<<   <   Page 3 / 3
Stevery 12/5/2012 | 3:44:51 PM
re: Optical Vendors Face PIC Challenges Seven: Well, Infinera claimed around $55M/quarter in revenue last year.

Under software accounting rules. So from the point of view of the accountants, they are the only software company on the planet with their own fab. lol.

Cost of part = Cost of making part/ Yield

The overhead probably dominates, since there is a huge fixed cost of maintaining a class 10 (or 100 or whatever) fab.

In rough numbers, their annual fab costs are probably about $50M regardless of whether they produce 10 or 100 or 1000 good parts per day.
greenbug 12/5/2012 | 3:44:50 PM
re: Optical Vendors Face PIC Challenges Does someone have price comparison between Infinera and other vendors?

In the end, there yield issue would reflect on the product price. I heard that there price is not cheap, but maybe super cheap if /10.
hyperunner 12/5/2012 | 3:44:48 PM
re: Optical Vendors Face PIC Challenges Hi Craig,

"Aside from the two you mention, I wonder if there's a third possibility: That Infinera gets some advantage out of owning its own fab. Even if yields are low, they don't have to wait behind someone else to get another group of wafers processed."

I think you're right about that.

I remember a time when ALU and NT would crow about having their own component divisions - especially NT with the early 10G gear.

I guess they sold those divisions smoewhere along the way, so now they have to wait in line for components from JDSU and Bookham - just like everybody else.

I also wonder if a third party fab would ever invest the time (and high failure rates along the way) to actually get this PIC stuff to work?

Reminds me of that scene in "Edison the Man" where Spencer Tracy is trying hundreds of different filaments for the light bulb, and it fails again and again and again :-)

These days I sit through one of the competitors presentations and I have to keep looking at the logo on the slide coz everyone seems to have the same Powerpoint slides - except Infinera. Those other guys have lost the ability to differentiate their products except on price.

hyperunner 12/5/2012 | 3:44:48 PM
re: Optical Vendors Face PIC Challenges Guys, I think there's a risk that this thread will descend into the usual "Infinera's creative accounting" discussion.

I don't know the details or costs of their fab process but I've explained my (rather simplistic) logic as to why I think it's working for them as a business model.

In terms of the cost of their gear:

- No carrier on the planet will buy a system that is significantly more expensive than the competing bids. In most cases lowest bid wins, and in some cases if it's close you can justify a small premium. So Infinera's gear costs out at "about the same" as everyone else. They don't need to be cheaper because they're way better.

- Proof of this is in the market numbers and their revenue growth. They are continuing to win market share over every other competitor.

- After years of trashing Infinera's strategy, two of our existing suppliers are now talking about PICs on their roadmap.

ninjaturtle 12/5/2012 | 3:44:48 PM
re: Optical Vendors Face PIC Challenges If the fab ran them $50M a year they would be out of business long ago. Never made an IPO. But then again Stevery you didn't think they couldn't make this happen 2-3 years ago and as usual you have been proven wrong. Most of your comments are very anti-INFN based so your comments are fairly worthless when it comes to INFN. INFN is the hottest company in the Optical space. Name another company that has gotten as much attention and praise in the last 12 months?
litereading 12/5/2012 | 3:44:45 PM
re: Optical Vendors Face PIC Challenges hR - can I assume you are referring only to 10G DWDM?
hyperunner 12/5/2012 | 3:44:44 PM
re: Optical Vendors Face PIC Challenges Hi litereading,

"hR - can I assume you are referring only to 10G DWDM?

Yes and no.

Obviously 10G is where the action is today, and for the near term future. Two concerns about 40G:

- Demand for 40G services is really low (mainly because the cost of 40G router interfaces is ludicrously high...like more than 4 x 10GE)
- It's unclear to me, and most of my colleagues, what the longer term prospects for 40G are vs 100GE

...so we're somewhat cautious about designing a next gen infrastructure around 40G waves.

Everyone's 40G story is the same:

- Use some exotic and untried modulation scheme so that 40G waves look like 10G waves.
- Let's see how many letters we can cram into an acronym to describe that modulation scheme!

Actually I'm not sure how much of their 40G story Infinera has made public so I don't really want to comment specifically on that.

Vent 12/5/2012 | 3:44:43 PM
re: Optical Vendors Face PIC Challenges I find it difficult to comment on the yields of the infinera PIc. What does it contain
in terms of pin diodes and lasers (wavelengths ?)
and other circuitry integrated in InP?
hyperunner 12/5/2012 | 3:44:40 PM
re: Optical Vendors Face PIC Challenges Hi litereading,

I guess I went to 40G because:

- 2.5G is irrelevant in the LH
- 100G is sci-fi for the next 2 years at least

So what's left?

Are they shipping 40G? Sure - as an i-mux implementation. Is it what we're looking for? Nope. Because it doesn't address the fundamental issue - 40G interfaces on the router are WAY too expensive. So right now I don't see how any optical vendor's 40G offers a solution to that.

Sorry, I'm not a basketball fan, what are you getting at?

litereading 12/5/2012 | 3:44:40 PM
re: Optical Vendors Face PIC Challenges hR - it's interesting that you went right from my comment to 40G. Wonder why?

So INFN has a 40G story, do they - are they shipping it?

Since you spent so much time researching me, let me a question about you. Are you like a NBA basketball player from NY who plays above the net?

<<   <   Page 3 / 3
Sign In