Luxtera, a global leader in Silicon Photonics, has enjoyed successful commercial production of its products for almost 10 years. The company recently scaled up its production quite extensively, and for good reason: many of Luxtera’s current products are in-use at and deployed by the world’s largest Cloud data centers, and delivered and validated by the world’s largest networking equipment firms. Luxtera’s business is growing, and the company is now in high-volume production.
Luxtera’s product family includes 100 Gb optoelectric transceivers, all single-mode and focused on applications in the 500m - 2km range. The public launch of Luxtera’s new 100 Gb PSM4 QSFP 28 module (LUX42604) was fully qualified and release to production in Q1 2016, and the company has already deployed more than 100k units, with plans to ramp up towards 1M units/year. This is a very significant development in the Si Photonics industry.
The Silicon Photonics market is expected to grow from a few tens of US$M today to more than US$1B in 2025, as shown in the last “Silicon Photonics 2016” Yole Développement report. To give an industrial perspective to this market, Yole Developpment recently spoke with Greg Young, Luxtera CEO, about the state of the industry, Luxtera’s current positioning, and the company’s future plans.
Yole Développement: After years of strong R&D effort and investments, we believe Si photonics technology is close to taking off as a solution to the interconnect bottleneck in data centers. Can you comment on this?
Greg Young: The global market opportunity for all optics is quite hot right now. Both Si Photonics and non-Si Photonics module manufacturers have tremendous opportunity to ship products into the cloud data center infrastructure and the overall communications infrastructure. Si photonics will benefit from this great demand, and the products that most effectively solve the customer’s problem will enjoy the greatest success. We believe that for some segments, Si Photonics delivers the highest promise in terms of value, especially in the 500m and 2km segments.
YD: We have identified SiPh as a very promising technology to further accelerate optical integration of sensors like LIDAR, which are highly desired for autonomous cars. Is Luxtera looking at non-datacom applications?
GY: Today we’re most focused on datacom applications.
YD: We see more startups coming, which is an encouraging sign. Do you think Si Photonics could be an exciting field for private investors?
GY: I believe it’s a dangerous space for private investors. The cost to deliver a product is extraordinarily high, and the technology lends itself to a broad set of applications that both do and do not make sense to invest in. And so extreme caution would be my first advice to private investors looking to invest in Si Photonics. The development time, cost, and complexity are big issues, and there’s a big risk that you’ll bet on a non-differentiated technology, or a technology with a market differentiation that is not so important. It’s easy to pitch, but if an investment opportunity turns out to be a quite small market, or doesn’t have enough incremental value over traditional optics to deliver meaningful return for investors, then you have a big problem. Many of the investors in Si Photonics companies have learned this lesson the hard way.
YD: Si photonics still poses technical challenges: laser integration, assembly cost, WDM, etc. How is Luxtera positioned to propose the best solution to these challenges?
GY: We have a laser integration strategy that we’re quite happy with. If you look at Si photonics, the first thing you have to think of is maturity level. The problems to be solved with Si photonics are very different than those in traditional optics, or semiconductors: there are also many different market segments, and different technologies associated with each of those segments. So each individual product, each individual market segment has different needs for laser integration and optical packaging. And so it’s not one size fits all. That said, we’re very pleased with the products we have, our approach, and our cost structure. And we have proven our ability to scale into high volume production. Our approach is an effective solution for our market segment.
YD: Do you believe Si Photonics has big potential for IC foundries in terms of optical wafer processing?
GY: The supply chain is complex, difficult, and filled with many issues that people don’t fully appreciate. Many Si Photonics applications lack the scale to make foundry operations profitable, and wafer volumes that fall below the minimum necessary to maintain a viable process.
YD: What can we expect from Luxtera in 2016 and 2017?
GY: Continued high growth and expansion of our product portfolio.
YD: This summer, Intel announced its entry into this market. Other companies are entering too. How do you see this competitive environment evolving?
GY: We congratulate Intel on reaching production for its first Si photonics product. They have a novel approach to Indium Phosphide on Si integration that many said could not be done. That said, it’s is still a baby step forward. The product, as we understand it, utilizes more traditional optical packaging, assembly, and discrete non-Si based components than Si photonics elements, and hence it likely has more of a traditional optical module cost structure. We believe that the journey forward for Intel to a more complete and cost effective Si Photonics solution will be full of some of the same challenges Luxtera has had to overcome. It will take time in order for Intel to complete their Si Photonics platform and to move their technology to a broader state of maturity. Regarding other companies, I don’t know of many others that have reached full commercialization of a complete Si Photonics technology platform.
President and CEO Greg brings over 20 years of executive management, marketing, and engineering experience in the semiconductor industry. His career focused on communications equipment applications and the utilization of leading edge technologies to drive broad based market transitions. Greg joined Luxtera in late 2007 to drive the commercialization of the company’s Silicon Photonics technology. He has raised over $200M in venture and industry funding. He has successful led the company through maturation of the revolutionary new technology platform introducing the world’s first Silicon Photonics commercial product in 2008, now having produced over 1Mu. With the introduction of market leading 100G products he has scaled the business to profitability and CFBE. Prior to joining Luxtera, most recently he was Vice President and General Manager of the High Speed Ethernet Controller business unit at Broadcom. While there, Greg led the growth of the Ethernet Controller business unit from concept to over $350M in revenue and the number one market share position. Prior to joining Broadcom, Greg was with Intel where he held several engineering, marketing, and business leadership positions in the networking and video business units. His most notable accomplishment was leading the growth of the 10/100 Ethernet Controller semiconductor business from $50M to $200M+ in revenue and number one market share position in under 3 years. Greg started his career as an inventor / entrepreneur in the early 90’s after attending San Diego State University with a focus in Applied Mathematics and Computer Science. He currently resides in San Diego with his wife of 16 years and four children.
Silicon Photonics for Data Centers and Other Applications 2016 - more
Have we reached the tipping point before silicon photonics grows massively in data centers?
Overview of Recent Photonics Trends|
Discover the presentation held by Jean-Christophe Eloy at EVG Technology Days, in Shefayim, ...
Thermal Imager with Microbolometer for Smartphone: Evolution & Comparison on the last Technologies Trends|
Discover the presentation held by Romain Fraux from System Plus Consulting during the Forum ...
Uncooled IR Imaging Market Perspectives|
Discover the presentation held by Eric Mounier during the Forum IR Imaging, the 8th of Septe...