch, wireless accesspoint and more).
The minor varities shows we talk about a winner, ok Asus could have done better in my opinion, with just one product name and increasing hardware level coding. Technicians and marketing people often have conflicting interests.
Do you believe e.g. the iPone 6 only comes in one hardware variant over many production years?

 

[RMerlin]

Which seems to be the same like used in the RT-AC1900P which is clocked at 1.4 GHz according Asus.

Yes, except it's only clocked at 1 GHz.

I have an RT-AC68U HW revision C1 here.

 

[RMerlin]

I'm still sticking with my theory The later revs (at least C1 and E1) are board level revs that allow them to use whatever components are available to simplify mfg. From an ASUS code perspective, it doesn't matter since the latest code supports them all. From a consumer/3rd party developer perspective, it's a headache.

On the RT-AC68U you can check the CPU by the "cpurev" nvram setting. It will be set to "c0" if it's a BCM4709C0 (which then requires a newer SDK that was used starting with 3.0.0.4.380_xxxx).

 

[escape75]

The RT-AC1900P that bestbuy.ca is selling is a 1.4 ghz ...
Unless asus changed it again, but they still market as that.

 

[escape75]

On the RT-AC68U you can check the CPU by the "cpurev" nvram setting. It will be set to "c0" if it's a BCM4709C0 (which then requires a newer SDK that was used starting with 3.0.0.4.380_xxxx).

I wish that was the case, I don't get anything on stock firmware ...

escape75@RT-AC68U:/tmp/home/root# nvram show | grep cpu
size: 45634 bytes (19902 left)

 

[escape75]

I wish that was the case, I don't get anything on stock firmware ...

escape75@RT-AC68U:/tmp/home/root# nvram show | grep cpu
size: 45634 bytes (19902 left)

However ...

escape75@RT-AC68U:/# cat /proc/cpuinfo
Processor : ARMv7 Processor rev 0 (v7l)
[..]
BogoMIPS : 1595.80
[..]
CPU implementer : 0x41
CPU architecture: 7
CPU variant : 0x3
CPU part : 0xc09
CPU revision : 0

CPU Part is 0xc09 I'm guessing that doesn't mean anything ...

 

[ColinTaylor]

CPU Part is 0xc09 I'm guessing that doesn't mean anything ...

No, just means it's a Cortex A9.

 

[john9527]

I wish that was the case, I don't get anything on stock firmware ...

escape75@RT-AC68U:/tmp/home/root# nvram show | grep cpu
size: 45634 bytes (19902 left)

I took a quick look at the code....cpurev is only set for 'C0' processors. Otherwise not present.

 

[escape75]

I took a quick look at the code....cpurev is only set for 'C0' processors. Otherwise not present.

Ok so that makes sense, mine is a C1 revision but running at 800mzh since it's not a C0 ...

 

[escape75]

Also seeing this that makes me uneasy, but I was told it's normal ...

escape75@RT-AC68U:/# dmesg | grep Bad
Bad eraseblock 317 at 0x0000027a0000
Bad eraseblock 601 at 0x000004b20000

 

[wouterv]

Guess earlier conclusions still stay: Asus has released different hardware revions of the RT-AC68 board, maybe for different production locations and end markets, which allow them to use a mix of various components probably based on availability and commercial agreements.
As long as we do not see firm hardware revision related issues, I would say they do a good job (and just manage to confuse us).

 

[john9527]

Also seeing this that makes me uneasy, but I was told it's normal ...

escape75@RT-AC68U:/# dmesg | grep Bad
Bad eraseblock 317 at 0x0000027a0000
Bad eraseblock 601 at 0x000004b20000

As long as the number of badblocks doesn't start growing, nothing to worry about. I also have 2 bad blocks since I first bought my AC68R over 2 years ago.

 

[WaXHeLL]

I am trying to compile an overview, sofar as follows (I know some models are still missing, like to W model):

View attachment 7743

In short:

• There 3 CPU types (800 MHz, 1 GHz and 1.4 GHz).
• There are 2 types of radio chips, the one with its own CPU shall have a better performance.
• For the older radios, there are two amplifier versions, not sure if there is a performance difference.
• The R types are Reseller only, some other models are BestBuy only.

Shoot on it, and I will try to update the table

[EDIT1] RT-AC1900P CPU and Radio chips types added.
[EDIT2] RT-AC1900 added (no details known yet).

This is somewhat related, just because there is a hardware table embedded in this thread, even though I am running DD-WRT instead of Merlin.

I just bought a RT-AC1900 with the following details:

Model: RT-AC1900
FCC ID: MSQ-RTAC68U
Base Firmeware: 3.0.0.4.380
H/W Version: C1

Could not flash DD-WRT directly from the Asus config (gave me the FCC error message), but was able to flash it from the CFE -- flashed the latest Kong build (30880M).

The DD-WRT UI detects the router as a RT-AC68U C1, with the CPU as a BCM4708.

Telneting into the router, I get the following
nvram get bl_version : 1.1.1.2
nvram get clkfreq : 1000,666

dmsg reveals the following:
eth0: Broadcom BCM47XX 10/100/1000 Mbps Ethernet Controller 7.14.89.21 (r524987)
eth1: Broadcom BCM4360 802.11 Wireless Controller 7.14.89.21 (r524987)
eth2: Broadcom BCM4360 802.11 Wireless Controller 7.14.89.21 (r524987)

 

[escape75]

This is somewhat related, just because there is a hardware table embedded in this thread, even though I am running DD-WRT instead of Merlin.

I just bought a RT-AC1900 with the following details:

Model: RT-AC1900
FCC ID: MSQ-RTAC68U
Base Firmeware: 3.0.0.4.380
H/W Version: C1

Could not flash DD-WRT directly from the Asus config (gave me the FCC error message), but was able to flash it from the CFE -- flashed the latest Kong build (30880M).

The DD-WRT UI detects the router as a RT-AC68U C1, with the CPU as a BCM4708.

Telneting into the router, I get the following
nvram get bl_version : 1.1.1.2
nvram get clkfreq : 1000,666

dmsg reveals the following:
eth0: Broadcom BCM47XX 10/100/1000 Mbps Ethernet Controller 7.14.89.21 (r524987)
eth1: Broadcom BCM4360 802.11 Wireless Controller 7.14.89.21 (r524987)
eth2: Broadcom BCM4360 802.11 Wireless Controller 7.14.89.21 (r524987)

Because of nvram get clkfreq : 1000,666
it must be a BCM4709 I would think ...

 

[escape75]

I've decided to return the AC68U and getting an AC1900P,
bestbuy.ca has it on sale for $180 CAD, so about $133 USD

(Ouch it's $199 USD on bestbuy.com, weird ...)

At least that one will have a 1.4 ghz cpu.

 

[escape75]

Little bit off topic but I've noticed the official firmwares for RT-AC68U and RT-AC1900P are binary identical:
RT-AC1900P_3.0.0.4_380_3831-g93dfe8c.trx = RT-AC68U_3.0.0.4_380_3831-g93dfe8c.trx

 

[RMerlin]

Little bit off topic but I've noticed the official firmwares for RT-AC68U and RT-AC1900P are binary identical:
RT-AC1900P_3.0.0.4_380_3831-g93dfe8c.trx = RT-AC68U_3.0.0.4_380_3831-g93dfe8c.trx

Yes, they're exactly the same firmware.

 

[wouterv]

I made a wrap up post in the generic ASUS AC Routers & Adapters forum:
http://www.snbforums.com/threads/rt-ac68u-rt-ac68p-rt-ac1900-rt-ac1900p.35759/