PLX chip carries 10GBase-T 300 meters

5/23/2011 12:01 PM EDT

Good to see that the CAT6a type cables are capable of carrying Ethernet data @10Gbits per second. But I wonder, was this only demonstrated in a clean environment, or this has gone through the necessary testing to prove its reliability in carrying data with expected level of integrity.

5/23/2011 11:24 PM EDT

Yes, nice to see the progress being made by modern silicon and copper cable design.

5/27/2011 3:45 PM EDT

CAT6A @10Gig has been around for years now and the debate has always been about power dissipation for Base-T ports. Many Datacenters wired up with 10Gig cabling, future-proofing themselves for the arrival of 10Gig hardware which were late in coming. From the looks of the news release, not much seems to have changed in power dissipation improvements since PLX acquired Teranetics! It is still at 4W per port (a small improvement, may be newer technology nodes may help; not sure what the TN8022 uses!).

@Rick Merritt: All that I can gather from the article is this: Base-T can address the reach challenges of 10GBASE-SR (300m) BUT at what cost? 16W per port is too much to user lower cost thermal solutions. The comparable optical solutions (with EDC) dissipate much less.

Also, @Rick Merritt, what was the wire gauge of the cables in the 300m reach demo, 24AWG? And the BER? Would be nice to know...

Dr. MP Divakar

5/27/2011 4:14 PM EDT

I am afraid that the information in this article could be interpreted incorrectly. One could leap to the erroneous conclusion that the PLX 10GBase-T transceiver has a higher power mode of operation (16W, as stated) that permits reach performance of 300 meters. This is not the case.
The PLX demonstration established the ability to span 300 meters without any increase in transceiver power dissipation. The per port power dissipation is approximately 4W.
In this demonstration, Power-over-Ethernet (PoE) technology was used to span power repeaters which were placed at 100 meter intervals and were span powered by a single transmitter. The key take-away here is that, at 4W per port, 10GBase-T technology is capable of implementing repeaters whose overall power requirements are in conformance to PoE standards and, thereby, do not require external power supplies. In addition, the demonstration showed that the noise injected by PoE circuitry on the copper line is rejected by the Digital Signal Processor in the PLX transceiver, assuring transmission with error rates that conform to the IEEE 10GBase-T standard.
I hope that this information corrects any misunderstanding.
Ron Cates – PLX Technology.

5/30/2011 6:58 PM EDT

@RonCates: thanks for the clarification. I am glad to reconnect with you after nearly a decade (I was also at Peregrine Semi's Optoelectronic Group where you were the VP).

Your response above leads to more questions: since the repeaters were powered in-band (16Watts for the 300m demo) leaving ~10W (with PoE+) for powering the edge device, is this really suitable for high-band width demanding applications like video surveillance? It seems like the data portion of your solution is well-answered but the power supply to the edge device is still an issue.

Does TN8022 re-clock at the repeaters? Does it also provide pre-emphasis and equalization?