new ASUS BRT-AC828 ( Wireless-AC2600 Dual WAN VPN Router)

[System Error Message]

Just wait till it gets to the UK, sellers in the UK are going to price this through the roof. If it costs more than £200 this is going to be a big no. Even small businesses in the UK wont spend much on networking equipment. The AC88U in the UK is priced at the AC5300 in the US.

The other issue, will wifi be locked? Will we get all the channels? Im not one who cares much about tx power rather i want empty channels i can use that the local ISP router brands wont use for some peace and quiet.

As for the CPU this is an improvement to the ARM A9 but broadcom using 64 bit ARM is going to be faster even for the smaller ARM 64 mainly because better per clock per core performance. Encryption performance would be great but openVPN software/libraries would likely not support hardware encryption yet.

I think CPU is moving too slowly in the consumer and SOHO market, we still see new routers that use single or dual caviums when there have been 8+ cavium MIPS cores available for many years now. As long as they dont call this router a 3, 4, 5,... core and call it a dual core unlike what they did with the broadcom routers with additional chips.

Maybe we should make our own router. Lets put the 9core tile, use SODIMM, mini PCIe, m.2, usb,...

2016 and many years of fibre optic available to consumers and still no SFP on consumer/SOHO routers.

 

[RMerlin]

FCC has a lot of info online now:

https://fccid.io/MSQ-RT0V00

 

[hggomes]

If it costs more than £200 this is going to be a big no.

That model estimated price will cost at least £350/400€ or more and you will only see it available on APR 17 on Europe.

 

[System Error Message]

That model estimated price will cost at least £350/400€ or more and you will only see it available on APR 17 on Europe.

Those $100-$200 tiny pfsense boxes are starting to look like a better deal.

 

[L&LD]

Those $100-$200 tiny pfsense boxes are starting to look like a better deal.

Cheaper, sure. But better? (I'd rather have the ASUS BRT-AC828 instead, assuming a budget can handle it).

 

[System Error Message]

Cheaper, sure. But better? (I'd rather have the ASUS BRT-AC828 instead, assuming a budget can handle it).

pfsense has some UTM features(and more features) that the asus doesnt have, not to mention a better CPU. Remember this is for VPN use.

 

[hggomes]

You mean Intel Atom/Celeron/Pentium boxes? Yes they are great speacially supporting 16GB RAM and CPU system power consumption of only 6-8W, but they miss the 8xLAN and 2xWAN on them, you would need a switch.

Also the wireless probably is not work as good as usual, but yes they are quite cheaper no doubts.

 

[hggomes]

Great box for the price: 160€ with full extras:

http://pcengines.ch/apu2c4.htm


I'm using Voyage Linux on it, works great.

 

[sfx2000]

You mean Intel Atom/Celeron/Pentium boxes? Yes they are great speacially supporting 16GB RAM and CPU system power consumption of only 6-8W, but they miss the 8xLAN and 2xWAN on them, you would need a switch.

And it's not just Intel/X86 boxes doing things this way - seeing more ARM boxes taking the same approach here.

Many are being focused as Router only, not as converged Router/AP's, along with switching... do one thing, do it well - be stable, be secure, be fast... there is an advantage in breaking things out - find the best of class for each function - and this is where some in the community are headed...

An external switch is not a bad idea - it's more to connect up, but once one outgrows the facilities provided by the "big honking router" boxes that most consumer AP's are, an 8 port managed switch isn't a big expense...

Also the wireless probably is not work as good as usual, but yes they are quite cheaper no doubts.

And there are some very capable AP's on the market these days... again, finding things that do one item well, vs. a compromise..

The business is shifting - back in 2014/2015, these was the development in the consumer AP segment - put as many ports, as many radios possible, along with every checkbox feature into a single box...

Thus we end up with abominations like the AC5300/AC3200 class devices that basically don't perform much better than a basic AC1900 class device - but it certainly adds to the bottom line, and the vendors have been very good at marketing this trainwreck...

It's encouraging to see that some vendors are breaking out of that mold - both incumbents as well as startups - let's see where the rest of 2016 takes up...

 

[L&LD]

If I remember correctly, maylyn has mentioned elsewhere that this will be announced/available in Q4.

We're there now, correct? Or is Asus' fourth quarter not aligned with the calendar year?

Either way, looking forward to this router with 10 ports (2WAN/8LAN) and built in M.2 SSD capability.

 

[Trentors]

Thus we end up with abominations like the AC5300/AC3200 class devices that basically don't perform much better than a basic AC1900 class device - but it certainly adds to the bottom line, and the vendors have been very good at marketing this trainwreck...

It is funny that everybody wants MU-MIMO because of the increased capacity. But the fact is that is it not working very well and it is difficult to find a compatible clients. Xstream routers are the perfect solution until MU-MIMO eventually works. They simply solve the capacity issue by adding an extra 5 GHz network. It is simple, works with every client (just avoid Smart Connect) and according to Tims testing they also seem to have give the benefit of faster 2.4 GHz.

Xstream is currently what MU-MIMO promises but cannot deliver yet. And yes with many clients two 5 GHz radios sure are faster than one.

I am seriously interested in what alternative to the BHR you recommend. Because after dipping my toe in the dedicated devices pool I came running back to using a BHR instead.

 

[sfx2000]

Xstream is currently what MU-MIMO promises but cannot deliver yet. And yes with many clients two 5 GHz radios sure are faster than one.

Broadcom is having some challenges with MU, but given enough time, they'll get it sorted.

My main issue with XStream is that it is not very efficient and provides little value to end-users... but that's my informed opinion - which may or may not be different from other opinions or experiences... while some might suggest that it improves capacity, I've found that most folks are not constrained on the capacity side, but more on the overall range side - and there, any AC1900 class Router/AP is going to be the best value solution...

 

[RMerlin]

If I remember correctly, maylyn has mentioned elsewhere that this will be announced/available in Q4.

We're there now, correct? Or is Asus' fourth quarter not aligned with the calendar year?

Either way, looking forward to this router with 10 ports (2WAN/8LAN) and built in M.2 SSD capability.

A quarter has three months. That means it could come out anytime between now and December 31st.

 

[L&LD]

A quarter has three months. That means it could come out anytime between now and December 31st.

True. Haha.

But they've already burned through the first 48 hours!

 

[Trentors]

Broadcom is having some challenges with MU, but given enough time, they'll get it sorted.

My main issue with XStream is that it is not very efficient and provides little value to end-users... but that's my informed opinion - which may or may not be different from other opinions or experiences... while some might suggest that it improves capacity, I've found that most folks are not constrained on the capacity side, but more on the overall range side - and there, any AC1900 class Router/AP is going to be the best value solution...

I am sure they will but that does not solve the fact that many clients do not support it yet. True that most peoples problem are not capacity - since most places in the world the WAN is limited to crappy ISP speed. But you could say the same for MU-MIMO so it is hardly a valid argument against Xstream. Even when it is ready on Broadcom based routers there is going to be a long time before MU-MIMO can solve any capacity issues unlike Xstream which works now.

 

[sfx2000]

you could say the same for MU-MIMO so it is hardly a valid argument against Xstream

Well - honestly... MU/XStream/Mesh - we're on the long tail - AC1900 is still the best bang/buck value solution - but if everyone has an AC1900 class router, then marketing folks need to prime that pump for demand...

 

[Trentors]

Well - honestly... MU/XStream/Mesh - we're on the long tail - AC1900 is still the best bang/buck value solution - but if everyone has an AC1900 class router, then marketing folks need to prime that pump for demand...

True. A good AC1900 router like RT-AC68U is enough for most peoples demand.
That does not mean there isn't room for a better product for people who want something slightly better. And the AC3100/AC88U, AC3200 and AC5300 perform better than the AC68U.

 

[W4RH34D]

I wonder if it can do multiple subnets over the ethernet ports. That's a pretty standard feature that is a requirement for business running credit cards and also an office + public wifi.

 

[RMerlin]

I wonder if it can do multiple subnets over the ethernet ports. That's a pretty standard feature that is a requirement for business running credit cards and also an office + public wifi.

Multiple subnets is worthless security-wise. What you actually need is multiple VLANs instead.

 

[W4RH34D]

Multiple subnets is worthless security-wise. What you actually need is multiple VLANs instead.

Worthless or not that's the compliance rules.

Having a firewall between subnet a and b is pretty nice to keep clutter down.

I never understood the allure of vlan. Maybe it's better in a virtual environment?

IDK especially for security - one switch with everything? Hmm IDK if I like that especially when port security is kinda a shirt-show with whether or not it actually enforces it.