Slashdot Mirror
Harald Welte Calls Out Netgear's Open Source Sham
Simon80 writes "Harald Welte, known for his involvement in various open source communities, has pointed out the shortcomings of Netgear's open source router hype. Netgear's own astroturfed community site reveals that the router requires the use of binary-only kernel modules for the wireless and ethernet hardware, which is supplied by Broadcom. Also worth noting are the missing features in third-party firmware versions supplied by Netgear."
I use Buffalo router with dd-wrt.
Solid for two years now!
Share connection with my neighbor, hee hee.
One of the open firmware shortcomings is "WPA and WPA2 are not working." That is a pretty big shortcoming.
This post climbed Mt. Washington.
Shortcomings:
-------
* WPA and WPA2 are not working.
Well, yeah, that's a _minor_ one. Sigh.
I just ripped a broadcom out of a Dell Mini 9, bashed it with a hammer, and put in an Atheros. Maybe that is possible with this router?
I guess I will chugging along on wrt54gl with Tomato.
Gone!
I don't know many people who would care a lot about with what software a router comes. Anyone with basic knowledge about routers will put DD-WRT or OpenWRT on it anyway.
This sounds like the same argument that's been going around within the Linux kernel comunity about the "Open-Sourceness" of all these drivers which use binary-blobs - in partuclar, a lot of wireless Ethernet drivers, and stuff like NDISwrapper. It's idealistic to want all software to be open - but for companies which pour a lot of intellectual property into their drivers and firmware, I find it understandable that they wouldn't want their work made available to competitors' products. If they're not using any open-source in their binaries themselves, it's no violation. My opinion is this - if you don't like it, don't use it.
Are there any routers available that do have completely open firmware/whatever else?
If so, buy those, the end. Activism means nothing, hate sites filled with spam like the FSF produce mean nothing. Businesses don't care about those, because they don't have to.
As the saying goes, money talks, and bullshit walks. Give yours to companies who produce the types of products that you want to buy.
No flaming, no flaps, no noise, no controversy. It's simple, it's quiet, and it can be effective.
Comment removed based on user account deletion
I made the mistake of buying their KWGR614 "open source" router a couple years ago, and boy did it suck. The firmware delivered with it basically did not work. It would drop connections after 15 minutes of being on and then stop working. Everyone else who purchased one of these lumps of shit corroborated this behavior. Their employees denied it on the message boards, and in the end said "it's open source, fix it". Which is weak, because I bought the thing hoping to play with it when I got a chance, not in order to use it at all. Anyway I found a french language site that described how to build the this device's firmware, then I replaced it's broken DHCP and DNS capability with dnsmasq and was able to get it to work to some degree, though it had to be rebooted once or twice to be (more) stable. It still needed a reboot here an there.
I think Netgear just called this an "open source" router because they didn't want to fix the junk themselves. I decided never to buy Netgear products again after that.
LS
There is a fine line between being a cultivated citizen and being someone else's crop. - A. J. Patrick Liszkie
Did Netgear claim it was an _only_ open source router, meaning everything in it is open source? I don't recall them claiming such a thing.
So...strawman.
And way for the Open Source community to encourage vendors to use OS. Shrill complaints are always helpful.
Everyone that has converted a router to OSS, raise their hand....
Everyone else....leave the room.
Ha! Oh netgear, you just don't market to freetards. They'll aggressively attack any company that attempts to cater to them or their platform. They will not be happy until you have given everything away and are completely unprofitable, like Sun. So Broadcom won't let you gpl their drivers? That's too bad, now the FSF is going to hang you from the highest tree and make an example out of you for making those drivers available to be used by the community. Instead of reaching a difficult to market to segment, you're instead going to be attacked by the very people you sought to please.
You just don't cater to people who are looking to undermine and exploit your product, it's not worth it. Save yourself some development time and just give them an unsupported breadboard with no firmware on any chips, some rough documentation, and tell them to have fun "hacking".
What a hilarious waste of marketing money. Selling to an anti-commercial market segment...
Didn't we just witness android os having the same issues? Many important aspects being proprietary (proprietary google apps) as well as the fact that the OS can't boot without the proprietary binary drivers from each handset device?
If you're in the US, go to the japanese website, download the HK blob.
Hey presto, your FCC compliany and completely binary-blobbed obfuscated wireless card now breaks the FCC regulations.
And this isn't some third party (Linux developers) doing this: it's the very manufacturers themselves.
Yet for some reason, despite this overt and explicit method of breaking FCC regulations by what the owner of said kit does by the aid and abetment of Broadcom, Broadcom remains free from jail.
Now please explain why this doesn't make your argument a mockery.
I've been ranting about precisely the same things with what's touted as the first Internet Tablets based on OPEN SOURCE with the Nokia tablets. There's a LOT LOT LOT of closed source modules and even applications that you can't uninstall on them despite the Linux kernel and generally Debian flavor of the OS. Meanwhile everyone using the pre-(not even yet released)-N900 tablets are stuck with an aged 2.6.21 kernel because many drivers are closed and Nokia would rather make a new tablet than properly open up the tablet architecture even though they like to go on and on about what an erection they get when they think about open source. I'm becoming less and less convinced that the driver modules that Nokia proudly takes credit for opening up are even BECAUSE of Nokia so much as that chip manufactures are opening it up on their own because of the positively evolving effects and opinions around open-sourcing hardware drivers. What Netgear did here seems to feel almost IDENTICAL to what Nokia has been doing for a few years now--talking up open source to get the attention of the geeks and developers and put up an open facade. It's still a walled garden despite the open facade.
I moved to alix 2d3 with pfSense.
http://www.pcengines.ch/alix2d3.htm and http://www.pfsense.com/
The thing can run circles around a wrt54g without sweating.
Yah 54g is great, I used it for a long long time, however 3 boxes I had always had some kind of issues with 3rd party firmwares dd-wrt, openwrt even tomato. From hanging to dropping WAN on DSL, I stayed frustrated. In due time, I figured my frustration had nothing to do with me living in my mom's basement.
pfSense + alix has been rock solid.
The best part about my alix board based router... its fully supported by lots of opensource initiatives (monowall, pfsense, zeroshell etc) and come time for N, I just replace the minipci card on it.
He works for VIA, and they do the same thing...
To be fair on him he has tried to make progress, but after a few years of big talk there is still no open source way to use the full features of VIA hardware.
So don't buy VIA because of the fancy features in the silicon - cos there is a good chance that you won't be able to use them.
So Netgear is responsible for the fact that a third party distribution is lacking?
I suppose I'm to blame because you are a moron?
Persistent Volume manager for Kubernetes - https://github.com/dwimsey/openshift-pvmanager
Who have lots of spare cash (not having bought all the non-FOSS stuff) are geeks and therefore love their toys.
Yeah, imagine the idiocy of a tech market seller ignoring the "wealthy geek" segment for selling their technological toys...
YOU aren't allowed to make modifications.
If you do and break the regs, you are breaking the law.
They don't put Buffalo in jail for it.
So why would they put you in jail for having the specs to write the driver in such a way as to make it break the spec?
As long as they provide the source code to these binary modules, then it's open-source. But you have to remember that open-source doesn't mean you can modify the sources and replace the original binaries. If you want that, you need free-as-in-freedom. But people arguing for that are just nut-jobs with their silly ideals, and open-source is fine for us practical types, right?
Surely it's a good idea to harangue a company for not being "sufficiently" open source. What a great way to gently remind them to have a positive attitude towards open source!
A professional might "appease" them by congratulating them on a first step and encouraging them to open up further. Hah! We ideologues know that only 100% compliance to our definition of open source can ever be correct. Anything else should be ridiculed openly.
After all, we would prefer a company be completely 100% closed source rather than have some features be open. Death to all infidels!
some manufacturing managers got together over tea and decided what to do with extra chassis/components for 1q2010 that weren't going to sell anyhow. Netgear is attempting to create a market the same way any other company creates a market, but is being shut down quickly in this case because the community is well informed and the technology is distinctly fraudulent by our definition of the "open source" term they have decided to embrace.
Open source will prove an uncomfortable venture for netgear however one that marketing has identified as "worth a stab" in terms of revenue.
whats worse is anyone interested in this product to begin with will likely agree the entire thing can be accomplished with a craigslisted wireless router, an old desktop, and a few ethernet nics. Better crypto, functional WPA, and i didnt have to buy anything new. Sure its larger, but what netgear fails to realize is my rights matter more than the size of an appliance that doesnt work.
Good people go to bed earlier.
Currently we have an old white box server with FreeBSD set up as a gateway/proxy. It's about 5 years old and we've not done anything to it in 3 years, but it's cheap commodity hardware and it has a 600W powersupply that sucks down a lot of juice. We wrote our own software that gives people wireless access for 3 hours when they buy a drink (coffee shop). We're talking a 400Mhz AMD K6-2 with 256mb of Ram.
We'd like to put that software onto a router and have been looking at Single Board Computers, but have yet to find anything that we like. All it has to run is Linux/BSD with an AMP stack.
Anyone have a recommendation that would be low power. I've looked at beagleboard and wall wart, but really we need 3 Ethernet ports and a wireless card.
"The problem with socialism is eventually you run out of other people's money" - Thatcher.
If copyright says that it is a derived work from the GPL work, it needs to be licensed appropriate for the GPL work.
If there is any confusion about whether this work is a derived one or not, this is the fault of your copyright laws, not the GPL.
LOL!
Show me ONE jurisdiction that has clearly defined "derivative work".
The very same Broadcom blobs are included in dd-wrt. It must also be noted that dd-wrt is supposedly GPL software, yet the evidence in SVN clearly shows that a large portion of the code is Copyright evil corporations such as Intel and Microsoft and that these corporations have NOT given permission to use the code under the GPL. It is in many cases not even clear if they give permission to distribute the code at all.
9/11: Never forget it was a false-flag operation
Because it seems like 180 degrees different from what a professional like CEO should do: whatever is possible to make the most money for the shareholders, whether ethical or not.
So why the difference?
"After all, we would prefer a company be completely 100% closed source rather than have some features be open."
No, we'd prefer to have a company 100% closed source rather than closed source and lying about it.
My WRT54GS has been stable for fucking years, absolutely years, rock solid for yonks, working its buns off moving packets. A couple months ago, I decided I was going to look for a new router that could do everything my old 54 can do plus wireless-n at 5.8ghz (maintaining g at 2.4ghz) and gigabit ethernet. I had to look at the $250+ range and I'm not even sure if those units would do it because I didn't bother scrutinizing the specs at that price. It may have been necessary to move into commercial grade equipment to get everything I wanted. Screw that. I can just hang a 5.8ghz 'n' WAP off a gigabit switch and plug that into my old 54 for a lot less money and not have to worry about unknown bugs, stability, etc.
In fact, I'm about to pick up a 54GL for my grandfather. I made the mistake of thinking a $20 TrendNet would be fine for him since he doesn't need traffic shaping or anything beyond a basic wireless router. Wrong. Damn thing quits every 5 or 6 days like clockwork. He has to unplug/replug it to get it going again. A 54 is worth the extra money because it just frickin' works. Linksys really hit the nail on the head with that line. As long as consumer broadband is in the 10-20mbit range, I'm not going to waste my time trying other routers.
....cares? This is all theoretical bullshit, guys. Move on.
Yes, Netgear is to blame as that very same third-party firmware supports WPA and WPA2 on all other supported routers but not on Netgear's. But of course the GP is a moron because he expected Netgear to be able to ship a firmware with the functionality it normally comes with.
USE HOT GRITS WITH STATUE OF NATALIE PORTMAN (NAKED AND PETRIFIED)
True. I went through the trouble of obtaining a WRT54G v4 for my family's home network as the load (two users who like HTTP and email and two who like BitTorrent and games) caused most cheap routers to crap out on a daily basis.
The thing is rock solid. I only need to reboot it very occasionally, maybe three or four times a year, and never because it hangs. On the other hand I have a Samsung router extending the network to the upper floor and the damn thing hangs once a month.
When the next router becomes neccessary I'll look very closely if I can find something like the 54G again.
USE HOT GRITS WITH STATUE OF NATALIE PORTMAN (NAKED AND PETRIFIED)
A lot of firmwares, like DD-WRT, have issues with binary only drivers and programs. I ran into it with the nas process in DD-WRT a few months ago.
I had decided to move to WPA2 Enterprise. It sort of worked, but there is a long standing bug in DD-WRT relating to WPA2 Enterprise. WPA2 Enterprise depends on Radius. The nas process will only try a Radius server once. If it fails, then it won't try again. The only workaround is to kill the nas process one way or another. Then to make it all the more fun, the nas process is binary only.
I ended up having to go back to the official Linksys firmware for my WRT600N to get working WPA2 Enterprise.
Havoc Penington, the bane of my Linux desktop.
Say you're a real company, with a real closed source code base and you decide to make an open source push.
You're going to have some code that you can't open, either for legal reasons or patent reasons or internal politics or because it's heavily patched code that only makes sense to a few key employees, and you want to clean it up before releasing it. So parts of it will remain closed. Otherwise, do explain how you'd sidestep all the legal, technical and political issues, while running a company and delivering a product.
I don't even understand the "astroturfing" charge since Netgear is promoting itself on its website. But say this is somehow relevant, for the sake of argument. Your participating developers are going to be your employees, at first. Maybe for a while. Seeing as how they're paid and the community is not, they're going to be the dominant influence. But if you really think there shouldn't be any paid developers, you need to explain how a company can shift gears from a closed source model to an open source model without taking the time to build up a community, and while not scaring off investors because they just fired their development staff.
I'm not saying they're above criticism, and yeah, I find the out of touch marketing message irritating and I don't understand why marketing types suck at life so much. But the open source fanboys here sound even more clueless, which is saying something. (Incidentally, I have nothing to do with Netgear, don't own or recall ever owning any of their products, and I work on databases, not networking.)
As the developer of a popular fork of Tomato, I'd like to address a few points:
Not all features supported
Specific to their Tomato port:
1 > WPA is not working.
2 > There is no support of SAMBA server .
3 > NAS is accessible only through command line using ftp. No GUI support to
access NAS is available till now.
1: Presumably, WPA2 is, which means that this isn't a showtopper, just a big annoyance. There's actually only one missing feature here, WPA support. The rest would not be expected.
2/3: Mainline Tomato doesn't support any of this on USB-supporting routers anyhow.
Binary kernel modules
This is no different than mainline Tomato, which also relies on binary kernel modules. In fact, most opensource firmwares DO.
Looking at this from the perspective of one of the authors of Tomato/MLPPP (bonding multiple DSL lines using a fork of Tomato), only WPA is really of any concern, and even then, you can work around it by using WPA2. This router adds support for 802.11N, more (MUCH FASTER) RAM, and a far faster CPU (200 -> 480MHz, plus other architectural improvements). Considering that memory throughput/latency and CPU power are our main bottlenecks when bonding multiple DSL lines, this router remains quite interesting despite the lack of WPA.
Hear, hear.
I have a 54GL, which I put ddwrt on a couple years ago (maybe more, definitely before Christmas 2 years ago). I haven't rebooted it since then. Period.
Hell, I haven't even logged onto the device since then. Every time I consider upgrading to a 'n' wireless system, the reason I don't is because the current one works so well.
Help! I'm a slashdot refugee.
The arguments he makes regarding Netgear's binary blob can also be made about VIA's famous on-again off-again open-source stance. Where are the complete documentation for their Chrome chipsets?
"Apparently they either did not do that, or they chose to ignore the values/rules by which this community works" ... hmm, VIA does this all the time.
"or they had somebody with limited understanding to advise them" ... and I thought Welte was supposed to advise VIA on Open Source?
"If anyone has a relationship with Netgear and contacts to the product manager responsible for this product" ... people have been contacting Welte forever. Did anything positive come out of it?
"Netgear, you can do much better than that!". Right, VIA can't, yet Netgear can ...
Is there a router with the following:
1.802.11b/g WiFi (N would be a bonus but not essential)
2.Ethernet (dont need Gigabit, 10/100 is fine)
3.ADSL2/2+ support
and 4.100% open source software with NO binary blobs for Ethernet, USB, WiFi or DSL
My current router has all those features (except possibly only being ADSL1) but it has binary blobs for the WiFi and DSL.
Some problems:
1. They are proclaiming it to be open source, which is deceptive. It's "open source" except where it matters (device drivers/modules) from a maintainability perspective.
2. Their employees are astroturfing
3. Releasing open source drivers does not in any way reveal your chip mask and hardware architecture. Atheros' real competitors have access to electron microscopes and everything else it takes to buy a router off the shelf and copy chips exactly; simply keeping the drivers closed is not going to deter, say, realtek or broadcom in the slightest.
Besides, Buffalo is supporting open source through action (money) not just in press releases - beating Netgear to the punch by a couple of years. Netgear is just playing the "me too! Signed, metoo@aol.com" game.
For years Microsoft has been paying companies behind the scenes, under the table and in the court rooms, to keep drivers closed. Broadcom is still on the Microsoft payroll, its board is colluded by Microsoft friends and it is standard policy to pull this crap.
The direct effect of this is to keep Linux from getting on the desktop by making sure its hard to configure hardware by the lay person by obfuscating and keep the hardware manufacturers under Microsoft's thumb.
This is not just for wireless devices either.
-Hack
The WRT54G series all use Broadcom chips pretty much identical to the ones you'd find in Netgear routers. See here:
http://www.dd-wrt.com/wiki/index.php/Supported_Devices#Netgear
I am not proud to admit this, but I took a CCNA years ago, and I've built literally dozens of wifi networks using various combinations of off-the-shelf (or off-the-refurbish-list) routers and stock/modified firmware. I am a minor authority on the subject of cheap-ass consumer routers.
Broadcom is what you'd call a "fabless semiconductor company," which means they design chips but don't actually manufacture them. Almost all consumer routers you can find today use Broadcom-based system-on-a-chips, which consists of basically a CPU, flash and DRAM, ethernet interface, and half a wifi-radio, all crammed onto a single CMOS.
Broadcom designs the chip, someone else leases the design for the chip (and all the accompanying drivers) from Broadcom, then the person that leased the chip pays a third person who owns a CMOS fabrication plant to actually manufacture the chip. Then the chips get sold to yet another party, like Linksys, Netgear, Trendnet, Asus (my pick!), Buffalo, and others. The chip has several dozen wires hanging off the end of it, and someone connects them to various external ports or devices on the router: ten wires make a bank of five Ethernet ports, two or four wires are connected to one or two antennas (more if you have MIMO), more wires are connected to the status LEDs and buttons, et cetera. The end manufacturer is also responsible for providing firmware, which historically they've done by combining Broadcom's drivers with some code they ripped off from the Linux kernel (some manufacturers, like Asus and Buffalo, are reputed to be good about providing source code when they do this). Then they put it in a box with a compatible power adapter, slap a lame warranty on it (because many governments and retailers require them), and sell it.
The end result is that pretty much all the routers you can buy are nearly identical in every way except firmware. Furthermore, almost no manufacturer can actually be bothered to provide long-term support for these routers (why fix a broken routers when they can sell you a new one?), and since firmware development is by far the most difficult and expensive part of what the end manufacturers (eg Linksys) actually DO, this is the area where most consumer routers really fail.
(The other problem is that most Broadcom chips only have about ~100 MB/sec of memory bandwidth on chip, tops, which is obviously less than one gigabit per second (~125MiB/sec). This means that there are no consumer routers you can buy that are actually capable of routing a gigabit of traffic per second- at best they all seem to crap out around 160 megabits per second, in my experience (note: you have even less bandwidth when traversing the NAT gateway, particularly with traffic shaping enabled). This is mostly a limit inherited from the CMOS manufacturing process they use, I think - it's the same process they use to make DRAM and flash, and while it's cheap relative to the number of transistors you get, the resulting chips are rather slow compared to what you get with optical lithography.)
As for your grandfather's router, I suggest you try running BitTorrent on a computer connected to it, and see what happens when you quickly spawn hundreds of new TCP connections. I'm betting it'll choke, because the onboard NAT has to keep track of each individual TCP connection, and your $20 Trendnet router (which is probably quite old indeed, regardless of how recently you purchased it) probably isn't expiring old TCP connections for a good 12 hours. There's probably a way you can set the NAT TCP timeout value to something more reasonable, like 15 minutes (if it's not in the web-based interface, try downloading the config file and editing it with a text editor - I shouldn't have to tell you the risk from doing this). You can also look up DD-WRT,
I lurves Tomato. What does yours do better?
This is Broadcom's doing. Blame them.
Just.. what are you talking about ?
A binary module isn't a first step, it's what open-source trashes away.
You're talking about Linux support, right ? Netgear boxes are all running Linux (you're probably confusing with the desktop situation, where some devices have no non-Windows OS support at all).
Harald's post ends on a courteous note because he suspects the false advertising comes from their PR. Other than that, Netgear "open-source router" is no different from any of their routers, even less friendly towards open-source than some of their previous models.
Netgear "open-source router" is no different from any of their routers, even less friendly towards open-source than some of their previous models.
Or almost, Netgear provides instructions for compiling our own apps : http://www.myopenrouter.com/article/13860/WNR3500L-Open-Source-Guide-Resources/
That's more open, but not much to do with open-source.
Also from the same page: "WNR3500L is running Linux 2.4.20."
Anyway, that brings Netgear closer to Nokia claims e.g.
But what a nice selection of open-source friendly parts, dear!
Apparently they care enough about the geek market to try to appeal to the router-modders and to try to build some momentum with astroturfing.
Now if they could just figure out that it's cheaper, more effective and more reliable to just do it right then everybody wins including them. Do it right and you don't have to astroturf - the grass roots want to grow.
Help stamp out iliturcy.
I'd mod you up if I could. This is an extremely informative post and is the reason I am currently using a full computer (using pfSense atm) for my routing needs.
Are you sure broadcom is still less than gigabit throughput? I mean, broadcom is a company not a product and it sounds like you're mentioning one individual chip not an entire line up. For example. the BCM4718.. you think it can handle full gigabit duplex?
When N is final spec retail and I can find a router that will run tomato gigabit + N with the hardware and usb to support squid proxy caching as well as anything else I want to do then I will buy but honestly I haven't found a router that matches that criteria successfully yet. Maybe I should splurge and build an x86 board and be done with it. I like pfSense as much as I like tomato.
I may be missing something here, but why is the first label GNU?
I wish Welte enabled comments on his blog so I could post this there.
However, the article summary was enough to explain everything. Netgear is using Broadcomm chips. I've worked in the embedded firmware arena before; Broadcomm does NOT release its drivers under open source. You only get to see the source if you and your company lawyers sign really nasty NDAs, perferably in blood. I'm pretty sure the specs for programming the chips are under NDA, too. Netgear does not have a choice about releasing the drivers as binary blobs if they are using Broadcomm stuff. The only way to get open-source Broadcomm drivers is to reverse engineer them, and Netgear probably isn't in the business of reverse-engineering their suppliers product. Hell, they're probably contractually forbidden to do so.
You will never get a fully open source product from a vendor that buys from Broadcomm, until Broadcomm changes its policies. Period, full stop.
---dragoness
Sorry for the typos in my above post. I wrote it over a glass of brandy.
I mean, broadcom is a company not a product and it sounds like you're mentioning one individual chip not an entire line up.
I suppose I didn't make that clear. Broadcom has a huge line-up of designs, which cover nearly every possible combination of the following features:
8, 16, or 32 MB onboard RAM
4, 8 or 16 MB onboard flash
100 or 1000 megabit ethernet interface
b/g wifi in single-channel mode, b/g wifi in dual-channel mode, b/g/n wifi in dual-channel mode, a/b/g/n wifi in dual-channel mode, a/b/g/n wifi in dual-radio mode (dual channels on each radio)
And then there are MIMO variants of the above.
They also vary in clockspeed from about 125 MHz up to around 400MHz. I've noticed that throughput seems to vary pretty much directly with clockspeed, regardless of most other features on the chip, so I'm guessing all these chips are using the same CPU/memory/hardware to do their routing, with the only major difference among them being the speed at which the hardware is running.
In my experience, the best predictor of a router's actual maximum throughput is the speed they advertise for the wireless interface. Chips designed with b/g in mind will tend to crap out somewhere north of 50 megabits per second, whereas chips designed for 2.4 GHz wireless N (270 or 300 megabits per second) can get all the way up to around 160 megabits per second (even if they advertise "gigabit ethernet").
If you're curious, the advantage of having a router that supports gigabit ethernet without actually being able to handle a gigabit of traffic, is that you can still have a gigabit of traffic flowing AROUND your router - just not over it. There is some small utility in this, particularly if you have a gigabit hub or switch.
For example. the BCM4718.. you think it can handle full gigabit duplex?
Hmm, BCM4718... that appears to be a dual-band dual-radio gigabit chip used in routers like the WNR3500. I don't think I've ever owned or used a router with that chipset, so I can't tell you how it performs. I do have some routers based on the 478x series, however, and that's where I got the 160 megabit/second figure. (This is the throughput I measured by trying a very large file transfer between two computes capable of reaching at least 60MB/sec when directly connected. When you include TCP overhead, the router's actual performance is probably somewhat better than 160mbps.)
I'd guess the BCM4718 will get you past 270 megabits but probably not quite all the way to 540 megabits per second. For reference, I have an Airport Extreme, which I understand uses TWO BCM43xx chips (I've never disassembled it to check), and although I didn't manage to max it out, I'm guessing it goes just past 500 megabits per second (downhill and with a favorable wind) - and this takes 12VDC * 1.8 amps. The BCM 4718 probably uses a lot of identical hardware, so assuming the clockspeed is similar, you'd need similar power consumption to get similar performance. I advise you to look at the power supply. [insert sly-looking emoticon]