- Show All
5 Photonics Technologies vying for business success
Why understanding market demand is so crucial to picking winners!
In June 2017, just over 170 scientists, researchers, government representatives and high-tech industry experts from 17 countries converged on the Dutch city of s’Hertogenbosch. They came to discuss next generation technologies the world is going to need in 2030 and beyond. That’s important now because we’re reaching the economic end of what’s popularly known as “Moore’s Law”. And particles of light (photons) rather than electrons will be the engine driving many new applications in communications and life sciences that we’ll soon take for granted.
Dr Michael Lebby worked with PhotonDelta on a comprehensive global market study as part of the preparations for the World Technology Mapping Forum in June. The full video of his presentation in 's-Hertogenbosch is now on line. Following the Photonic Integration Conference in Eindhoven recently, PhotonDelta's Jonathan Marks caught up with Michael Lebby for an update.
The technologies that are going to triumph are the ones that become the most scalable while also getting their costs down. It is unlikely that all five technologies will be used. Frankly, the datacenter customers don’t care that much about what is under the hood.
Michael, in a nutshell, why was such a study needed and what did it reveal?
When you put an effective roadmap together, you need to understand the directions where the customer wants the market to go, or the kind of price-points they are looking for. So, it is essential to have an accurate high-level non-competitive overview of where the markets are now, the trends in the community and what the forecasted revenues might be over the next decade. From that you can start to make estimates as to the types of products that will be needed, and some idea of which product platforms are going to lead that revenue growth in the market.
What trends did the numbers uncover?
Our market analysis clearly shows that there are vast demands being put on the Internet infrastructure to cope with an exponential demand for bandwidth. Global Internet Traffic is set to double in the next 3 years, from around 100 Exabytes per month in 2017 to an estimated 200 Exabytes a month in 2020. Multiply this by 12 and we are in the ‘Zetta-byte’ era with the view of ‘Yotta-bytes’ not that far away. Much of the growth reflects our increasingly video-hungry world, with huge growth in the stats for mobile video data. That means the global Internet infrastructure needs to work a whole lot faster, cheaper and greener to cope.
There is no doubt that integrated photonics rather than electronics is going to be key to the solution as Moore’s Law for electronics begins to saturate. Beyond Moore’s Law, we will require huge volumes of reliable, integrated photonic components in optical transceivers over the next decade. It turns out that it is very important to understand the cost performance metrics of these integrated photonics “engines”, if you like, inside the optical transceivers that will get used in the data centre market.
Picking winning technologies
Is it going to be an easy route?
No, it won’t be all plain sailing! I believe there are some important technical hurdles that need to be overcome. Because there are gaps appearing between what the market expects and what the research and subsequent chip and technology manufacturing can deliver.
There are five major technologies vying for a stake in the business opportunity.
- There is the incumbent indium phosphide (InP),
- there is a new technology called silicon photonics (SiP),
- we've got Gallium Arsenide (GaAs),
- we've got polymer photonics (PP),
- and we've got dielectric photonics (DP).
There are going to be hybrid versions of these technologies as researchers find ways to combine the best of, let’s say, Indium Phosphide with Silicon into a single system-on-a-chip, rather than trying the connect two chips side-by-side. The closer components can be put together, the lower the losses, the faster they can operate and the lower the power consumption. Furthermore, both polymers and dielectric materials can be integrated easily onto InP, GaAs, and SiP, which makes them especially useful to address innovative technical challenges.
But the key factor that will decide the winners and losers is the scalability of these various PIC programs.
And over the next ten years, the hyper-scale data centres operators like Google, Amazon, Microsoft, & Facebook have very clear targets for their cost versus performance. They believe it is in the region of 1 US Dollar per 400 Gigabit per second ($1/400Gbps): which in real terms means that an optical link will need to run at 400Gbps and have each transceiver at the link ends be cost no more than US$200 apiece. After normal profit margins, the bill of materials may well be less than $100 per transceiver – all operating at 400Gbps – which turns out to be a very tough target for sure.
So, the technologies that are going to triumph are the ones that become the most scalable while also getting their costs down. It is unlikely that all five technologies will be used, and frankly the datacenter customers don’t care that much about what is under the hood.
As power consumption of datacenters becomes a very serious environmental issue, photonics has a clear advantage over existing electronic solutions. But we're certainly looking closely at some of those technologies to judge their scalability. The “dollars per gigabit per second” metrics are going to be important deciding factors for those building datacenters for data communications folks.
You also hinted in your presentation that the future of photonics is not just about building chips for data centres.
Once the high speed, low power, integrated engine is created, then other markets will follow. Healthcare, automotive and consumer (mobile, other battery-operated devices) will each reap the advantages of light-based technologies. Today, although the technology is proven, optical technology is often outdated. The devices themselves are housed in large bulky boxes using discrete components (i.e. not integrated) which brings with it a hefty price tag to manufacture.
Datacom accounts for about 60% of the profits in the photonics market in 2017. Many other opportunities will grow once folks see the power of an integrated photonics engine the size of a “system-on-a-chip”.
But there are numerous projects coming out of the lab which can use light-enabled sensors to detect diseases without having to puncture the skin or guide a surgeon’s needle to ensure more accurate removal of tumors. On farms, veterinary technicians and inspectors are hoping to use photonics to detect pathogens in food.
The inSPECT project video
A new integrated photonics engine will certainly open doors for say, hand-held battery powered devices for PoC (point of care) in the healthcare industry. We also expect to see small intelligent sensing devices that can be accessed via the Internet generating huge amounts of useful research and monitoring data.
There is also interest from military clients who need smaller, lighter, greener integrated photonics technology for imaging and chemical detection. The defence industry is therefore utilizing the same metrics to guide more rapid development of some of their applications too.
How serious is the price-performance gap and how can that be fixed?
This gap is interesting because it only became apparent to us at the WTMF in June. A couple of very detailed photonics systems roadmaps were done in United States about three years ago. And we compared their predictions then with customer expectations. We noticed that the technology has not scaled as fast as people had expected. Chip companies haven't produced technology for example that is anywhere near 1 dollar per 400 Gigabit per second which is a figure being thrown out by the Datacom industry.
So, I think we need to look at new paths to get to those targets faster. For some, that may mean some radical thinking away from the integral step-by-step improvement we’ve seen in recent years.
So which technologies should public funders in regional governments and the European Commission be betting on?
In the Netherlands and Belgium there is an experience base of semiconductor chip production, and integrated photonics core skills and competences going back at least 20 years. So, it would be wise for this region to take advantage of that leadership by driving the integrated photonics technology to the next level.
Cross border collaboration is becoming increasingly important in what traditionally has been a very fragmented industry. And so, we’re seeing the emergence of “digital innovation hubs” where each hub is developing its own expertise cluster. In Bretagne, France that seems to be aerospace photonics, in Jena, Germany they’ve just opened a new institute which is using photonics to speed up infection diagnostics.
A lot of work is also going on to improve the toolkits to make these chips. Chip Design software is rapidly improving to bump up the reliability of chips going into production. And in Tyndall, Ireland the Science Foundation Ireland has built Europe’s leading expertise centre in photonics chip packaging.
Yet more intensive collaboration will be needed if Europe is going to stay ahead. There are massive investments being made in China, Taiwan and the USA in applied research into photonics as well as scaling up the chip manufacturing process.
You are now CEO at Lightwave Logic, which is described as a “public material and device development company based in Longmont, Colorado”. Lightwave Logic (LWLG) is commercializing it's P2ICTM organic polymer systems for a variety of electro-optic devices. Can you explain briefly what polymer photonics is all about and how it fits in to what we already know?
Polymers are so-called spin-on materials. They can be spun onto InP as well as SiP (even GaAs) PICs to create very high performance, very low power integrated photonics solutions. The polymers that Lightwave Logic have created exhibit high temperature stability. So, they are ideally suited for the heavy data applications markets such as datacenters - as well as high performance computing. Our company has already demonstrated 50Gbps capable modulators for integrated photonics and plans to further develop its P2IC® (polymer PIC) platform over the next few years.
The real fascination of Polymer based photonics is that fact that it is very low power and is capable of very high data rates – both qualities that are critically needed by the community.
Donald Trump & Photonics
The Trump administration seems to be unpicking many of the schemes launched by the previous Obama Administration. Is there a danger that some of the funding to AIM Photonics could be cut back?
I think the jury's is still out. But what you can say for sure is that the AIM photonics program was put in place because the Americans realized that Europe had a leadership position in integrated photonics and they needed to catch up. We’re now starting to see the fruits of that investment in places like Rochester and Syracuse, NY. But Photonics is a global industry, so I hope the collaborative spirit which has blossomed between those leading photonics research communities on both continents will be allowed to flourish. We need world class road maps to really propel both Europe and North America into leadership positions.
A final question. We heard feedback at the WTMF that not enough engineers are being trained or are available now to make sure that photonics can scale up as planned. What needs to happen?
Photonics does indeed have a major requirement for more engineers who understand integration technologies for PICs. Remember that the technical challenges in scaling requires a whole range of skills not only in designing the chips but also developing the packaging technologies to mount the chips onto modules. Our industry sector needs academics to work on the fundamental science. But there are already plenty of jobs emerging demanding high-tech engineers with commercial, industrial production skills.
The most interesting observation about our skill base is that integrated photonics is quickly becoming the ‘next generation IC’, much like IC’s or integrated circuits with silicon over 50 years ago. So, getting into photonics now is a brilliant career move.
Save The Dates
New dates and a location for the second World Technology Mapping Forum have been announced. It will be held on the campus of the University of Twente in Enschede, The Netherlands from June 20-22nd 2018. More practical details on the forum programme and hotel information will follow at the start of 2018. In the meantime we suggest signing up for the free WTMF Newsletter to stay informed of developments.