While I understand your enthusiasm, some corrections to perception are in order.
1st, Infinera has almost nothing to contribute to the Intel research into making an optical bus.
lets talk about their mighty InP technology advantage.
They make an array of detectors. 1 chip. they make an array of EML's on DWDM wavelenghts. Well, if you look at Santur's tunable laser, they make an array of DFBs. Not too far a leap for an array of EML's.
Those two arrays are the only part of their pict that is monolithically integrated. The rest of it is simply smart hybrid integration that anyone with 300M in investment could EASILY duplicate. They use an off the shelf mux/demux based on AWG. So, the mux is not part of the pict monolithic design. They pick off the incoming signals from the array of detectors and shunt them over to standard electronics for massaging, retiming, reshaping. They do have some small amount of special sauce in how they do this, but it is not monolithic. OFF the shelf electronics. After the cleaning up of the signal and the associated performance monitoring, the re-inject the signal into their array of transmitters. They are reach limited based on their technology. They use commercial dispersion compensation and EDFA to make sure they even reach 80-100km.
So their basic technology advantage from the InP platform is limited. What they do have is some genius for integration and for taking a somewhat fresh systems approach.
While I think they are a great company and will probably do OK in an IPO (lots of hype, lots of press, lots of mystical belief in them), I do not think their technology is all that "groundbreaking".
What they have could rather easily be duplicated by anyone with a couple hundred million to spend.
I do wish them luck and want them to succeed. I simply want to point out that Infinera is a systems play! Not a fundamental earth shattering technological InP breakthrough.
re: IPO Alert: Infinera Files Its S-1I am not sure what S-1 you guys have been reading, but Gross Margin was positive in Q4. There is a huge amount of deffered revenue that is keeping apparent gross margin very negative. The big revenue gain in Q4 is that they were able to recognize $40M of past shippments. According to the S-1, they Shipped $140M of equipment last year and recognized $53M in revenue.
I have not gone back and calculated their actual margins and net income but it will be a LOT better than the posted Income Statement.
I hold zero personal animosity for the leadership team at Infinera.
That's not to say I don't hold the actions of some of them in contempt, it's just that it's not personal, as in irrational or prejudiced or "grudge". It's based on their actions, not the color of their hair or whether they like me or I like them.
I have said before and repeat: I have several friends over there whom I have known many years, respect highly, and would happily hire, work with or for in the future. I also said a long time ago the dream of high yield low cost highly integrated monolithic anything was crazy and would auger in.
They scrubbed a few VCs clean, no big woop, part of the game, more power to 'em.
That money is gone, now they want to scrub the public. Big woop.
re: IPO Alert: Infinera Files Its S-1"There is a huge amount of deffered revenue that is keeping apparent gross margin very negative. "
Are you implying that they recognized cost associated with the deferred revenue but not the revenue itself? This is highly unlikely as accounting standards require that costs and revenue be recognized in the same period (there are exceptions but they tend to be rare).
While I think there is far too much negativity on this board, particularly from why who obviously has some personnel animosity toward the leadership team at Infinera, the S-1 was far uglier than I expected. 7% GM in Q4 despite a massive increase in revenue, high opex and extreme customer concentration (looks like Sycamore circa 2000) are all flashing red to me.
It will be interesting to see how the NASDAQ holds up in the next few months or many of the IPOs in the pipeline may have to be shelved.
don't get me wrong. I think there is value in Infinera's technology. But it is a mix of some InP integration (array's of detectors and lasers / modulators) along with some original thinking and elegance at a systems level. Fundamentally this is a systems play. they have a nice solution for some kinds of links. but not a ubiquitous replacement for all the other DWDM systems out there that have better performance for many kinds of links. I wish Infinera well on their IPO. It would be good overall for the industry I think. I have some doubts about it, but those are based on the financials that have been widely discussed on this board.
Regards Luxtera, from what I can see from the press releases, they have very interesting technology. But on some levels I see them pursuing a crowded field with their new entry into PON devices using their CMOS platform. I wonder if the applications space they were pursuing is taking too long to mature, hence the decision to jump into PON. My issue with PON customer premise optics is that it seems like a race to the bottom in terms of both price and margins. Too many players with too many technology platforms from simple TO can based TOSA / ROSA to PLC based approaches like NeoPhotonic and others are pursuing to what Xponent and Enablence is doing, to Luxtera's approach.
I am intrigued by their earlier stated applications: optical bus technology for high speed devices (10G x 10 or 20 optical lanes is far faster than the standard electrical bus used in most high end servers and other computing platforms).
I think Luxtera has a great team. But wonder about when the market will be ready for the optical bus stuff. This market could be gigantic.
I wonder as well as the original work they were doing on making an optical gyro monolithically in a CMOS chip. That could be very interesting device integrated into on board automotive navigation systems.
Makes me think the recent PON decision was to find some revenue as I am guessing the other markets are not maturing as fast as they would like. On the other hand, if their secret sauce reallly gives them a competitive advantage on pricing, then they may just have a winner. TBD. CMOS processes certainly could be a huge cost advantage if they figure out a low cost package to go with the OSA.
sailboat.
p.s. Dave: are you still doing recruiting? or are you into something else now?
re: IPO Alert: Infinera Files Its S-1I somewhat agree with the statement, "I think there is value in Infinera's technology". However, in addition to cutting edge, unique technology that actually works, there MUST be a sustainable market environment for the technology in question. If the market is weak for the technology, is there truly "value"?
If you look at Jim Duffy's article in Network World (Feb 19th issue, page 8), as well as, several other independent research reports, it appears there's still a substantial fiber glut; thus obviating the need, in most cases, for DWDM systems. Case in point - "...carrier revenue on wavelength services to deliver such content is expected to stay relatively flat since bandwidth supply remains plentiful." - Insight Research
Certainly, Infinera's revenues to date are minimally respectable for a privately company. But, for the general public and fund managers, I just don't see the net income, margin or any sort of appreciable growth - and thus, the appetite for an investor to look at this as a long term, or even short term growth investment. C'mon, how many of you out there can honestly say you'd take a position in this company and park it in your 401K for any amount of time?
Are there any system's companies making hundreds of millions or billions in the DWDM market?
While I understand your enthusiasm, some corrections to perception are in order.
1st, Infinera has almost nothing to contribute to the Intel research into making an optical bus.
lets talk about their mighty InP technology advantage.
They make an array of detectors. 1 chip. they make an array of EML's on DWDM wavelenghts. Well, if you look at Santur's tunable laser, they make an array of DFBs. Not too far a leap for an array of EML's.
Those two arrays are the only part of their pict that is monolithically integrated. The rest of it is simply smart hybrid integration that anyone with 300M in investment could EASILY duplicate. They use an off the shelf mux/demux based on AWG. So, the mux is not part of the pict monolithic design. They pick off the incoming signals from the array of detectors and shunt them over to standard electronics for massaging, retiming, reshaping. They do have some small amount of special sauce in how they do this, but it is not monolithic. OFF the shelf electronics. After the cleaning up of the signal and the associated performance monitoring, the re-inject the signal into their array of transmitters. They are reach limited based on their technology. They use commercial dispersion compensation and EDFA to make sure they even reach 80-100km.
So their basic technology advantage from the InP platform is limited. What they do have is some genius for integration and for taking a somewhat fresh systems approach.
While I think they are a great company and will probably do OK in an IPO (lots of hype, lots of press, lots of mystical belief in them), I do not think their technology is all that "groundbreaking".
What they have could rather easily be duplicated by anyone with a couple hundred million to spend.
I do wish them luck and want them to succeed. I simply want to point out that Infinera is a systems play! Not a fundamental earth shattering technological InP breakthrough.
sailboat