In the past, we’ve focused our annual forecasts on electronics (Ic and EDA), but we’ve recently shifted our focus to photonics, so my 2020 forecasts are focused on this area. Historically, photonics has been the technique of gallium arsenide. Past, present and will always be the technology of the future. Analysts are always predicting the rise of photonics. Next year, as Moore’s Law ends or slows down in the electronics world, photonics will become one-of-a-year-high, into the hockey stick phase of its development. Although photonics is playing an increasingly important role in our technology ecosystem at an alarming rate, explosive growth has not yet been achieved.
Analysts are always predicting the rise of photonics; next year, as Moore’s Law ends or slows down in the electronics field, photonology will establish itself as it moves into the hockey stick phase of development (Hockey-stick Effect), which means that in a fixed cycle, early sales are low. There will be a sudden increase in sales by the end of the period, and in successive cycles this phenomenon will be repeated, with demand curves shaped like hockey sticks, so they are known in supply chain management as hockey sticks). While photonics is playing an increasingly important role in our technology ecosystem at an alarming rate, the expected explosion has not yet occurred.
Why is that? There are several reasons. First, photonology does not conform to Moore’s law: the wavelength of light is the wavelength of light. This is a constant. It’s not halved every two years, so the amazing advances in electronics driven by Moore’s law simply don’t apply to photonics.
Next, the electronics engineers will be as smart as ever, and every time Moore’s Law comes to an end, new approaches emerge, and engineers continue to break through what was once considered insurmountable barriers in the electronics field. As a result, the use of photonics instead of electronics is still being replaced by increasingly clever electronic design. Replacing electronics with photonics may still be inevitable, but the timeline is still moving forward.
Finally, over the past half-century, electronics has developed into today’s complex and well-functioning design and manufacturing ecosystem, but this development has not yet taken place in photonics. Today’s photonic ecosystem sysiphoto is still most similar to the electronic ecosystem of the early 1980s.
In 2019, more major companies acquired major optoelectronics suppliers, such as Cisco for Luxtera, and now Acacia Communications, and Finisar, the world’s leading supplier of optoelectronics products, Broadcom acquired their optical messaging assets from Foxconn. Many believe this trend is an acknowledgement of the advent of photonics and the need for leading photonic solutions from major telecommunications providers.
What about this year? Will the photonics society be the focus in 2020? Next we’ll predict several trends that affect when the inflection point comes.
In the shorter and shorter distances, photonics becomes more and more important, and then becomes mainstream, and finally becomes dominant. Today, telecommunications operators provide thousands of meters of long-distance communication to your home and business travel via fiber optics. Photonics has now been transferred to the data center. Large, hyperscale data centers around the world struggle with power, cost, heat, bandwidth, and data latency. Replacing copper wire with fiber solves all these problems. Optical fibers are cheaper, faster, have shorter delays, have higher bandwidth, and consume less power than copper, reducing heat and power costs.
Fiber takeovers between data center racks are almost complete, and fiber has been moved to interconnect servers in the same rack. As a result, photonology has evolved from a dominant position at one kilometer to a distance of ten meters to a distance of one meter. By 2020, photonic integrated circuits (PICs) will become more common in the market, making photonology meaningful within millimeter distances. The integration of photonics, including lasers, with electronic chips is under way to reduce the relevance of photonics to micron distances.
As Ethernet data transfer speeds continue to migrate from 100G to 200G by 2020, photonics is becoming increasingly attractive to transceivers on either side of the fiber. The expansion of 100G has been largely completed. By 2020, the transition to 200G will be smooth, and the transition to 400G will be smooth for early adopters. With clever engineering, electronics can continue to operate well at 100G, but photons are already competitive and there are many 100G photon transceiver designs on the market. We’re going to see more intelligence from IC designers, but at 400G, electronics will lose more control over the transceiver market, and photonics will begin to move from relevance to universality. By the time we reach 800G and 1T (Note: this will take a considerable amount of time to reach), photon technology will dominate and will have little access to electronic transceivers.
THE PHOTONIC DESIGN TEAM AT FANG (FACEBOOK, AMAZON/APPLE, NETFLIX, GOOGLE) WILL FOCUS ON ADAPTING PHOTON TRANSCEIVERS TO THEIR OWN SPECIFICATIONS, WHICH WILL ACCELERATE PHOTON’S DOMINANCE IN THE DATA CENTER. Operating a huge data center, they will benefit greatly from photon design, which meets their specific needs, just as we see fangs designing their own integrated circuits. By 2020, we may not see the fruits of this activity, but it will significantly improve the world’s photon design capabilities and accelerate the evolution of commercial photon ecosystems.
Today’s photon foundryis are dominated by small commercial or research and development fabs such as SMART Photonics, LionX, Ligentec, imec, Leti and AIM, which are typically targeted at research and development or MPW, rather than large-scale commercial production. While solid-state fabs are a leader in their optoelectronics, such as phosphated radon for lasers, they do not yet have the capacity to drive large commercial markets, and they do not yet have the opportunity to develop the extreme customer support processes that large semiconductor fabs have built over the past few decades.
FANG is a major customer of the world’s leading semiconductor foundry. These foundrys have noticed an increase in photon design projects and have entered or are considering entering the photon business. These plants will use their production knowledge, experience and skills to build mature ecosystems that will accelerate their commercialization.
In the past few years, leading semiconductor foundries such as TowerJazz and GlobalFoundries have begun to serve the photonics business. By 2020, other major semiconductor foundrys will enter the photonics business. The addition of these foundrys will accelerate the commercial ization of silicon photonics. One of the attractive advantages of photonics for semiconductor foundrys is that they do not require cutting-edge technology and therefore require large-scale electronics research and development and capital investment. On the contrary, photonics can make a handsome profit from fully capitalized semiconductor manufacturing equipment.
One indicator of the maturity of photonics is the emergence of the process development kit (PDK) for the foundries. The first photon PDKs appeared about two years ago, and they became popular as the foundry offered PDKs for a variety of design tools. These PDKs are still relatively primitive compared to the libraries provided by semiconductor foundrys, but they are an important and important step in the maturation of the commercial photonic ecosystem and require a strong partnership between the foundry and the photonic design automation (PDA) company. Together, they produce new PDKs and drive the development of existing PDKs. As more photonic ICs are manufactured and tested, more and more data can be used for statistical analysis. In 2020, a statistics-based PDK will appear, enabling Monte Carlo and process angle static analysis in more advanced PDA simulation tools. This will lead to more robust design, with a new focus on manufacturability, another requirement for the commercialization of photonics.
Mature EDA suppliers are noting the emerging optoelectronics market. They provide design tools for this market and form key alliances with leading PDA companies to provide a complete integrated design process. Last year, Mentor launched LightSuite Photonic Compiler, while using its Tanner tool to provide schematics and layouts. Cadence introduced curves through CurvyCore to enable its industry-leading custom design platform, Virtuoso, for photonics. Both Mentor and Cadence integrate their design processes with Lumerical, a leading photonic simulation provider. Cadence, for example, provides collaborative simulation capabilities that enable the entire design process to be driven through Virtuoso. Synopsys has adopted more independent strategies.
We expect the addition of major EDA suppliers to indicate higher prices for PDA tools. The average price of popular EDA tools is much higher than that of PDA tools. This imbalance is not sustainable in the long term, as it will hinder the required investment in PDA and the ability of PDA companies to compete in the new EPDA environment. Although the change does not happen abruptly, it will be consistent.
Last year there was an integrated electronic-photon design automation (EPDA) process that will become more complex by 2020 with increased statistical and manufacturing design (DFM) capabilities. Statistical considerations will require more computing power, so in 2020 we will also see high-performance computing applied to PDA, Amazon AWS, and Microsoft Azure as important players in providing photons in the cloud by leveraging all these photons in their data centers.
Compared to electronics, photon designs contain only a few elaborate components. Many of these components can be found in the PDKs provided by the foundry, but each cutting-edge photon design will always contain key components that the more common foundry PDK cannot provide. This provides opportunities for well-positioned companies to build photonic IP (PIP) businesses. Well-managed companies with superior photon design capabilities and focus, as well as low-cost access to design tools, are likely to drive the rise of this market. With their tireless efforts in photonic design, these companies will deliver superior designs. Companies that want to provide leading photonic designwillwills will work with these PIP vendors to outsource their component designs to focus their resources on other value-added areas, such as THE entire PIC design. In the early stages, the PIP business may maintain business similarity to custom design services.
Breakthroughs in photonic design methodology will provide higher quality, more manufacturable designs, and will reduce barriers, so photonic design no longer requires a Ph.D. in physics. Better design will drive photonics to compete and win applications. Qualified designers will give the company greater ability to equip its photonic design team, leading to greater competition, resulting in better products and faster development.
We’ve seen the impact of photon inversion design from sources from Stanford University and Lumerical working with the open source community. We are beginning to see that the design cycle can be greatly shortened by a significantly improved quality factor, and component design can be done more simply.
Even the best-released designs, we’ve seen improvements that usually take a few days to complete. We see an quantitative improvement in components. Photonic Inverse Design’s simplified automated design approach will replace today’s manual iteration process and will be applied to a variety of photon components by 2020. PhotonicInverse Design’s influence on photonics will be similar to the effect of logic synthesis on IC design. 1980s. This will expand the circle of qualified photonic designers and speed up the time to market for photonic design. We think the transition to photonic inversion is similar to increasing the level of abstraction of designer work. Just as increasing the level of abstraction in IC design frees up the workload of IC designers,
Application of photonics
Transceiver: The trend of photonology taking over in the data center will continue in 2020. This will be even more apparent in 2020 when we upgrade from 100G to 200G and reach 400G Ethernet transmission speeds.
LiDAR (lidar): In 2020, we will see the introduction of multiphotonic-driven lidar designs. Lidar is a key technology for autonomous vehicles, and a number of start-ups are focused on reducing liDAR size (reducing a deck of cards) and reducing costs (an order of magnitude), some of which will showcase their designs in 2020. In addition, we will see a liDAR based on photonics designed by at least one well-known leading LiDAR company.
5G: By 2020, we will actually see 5G built. This will facilitate mass production of photon chips, as new photon technologies such as NG-PON2 are deployed in both the forward and rear processes. With the deployment of 5G mm wave, there will be a hockey inflection point in 5G. But this broader 5G construction won’t really happen in 2020.
Sensors: It may be a bit boring, but this is an area where photonics is developing steadily, and this progress will continue until 2020. Medicine is a particularly interesting area for photoelectric sensors with great opportunities. Progress in the medical field will depend more on laws and regulations than on technology, and there will be no breakthrough in this area in 2020.
AR/VR: The easiest way to predict the future of our technology is to watch Star Trek. All of Star Trek’s technology will eventually be realized. This is good news for photonics. Photonics will play an important role if we are to travel through the fence walls.
Quantum computing: Quantum is another application that will drive photonics. 2020 will not be a quantum year, but we do expect at least one important quantum statement to shock everyone.
The future of technological photonics will be steadily developed in 2020. The growth rate will be impressive and a large number of applications will be the focus. The exhaustion of engineers’ ingenuity in the field of electronictechnology and the evolution of the photonic ecosystem will further constrain the growth of electronics. With the addition of commercial foundries and the maturity of design automation, signs of maturity are becoming more common. 2020 is the year of commercialization of photonics.