Infinera’s Crack at Quasi-Reconstruction

June, 2017

Actions speak louder than words. The rather isolated introduction of the XTM II is a new attempt by Infinera to straddle the line between joining the rest of the community of optical system vendors, which determined a long time ago, the lack of merit with choosing a Photonic Integrated Circuit (PIC) approach, and the supplier’s stubborn adherence of continuing to push a PIC template in its marketing efforts. fibeReality still believes that these PICs not only fail to provide substantial differentiation, but can actually be counterproductive. Similarly, in our opinion, the small minority of financial analysts, who have been outspoken advocates of the “optical sypercycle,” which we have referred to previously as a “farce,” are striving to reconstruct their credibility, especially with the huge dip presently in the China optical space, as they are being forced to straddle the fence themselves, by unconvincingly pointing to Infinera as a sign of an impressive growth spurt in the market.

As we pointed out to followers of our second blog a couple of weeks ago, “Infinera could have better protected its position in the DCI space earlier by taking advantage of the XTM technology sooner, as we advocated in the past. Now the industry will deal with splitting hairs: [calling the XTM II a] 'lower range' DCI."

We use such a characterization because there would have been nothing to stop the ex-Transmode, metro-centric engineers from developing a 1RU box, which would have been at least in the running with the most state-of-the-art DCI solutions today. Ironically, Infinera’s countless expressions of regret about its supposed mistake in not developing its next-generation chip sooner, should have been a fixation about a different error.

For one thing, there is an emphasis by Infinera on the TMX II and its “Open Line System” capability, which is heavily pushed as a requirement, predominantly by the hyperscale data center operators. In contrast, the vendor’s only alternative with its PIC-based, “CX family,” is apparently still an unattractive band-aid scheme, interfacing with another supplier’s white box.

Secondly, Infinera has more than implied at times that its PICs were an essential enabler of its “Instant Bandwidth” offering. Yet, evidently, it was always about software, as the TMX II now has this capability as well.

It was probably no accident that Infinera announced the TMX II with the summer months approaching, and the possibility of only a limited number, if any, public events with financial institutions (with the exception of an August earnings call), until after Labor Day. A lot of questions about further development of a product line without incorporating its PICs will be avoided. Still, the insistence by the vendor that such a possibility is still on the table, when it was totally illogical with just the first generation of the TMX, demonstrates how trapped the company is with its historic rhetoric, and how potentially detrimental it will continue to be to its long-term survival.

Moving to the “optical supercycle” proponents, Infinera “insist[ing its] deal pipeline is the strongest in the company's history…,” [and] suggesting a reacceleration in demand,” is hardly a compelling conclusion. A more prudent explanation is that it predominantly reflects the pent-up demand from the tremendous amount of the company’s business being held up for a while from M&A activity by its customers.

Once again, the tremendous amount of irony is unmistakable. There would have not been any “megacycle” designation without the recent boom in China for optical components, which had a lot to do with aberrations, resulting from governmental and political activities, leading to constraints in purchases.

Please find details on our “Clash” market reports here.

[written by Mark Lutkowitz]

Share Article

Add Comment




Read Comments

S Reddy
September, 2017
More from the same
Vijay Bhagavath

Greg, the question of competitive advantage and more to come often in investor conversation, because this is a difficult space to invest in fairness, so lot of -- at least the longer term investors are looking to see, which of these names have a sustainable differentiation, have a deep motor on their business. How do you respond it from a Oclaro’s point of view?

Greg Dougherty

I think it’s very important that in this space that the technology differentiation is fundamental to any company. And when you really look at what's driving the advances in the space there’s two elements; one is digital signal processors technology, which Oclaro does not have, so just to give you a full answer. So DSP technology is one driver of differentiation. The second is the indium phosphide component three, specifically lasers, modulators and receivers; the ability to do more and more on a single indium phosphide chip is critical in terms of higher and higher speeds, smaller and smaller sizes and lower and lower power consumption, which is the three things that the hyperscale guys are all obsessed with. And so the two things to look for investing in this field are people that have best in class indium phosphide technology or DSP technology. Those are the two critical elements in future looking optical system in optical component tree.
You know who has the best in class Indium phosphide technology it is INfinera they need to increase cadence on DSP little bit. Interesting comments from component vendor.

S Reddy
September, 2017
Hi Mark what do you think of these comments by CEO of Oclaro this week at Deutsche bank conference with Vijay bhagwath transcript is on Seeking alpha

when asked about future of the industry and differentiation he thinks to get to 600g speed PIC on Indium phosphide is very imp infact Ciena buys them from Oclaro now they mainly develop DSP. You know Infinera is the best at this.
Interesting to hear what your take on this is?

Here are the excerpts..

Silicon photonics, it's clearly not going to be the solution for higher speeds. Silicon photonics, I think, you've seen the hiked down over the last two years. Silicon photonics has much higher los than indium phosphide. It's much more power hungry than indium phosphide, and requires external component tree like our beam amplifiers as you go to higher speeds. So basically, we've been able to do, for example, of 1.2 terrabits by integrating lasers, modulators and amplifiers onto a single chip, we're able to offer 600 gigabit per channel solution that doesn't require any additional component tree and has much less power consumption as silicon photonics solution.

So it's clear that as you look forward the indium phosphide will be much more important for higher speeds. As you look at very, very short lengths in parallel optics, that's where silicon photonics has some advantages, because there it's more concerned with -- the distances are so small that you don't need the performance and the loss, or the optical efficiencies as indium phosphide.