10GbE.TX — Crazy Development Costs or Just Right?

Yesterday GigaOm had a post on 10GbE.TX developer Aquantia raising $26M in their 2nd round of financing. Another 10GbE-Base-TX developer SolarFlare has raised ~$100M. I can’t seem to find the financing number for Teranetics.

I’ve always been curious about the market for 10GbE physical layer devices. Here is a “napkin” market analysis. The ethernet copper market (E.TX) market historically starts in big data centers with short connections from modular Servers and Switches that were already using optics, then it adds Switch-to-Switch connections, and later the market includes the much larger Computer Motherboards segment (LOM Lan-on-Motherboard) ( big because this is a 10/100/1000/10,000 … it goes everywhere). E.TX also rules the fixed form factor market which comes along around the same time as LOM. Today we are entering phase 1 – modular servers to modular switch connectivity – which is going to last for a few years.

A few weeks ago Cisco said the development cost of the 10Gig-centric Nexus has been ~ $250M. Dell’Oro says that Chassis Ethernet Switch market will be ~ $10B in 2012. The Nexus should get a big chunk of that. Dev cost of Nexus seems sane in this context.

A few years back Cisco said the CRS-1 development cost was ~ $500M. Dell’Oro says that Service Provider Router Market was ~$2B for 3Q07. ( ~ $8B this year ). Let’s say $10B for 2012. Dev of cost of CRS-1 seems sane in this context.

Returning to 10GbE.TX components. Back of the napkin market analysis indicates that the “Internet” Switch & Router Market is gonna be ~ $20B in 2012. The Physical Layer content is usually around 1%, or $200M. The 10GBE.TX segment will be ~ 30% ( if it works ). Thus my napkin’s scribbling indicates that 2012 Switch & Router 10GbE.TX PHY market is ~ $60M. The server side of these early connections is usually a bigger market. Let’s say it’s double for $120M. The “napkin” says that 10GbE.TX PHY market is ~ $180M in 2012 for phase 1.

❓ I wonder if I’m even close to the projections of 10GbE.TX vendors ❓ Anyways, the real market opportunity for 10GbE.TX is in the later phases when LOM and fixed form factor switches ship. Today 10GbE.TX is strategic. It’s not clear that the development cost of 10GbE.TX devices is sane in the context of the phase 1 market.


3 thoughts on “10GbE.TX — Crazy Development Costs or Just Right?”

  1. * thx Murat … that makes > $200M for these 3 and there are a few others too.
    * It is ironic that a “no-brainer” technology like 10GbE.TX doesn’t immediately turn into a “no-brainer” business.

Comments are closed.