Do I need a gigabit ethernet connection?

It's not really sleazy marketing; it's taking advantage of customers' ignorance about the tehcnology.

The 500 mbps rating for the powerline adapter is its link speed, not its data transmission speed. That is, the powerline adapters talk with each other at up to 500 mbps. Due to interference and signal loss from path length, its actual data transmission speed is a lot lower. Exactly like how 802.11n has a link speed of 300-600 mbps, but its actual data transmission speed is closer to 100-150 mbps.

Wireless typically transmits data at 1/3rd to 1/5th its link speed because it has a ton of error correction to guard against RF interference. Powerline is even worse because your house's electrical lines are unshielded. They're basically gigantic antennas picking up every tiny stray bit of RF out there. Your appliances don't care about 100 MHz FM noise in their 60 Hz electricity. But it wreaks havoc with powerline ethernet. That's why they put 100 mbps ethernet ports on it - they knew its real data transmission speed would never be faster than 1/5th its link speed.

Go to www.smallnetbuilder.com and look though the router charts and the powerline charts.

Look how much a "lie" they have on wireless speed compared to tested rates.

They DO market switches like that. PoE is very common to not be able to run full power on all ports. When you look at switch speeds you actually do need to be careful of the rates. Say you have a 48 port gig switch. To really use this the switch needs to be able to pass 48g up and 48g down for a total of 96g of speed. If the back plane can not do this then the switch is not considered non-blocking or wire speed. You must really read the fine print. This used to be a huge issue in cheap consumer switches but now that the parts are so cheap you seldom see a 4 or 8 port switch that is not wire speed. Larger commercial switches though you still have to watch for this....especially on the ones that have 10g ports.

The reason I wouldn't go so far as to call it sleazy marketing is because the link speed info is actually useful for us techies. The link speed is an absolute max speed limit imposed by the technology being used. So I know that 500 MHz powerline adapters are using better (newer) tech than 200 MHz powerline adapters. This cannot be fudged by marketing.

The real-life data transfer speed is subject to a ton of variables - distance, background noise, device intercompatibility, wire composition and quality, error correction algorithm used, etc. As there's no standard way to test such speeds, each manufacturer would be free to come up with their own test. One manufacturer might use a really easy test (connect the powerline adapters with 6 inches of high-quality wiring!), while another comes up with a more realistic test (50 ft of 12 gauge wiring). You then you end up in a position where you have to carry this mental mapping - 10 MB/s in brand A is equivalent to 25 MB/s in brand B is equivalent to 3 MB/s in brand C - in order to compare the products.

So really, labeling these things with actual data transfer speed would be more confusing than not labeling them at all. I'd much rather the manufacturer just label the link speed, while I rely on third party reviewers (like smallnetbuilder.com) which test and compare all these products under identical, reproducible conditions. That way I know the MB/s data speeds are comparable between brands.

But I completely agree with you - there should be a huge disclaimer on these things that (for devices operating in noisy environments like powerline or wireless) actual data speeds will be nowhere near the link speed.


As I already mentioned, wireless has a ton of error correction applied to it to make it function in a noisy real-world environment. That's why you typically only get about 1/3 to 1/4 the link speed.

Hard drives actually function the same way. When you save a MB of data, you aren't writing one MB of data to the disk. That single MB becomes probably 1.5-2 MB of data actually written to the platter as magnetic data once all the error correction is applied. The difference is that HDDs operate in a constant environment where the signal's journey - from magnetic fields on the platter to bits sent to your computer - is consistent and constant for the life of the product. That's why you can label HDDs in terms of MB/s data transfer speed.

OTOH, wireless and powerline have to deal with completely inconsistent and always changing environments. So the MB/s data transfer speed is highly unpredictable, while the link speed is still constant.

I did not doubt you understood this I put this in for danageis.

Many people have not seen these charts that show real world test numbers....I really wish more people do.

It was only a week ago someone wanted to know if he need 10g ports on his router because the 802.11ac said it could do 3200meg. It is so sad the amount of education it takes to understand what that number really means.