disruptive technology - vladimir kozlov
Is Silicon Photonics a disruptive technology?
Silicon photonics is showing great potential as a new communications technology, with its blend of optical and low-cost CMOS semiconductor processing. It enables a new breed of monolithic opto-devices but will it deliver enough to disrupt the dominant technologies, asks Vladimir Kozlov.
per month and generate a profit. This will be hard for any new technology to beat. Optical interface modules for 100G DWDM are the most complex in the industry and integrating some of these functions makes a lot of sense. Such integration does improve port density, but it does not guarantee lower cost. Many equipment vendors are still using discrete 100G DWDM optics to minimise cost. There is very little doubt that silicon photonics will find a place in the market. There will be many products benefiting from this technology, but it is unlikely to challenge the dominance of InP and GaAs optics in the next 5-10 years. Predicting technological advances on the longer time scale is for science fiction. Almost anything can happen. Increasing funding for silicon photonics around the world is very encouraging. It gives a chance for engineers to build new devices and discover new applications, which we can not even conceive of yet. Market research works well for estimating the potential of existing technologies, but it is up to engineers to discover and launch new industries. The optical communications industry owes its existence to such discoveries made not that long ago.
manufacturing cost, enabling very compact and power efficient optical interfaces would probably do this. None of the current products combine silicon photonics and electronics on the same manufacturing platform. Yes, it is all CMOS, but optics is made in older 90-130 nm fabs, while high-speed electronics requires 22 nm processing. It is unlikely that we will see products combining optical and electronics on the same silicon wafer in the next 3 years, but some vendors are certainly working on this and may be successful in the next 5-10 years. Lowering the cost of manufacturing has been the most publicised advantage of optical integration. However, none of the successful integrated products delivered significant savings in cost of production, testing and even packaging. Integration has been more successful in increasing functionality, improving port density and power consumption than reducing manufacturing cost so far. Suppliers of FTTx optics achieved very impressive cost reductions using conventional discrete lasers and detectors. For example, selling prices of GPON bi-directional subassemblies are only $10 now. Leading vendors make these products in millions
amplifiers by Raman because of the number of performance advantages. This never actually happened, but Raman amplifiers evolved into a key complementary technology. Xtera Communications is an example of a pure-play company focusing on this technology, but all major equipment vendors use Raman amplifiers today. Optical integration enabled several commercially-available products of the last two decades. Electro-absorption modulated lasers (EMLs) is one of the first success stories. Indium Phosphide photonic integrated circuits, brought to market by Infinera, is a more recent example. Where these developments truly disruptive? Would the industry be radically different without them? Probably Silicon Photonics technology launched several successful products over the last few years. Luxtera’s 40G optical engine is one the early examples. Lightwire’s modulators, now used by Cisco, is a more recent one. Many other vendors joined the silicon photonics race with 100G products in 2014-2015 and they will also have a chance to prove themselves in 2016. Is silicon photonics a truly disruptive technology? Just like some of the examples discussed above, silicon photonics modulators offer a number of performance advantages. What would it take for this technology to really change the optical communications industry, replacing InP and GaAs products dominating the market today? Combining optics and electronics on the same manufacturing platform, drastically reducing not. Were these products commercially successful? Absolutely!
Vladimir Kozlov Founder and CEO of LightCounting S ince Clayton M. has embraced the concept. Hardly any new venture is now started without planning to bring a new disruptive technology to the market and impact the world. However, does a technology need to be disruptive in order to be successful commercially? It is hard not to get excited by the success of silicon transistors and digital cameras, both of which truly changed our world. Invention of the optical fiber and semiconductor lasers are among selected few technologies that essentially enabled the optical communications industry. Each one of these major inventions was followed by a long string of innovations. Very few of them were truly disruptive, but many reached commercial success. Success came early to inventors of optical amplifiers and DWDM technology, or at least it seemed so from a distance. Many other inventions took longer to prove themselves, reach commercial success and make a difference. Very few of these were really disruptive, but just as important for the industry in complementing the existing technologies. The Raman optical amplifier is one example. It was considered to be disruptive to the optical amplifier market back in 2000. Many industry analysts, including the author of this article, were anticipating complete replacement of Erbium-doped Christensen introduced the notion of disruptive technologies, industry
Silicon Photonics in Numbers
$300
Investments Product sales
$200
$100
$0
2010 2011 2012 2013 2014 2015 2016
ISSUE 6 | Q1 2016 22
Silicon Photonics in Numbers. Source: LightCounting Market Research
Made with FlippingBook - professional solution for displaying marketing and sales documents online