2
\$\begingroup\$

I'm implementing an Ethernet system on a Xilinx FPGA, and I've been reading IEEE 802.3-2015, but nowhere have I found mention of RGMII which is the interface to the PHY. After extensive searching, I have been unable to find a source for where RGMII came from and whether a standard exists, I've just been able to find manufacturer's specs for the interface. So my question is how did RGMII come to exist and is there a common standard out there?

\$\endgroup\$
  • 1
    \$\begingroup\$ If you haven't already, try googling "Reduced gigabit media-independent interface" \$\endgroup\$ – Tut Feb 16 '17 at 15:03
  • \$\begingroup\$ I have, I know what it is and how it works, this question is more asking how did it come into being and if there's an actual standard, i.e. the history of RGMII. Did one manufacturer decide to produce a chip with this interface that they defined and it's just so happened to have been adopted by other manufacturers? Is there no formal standard produced by IEEE, IEC, ITU, etc.? \$\endgroup\$ – JMercer Feb 16 '17 at 15:09
  • 1
    \$\begingroup\$ Well you're not going to find it in a history book. Here is the standard referenced by Wikipedia. It would appear to have come from a small consortium led by HP: web.archive.org/web/20160303171328/http://www.hp.com/rnd/pdfs/… \$\endgroup\$ – Tut Feb 16 '17 at 15:23
  • \$\begingroup\$ Ah, perfect! I'd seen that document earlier and thought it was just HP's guidelines on it, not the original interface spec, but I never looked at the Intellectual Property section at the end, I guess that's why it's not mentioned in any of the Ethernet Standards documents. Thank you! \$\endgroup\$ – JMercer Feb 16 '17 at 15:27
2
\$\begingroup\$

RGMII was born the same way the original Ethernet was. Some interested parties (businesses) brought together and prepared a specification (a type of technical standard) for the MAC-to-PHY interface they were in need of.

Why did not they send it to a (inter)national standardization body like IEEE, ISO, IEC, ITU, other "official"?... My guess is because the "official" standardization process is (very) costly (and also not so fast) and the interested businesses did not see any extra profit resulting from the "official" standardization.

Despite of this, RGMII is very widespread today. Maybe it is not a de-jure ("official") standard but thers is no doubt that it is a de-facto ("useful") one.

\$\endgroup\$
  • \$\begingroup\$ RMII, the older non-gigabit variant, is also an "industry" standard not endorsed by the IEEE. \$\endgroup\$ – TEMLIB Nov 10 '17 at 22:56
  • \$\begingroup\$ @TEMLIB yes, the series of popular but not "officially" standardized Ethernet-related features is long: GPSI (SNI), Turbo MII, SGMII, QSGMII, Auto-MDI/MDI-X, etc... \$\endgroup\$ – asndre Nov 11 '17 at 14:16

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.