Ethernet MAC address

[Pszemol]

"My product is a sticky label with a MAC address on it."

)) I like it!

 

[Pszemol]

The $1650 is for a whole OUI (16777216 addresses). If you go to
http://standards.ieee.org/regauth/oui/pilot-ind.html there's a form
for an IAB which is only 4096 addresses and costs $550.

Thank you - I had to missed it. $550 looks much better.

 

[Spehro Pefhany]

Thank you - I had to missed it. $550 looks much better.

I like the idea of buying 16777216 for $1650 and selling blocks of
4096 for $550. Actually $99 would be fine too. ;-)


Best regards,
Spehro Pefhany

 

[Joe Chisolm]

When a company designs a device with Ethernet it needs to programm MAC in
it. I know IEEE assigns MAC numbers and charges for this fee.
I know also the meaning of bit 46th means:
=0 - address given by IEEE
=1 - address NOT given by IEEE
Does it mean there is a legal way to not go to IEEE for that number and
program the Ethernet MAC address with a random number with 46th bit set?
I know I would risk my address being not unique, but this is very small
volume production and it would not be a big problem. I am more concern
with a legal side of this issue. Is there a legal force to go to IEEE for
MAC? Anybody here experienced in manufacturing Ethernet device?

Reading from my IEEE letter where we got a OUI, they state there is no
guarantee that duplicate assignments wont occur.
For a small number of units you could use the Locally Admin bit set, but
you would need a way to let the user re-assign the MAC in case of collision.
This would not occur if they *do not* have any other locally admin'd MACs.
Use any number for the first 24 bits (with the local admin bit set) and
a serial number of something like that for the low 24.

Here are a couple of links that may help:

http://standards.ieee.org/regauth/oui/tutorials/lanman.html
http://standards.ieee.org/catalog/olis/802-1990.pdf

 

[Keith R. Williams]

Reading from my IEEE letter where we got a OUI, they state there is no
guarantee that duplicate assignments wont occur.

Thank you. This is a fact. Manufacturers have run out of
assigned blocks and are reusing them. It's not a big deal.

For a small number of units you could use the Locally Admin bit set, but
you would need a way to let the user re-assign the MAC in case of collision.
This would not occur if they *do not* have any other locally admin'd MACs.
Use any number for the first 24 bits (with the local admin bit set) and
a serial number of something like that for the low 24.

Seems to be good advice.

 

[Joe Seymour (Joseph M)]

: Thank you. This is a fact. Manufacturers have run out of
: assigned blocks and are reusing them. It's not a big deal.

How did we run out of two hundred trillion addresses?
Joe

 

[Keith R. Williams]

: Thank you. This is a fact. Manufacturers have run out of
: assigned blocks and are reusing them. It's not a big deal.

How did we run out of two hundred trillion addresses?

It's not a flat address space.

 

[Joe Chisolm]

It's not a flat address space.

The 48 bits are are divided into OUI (well actually
the L/G and Local Admin then OUI) and 24 bits that
the manufacturer can assign. So you really only
have 2**24 units until you either get another OUI
or use the address again.

 

[Keith R. Williams]

The 48 bits are are divided into OUI (well actually
the L/G and Local Admin then OUI) and 24 bits that
the manufacturer can assign. So you really only
have 2**24 units until you either get another OUI
or use the address again.

Right. That's 16M devices. Look at the list and pretend that no
one on it has manufactured 16M devices. Sure. They are *not*
unique. It does *not* matter.