Trending News
Home / Technology / Nanotech / Silicon Photonics breakthroughs enable large data centers, Real time cloud-computing, Exascale supercomputers and cognitive computing

Silicon Photonics breakthroughs enable large data centers, Real time cloud-computing, Exascale supercomputers and cognitive computing

The technology industry is entering a new era of computing that requires IT systems and cloud computing services to process and analyze huge volumes of Big Data in real-time, both within data centers and particularly between cloud computing services. This requires that data be rapidly moved between system components without congestion. The enormous growth in both data storage and the demand for high-performance computing is also leading to the requirement of much higher bandwidth density for inter-chip communication than ever before (expected to surpass 40 Gbps per interconnect by 2020).


Equally important, data centers are faced with power consumption challenges. Bandwidth demand for computing approximately doubles every two years. However, electrical performance-scaling is not keeping pace with bandwidth demand. Moreover, electrical I/O energy efficiency has slowed recently, resulting in an approaching power wall. Specifically, the I/O power is trending to be larger than the total available socket power, leaving nothing for computing.


Traditional electrical interconnects are not up to the challenge largely due to limited bandwidth, electrical cross-talk, and low input/output pin density.


To solve this problem, researchers started exploring ways to use light to transmit information between and within microchips. Researchers’ goal was to eliminate electrical interconnects within a chip altogether, replacing them with optical waveguides that carry information encoded on photons.


Silicon photonics refers to the application of photonic systems using silicon as an optical medium. Silicon photonics uses photons to detect processes and transmit information more efficiently than electrical signals, and yet have low manufacturing costs as a result of using conventional silicon-integrated-circuit processes.


Silicon photonics (SiPh) is a material platform from which photonic integrated circuits (PICs) can be made. These Photonic Integrated Circuits (PICs)  enable, extend, and increase data transmission. PICs may consume less power and generate less heat than conventional electronic circuits, offering the promise of energy-efficient bandwidth scaling.


Silicon photonics combines technologies such as complementary metal oxide semiconductor (CMOS), micro-electro-mechanical systems (MEMS) and 3D Stacking.  It uses silicon on insulator (SOI) wafers as the semiconductor substrate material and most of the standard CMOS manufacturing processes can be applied.  The silicon material used in such photonic systems is designed with sub-micrometer precision and is deployed into the microphotonic components.


SiPh is compatible with CMOS (electronic) fabrication, which allows SiPh PICs to be manufactured using established foundry infrastructure.  The same processes have made transistors, microchips, and integrated circuits from silicon to be small, affordable, and energy-efficient devices that could execute complex tasks.


Silicon photonics has emerged as a disruptive technology to address data bottlenecks inside of systems and between computing components, enhancing power efficiency, improving response times and delivering faster insights from Big Data.


Silicon photonics allows the transfer of large volumes of data at very high speed between computer chips in servers, large data centers, and supercomputers, overcoming the limitations of congested data traffic and high-cost traditional interconnects.   In addition, features such as low environmental footprint, low heating of components, low operating cost, high optical functions integration, high density of interconnects, low error rate and spectral efficiency are adding value to the silicon photonics products.


Katharine Schmidtke, strategic sourcing manager, optical technology at Facebook says, “Silicon photonics can produce integrated designs, with all the required functions placed in one or two chips. Such designs will also be needed in volume, given that a large data centre uses hundred of thousands of optical transceivers, and that requires a high-yielding process. This is a manufacturing model the chip industry excels at, and one that silicon photonics, which uses a CMOS-compatible process, can exploit.”


Schmidtke is upbeat about silicon photonics’ prospects. “Why silicon photonics is attractive is integration; you are reducing the number of components and the bill of materials significantly, and that reduces cost,” she says. “Then there is all the alignment and assembly cost reductions; that is what makes this technology appealing.” The silicon photonics market is expected to reach $1.9 billion by 2022, at a CAGR of 22.1 percent between 2016 and 2022, according to a report by Reportlinker, a market research solution.


Silicon Photonics shall also benefit other application areas such as fibre-to-the-home (FTTH) or fibre-to-the-premises (FTTP), environmental monitoring, biological and chemical sensing, microwave photonic circuits, medical and military applications and astronomy. Eventually, the technology could reach home computers and mobile devices and improve applications from gaming to video streaming.


Photonic integrated circuits (PICs), the optical counterpart of traditional electronic integrated circuits, are paving the way toward truly portable and highly accurate biochemical sensors for Department of Defense (DoD)-relevant applications.


Military hopes integrated photonic circuits will eventually be the core of new communication devices that work faster, use less energy and can zip data across long distances, said Adele Ratcliff, director of the Defense Department’s manufacturing technology program. Mark Wright, spokesman for the Defense Department, noted how fiber optics are used to send digitized information between computers across the country, a technology that serves as the basis of the internet. Fiber optic cables help telecommunications companies sell high-speed internet to customers. The military could use that technology too, Mark Wright, said. “Despite the heavy interest of the commercial sector in this area, the DoD has planes, ships, and other platforms that need to move information around on them and adopting the solutions developed in this area are certainly of interest,” he said.

Recent advancements overcoming challenges

Recent advancements in Silicon Photonics enable tighter monolithic integration of many optical functions within a single device is leading to higher yield, high reliability, and low costs common in the electronics industry. Along with the waveguides came optical modulators, oscillators, and resonators that could precisely modify the properties of light. On-chip spectrometers have become available that can extract molecular and atomic signals from incident light on a device that is smaller than a fingernail.


Chips with onboard photonics together with novel packaging schemes to connect chips with optical fibers and standards to ensure interoperability is making accessible a large range of applications from fiber optics based communication systems to intra-chip communication or as optical interconnects.


These photonic devices have an enormous range of potential applications. The most consequential is the replacement of integrated circuits (ICs) that run on electrons with photonic integrated circuits (PICs) that run on photons. The best case would be a fully integrated silicon photonic circuit with no electronic components whatsoever. The next best option is a hybrid silicon photonic circuit that coexists with electronic transistors. To match the high production volume and low-cost efficiency of the current generation of microprocessors, these photonic circuits would have to be compatible with CMOS fabrication technology.



Silicon photonics involves patterning a silicon wafer with optical components using existing semiconductor manufacturing techniques. These miniaturized photonics systems show significant promise for next-generation optical communications technology. A key component of these devices is optical waveguides, which ideally guide light with minimal loss of energy.


Waveguides are components used to route light signals on a chip. The high refractive index of silicon enables the high confinement of light in the waveguides and leads to the reduction of the footprint of the silicon-based photonic devices. Silicon-on-insulator processes, which are now widely available for CMOS, can easily create a low-loss waveguide by patterning the same semiconductor layer used for transistors and then covering it with insulator.


However, the high thermo-optic (TO) coefficient of silicon which is around 1.86 × 10−4/K and the high confinement of light in the waveguide core leads to high temperature sensitivity of the silicon photonic devices. Researchers from University of California, Santa Barbara (UCSB) and the California Institute of Technology (CalTech) came up with new microchip-scale, integrated waveguides for photonic delay employing silicon processing. Such photonic delays are useful in military applications ranging from small navigation sensors to wideband phased array radar and communication antennas, according to DARPA


Rajesh Menon and other Engineers at the University of Utah (Salt Lake City, UT) have developed an ultracompact beamsplitter—measuring only 2.4 by 2.4 microns or one-fiftieth the width of a human hair –for dividing light waves into two separate channels of information. In 2015, researchers at the Harvard John A. Paulson School of Engineering and Applied Sciences (SEAS) developed the first on-chip metamaterial with a refractive index of zero, meaning that the phase of light could be stretched infinitely long. The metamaterial represented a new method to manipulate light and was an important step forward for integrated photonic circuits, which use light rather than electrons to perform a wide variety of functions.


Now, SEAS researchers have pushed that technology further – developing a zero-index waveguide compatible with current silicon photonic technologies. In doing so, the team observed a physical phenomenon that is usually unobservable — a standing wave of light. When the refractive index is reduced to zero the light no longer behaves as a moving wave, traveling through space in a series of crests and troughs, otherwise known as phases. Instead, the wave is stretched infinitely long, creating a constant phase.  The phase oscillates only as a variable of time, not space.


This is exciting for integrated photonics because most optical devices use interactions between two or more waves, which need to propagate in sync as they move through the circuit. If the wavelength is infinitely long, matching the phase of the wavelengths of light isn’t an issue, since the optical fields are the same everywhere.


“This adds an important tool to the silicon photonics toolbox,” said Camayd-Muñoz. “There’s exotic physics in the zero-index regime, and now we’re bringing that to integrated photonics. That’s an important step, because it means we can plug directly into conventional optical devices, and find real uses for zero-index phenomena. In the future, quantum computers may be based on networks of excited atoms that communicate via photons. The interaction range of the atoms is roughly equal to the wavelength of light. By making the wavelength large, we can enable long-range interactions to scale up quantum devices.”


Ideal light confinement requires cladding material with a very low refractive index. While silicon dioxide works well as a cladding material, a convenient method of thin-film deposition — called plasma-enhanced chemical vapor deposition (PECVD) — tends to produce suboptimal silicon dioxide films.


Scientists at the Naval Information Warfare Center Pacific (NIWC-Pacific) have recently invented a new type of optical waveguide deposited on a silicon wafer that results in greater confinement of light, as reported in Sep 2020. The waveguide has applications for low-loss, high-bandwidth data processing in modern computing systems.


To this end, Navy researchers have developed a way to manufacture an optical waveguide with silicon dioxide cladding using PECVD. First, a waveguide is patterned onto a silicon wafer using electron-beam resist lithography or photoresist etching. The waveguide and wafer are then covered in a thick silicon dioxide cladding through a particular PECVD process. The cladding has a greatly reduced refractive index and thus achieves better light confinement than traditional PECVD waveguides.


Integrated semiconductor optical amplifier

As we focus on reducing total power consumption, integrated semiconductor optical amplifiers are an indispensable technology, made possible with the same material used for the integrated laser.


Optical Interconnects

Today, optical interconnects over fiber dominate long-distance interconnects, while electrical dominate short distances. The current I/O performance of electrical interconnects are running into practical performance limits. Silicon photonics offers a potential breakthrough in optical interconnections that provide bandwidth and power performance, to remove communication bottlenecks within integrated circuit chips that incorporate over a billion transistors, and between integrated circuits in single boards that provide multi-teraflop (1012) computing capacity.


That trend is the optical interconnection of components, now moving from systems to boards to chip packages to chips themselves, says Lionel Kimerling, the Thomas Lord Professor in Materials Science and Engineering and director of the MIT Microphotonics Center.  “There are significant challenges for each one of those steps. Cost, bandwidth density, and power efficiency are the big three, and the cost is the one that’s really controlling the entry of photonics into the system.


For example, there are 100G optical transceivers being used for rack-to-rack connectivity — connecting row switches and director-class switches throughout data centers. The goal is to advance optical performance and replace electrical by bringing optical I/O directly into servers for board-to-board and package-to-package connectivity.


Optical I/O has the potential to be more efficient than electrical I/O while simultaneously providing >1000 times more reach. By tightly integrating optical into a server package using silicon photonics, we can gain three benefits in a smaller footprint: (1) lower power (2) higher bandwidth and (3) reduced pin count,  says Jeff Hockert is a senior marketing manager at Intel. We call this research vision Integrated Photonics and believe it will fundamentally change our data center network architectures. It will free data to move around the data center much more efficiently — at both lower power and lower latency. Our research goals are 1Tb/s per fiber at @ 1pJ/b with up to a 1km reach.


Researchers of the University of Twente (UT; Enschede, Netherlands) have, for the first time, succeeded in connecting two parts of an electronic chip using an on-chip optical link, all fabricable with standard CMOS technology — a long-sought-after goal, as intrachip connection via light is almost instantaneous and also provides electrical isolation. Such a connection can, for example, be a safe way of connecting high-power electronics and digital control circuitry on a single chip without a direct electrical link. Vishal Agarwal, a UT PhD student, created a very small optocoupler circuit that delivers a data rate of megabits per second in an energy-efficient way.


US-based startup Ayar Labs announced in Nov 2020 that it has successfully secured $35M Series B funding with a host of investors, including Downing Ventures and BlueSky Capital, to develop an optical interconnect (OIO) technology. In 2015 for Ayar Labs researcher Chen Sun co-authored a paper detailing how to bond an electronic microprocessor with an optical processor in a 3 mm x 6 mm chip. The foundry-fabricated chip includes 70 million transistors along with 850 photonic components. “This is a milestone. It’s the first processor that can use light to communicate with the external world,” said Vladimir Stojanović, an associate professor of electrical engineering and computer sciences at the University of California, Berkeley, who led the development of the chip. “No other processor has the photonic I/O in the chip.”


A technical brief authored by Ayar Labs in 2019 offers concrete details about the potential for the hybrid processor I/O:

  • Consumes less than 5 picojoules per bit of power (less than half of modern SerDes at 112 Gbps)
  • Improves latency by a factor of ten
  • Increases bandwidth up to 100 Tbps long-term, beyond what is possible for electrical interconnects

Ayar Labs' SoC representation demonstrating optical interconnections for a Terabit PHY monolithic-in-package optical (MIPO) I/O


Integrated laser Sources

The ideal scenario for silicon photonics is a light source made from silicon. Such an LED or laser can be easily integrated on a chip and be readily fabricated. Wide-spread adoption of silicon photonics has been hampered in part by the lack of monolithically integrated laser sources. However, the fundamental obstacle towards this photonic future is the indirect band gap of silicon. What this means is that an excited electron in silicon cannot return to the ground state without a push from somewhere – usually through lattice vibrations called phonons.


It’s not very hard to excite electrons in silicon—any blue laser will do the job. However, these electrons will wait for a phonon to come along, which can take somewhere around 10 to 30 milliseconds. By this time, the wandering holes and electrons in silicon would have combined with the defects in the crystal lattice to produce heat, not light. In fact, a million electrons need to be excited in silicon to generate a single photon. By comparison, GaAs only requires two excited electrons per photon. To summarize, silicon can emit light, but with very small yields.


A way out of this conundrum was suggested in 1990 by Leigh Canham, then affiliated with the Royal Signals and Radar Establishment. He showed that crystalline silicon with over 80 percent porosity showed photoluminescence in the red when exposed to diffuse blue light.This work led to many low-dimensional silicon systems, including nanocrystals, semiconductor-insulator superlattices, and nanopillars being actively investigated as a means of improving the light-emission properties of silicon.


In 2003, Jalali and his team at UCLA developed a Raman silicon laser. This is widely considered to be the first true silicon laser. Like porous and nanocrystalline devices, these lasers also required a light source for excitation. However, this light interacts with phonons in silicon to excite light emission through the Raman effect. The Raman lasers from Jalali’s group only produced intermittent pulses of light. The first continuous silicon Raman laser was created in 2005 by a team at Intel led by Haisheng Rong. These lasers were made from a narrow silicon waveguide cut into a dielectric medium.


Despite these advances, no silicon laser or LED has been able to match the optical gain of III-V lasers. This has led some to question whether a fully silicon light source is either practical or desirable.  Whilst there has been a range of microminiature lasers incorporated directly into silicon over the years, including melding germanium-tin lasers with a silicon substrate and using gallium-arsenide (GaAs) to grow laser nanowires, these methods have involved compromise.


The integration on silicon of efficient indium phosphide based light sources, currently driving long-range telecommunication networks, is known to be very challenging, owing to the large mismatch in crystal lattice constants between both materials. In recent years there has been some progress in this direction. John Bowers at the University of California, Santa Barbara, has shown that it is possible to bond GaAs and InP to a silicon wafer. There are still challenges to solve, such as growing a 300 mm III-V wafer that is compatible with CMOS. If a silicon laser does not materialize, this might be our best shot at integrated silicon photonics.


Integrated multi-wavelength lasers: Using a technique called wavelength division multiplexing, separate wavelengths can be used from the same laser to convey more data in the same beam of light. This enables additional data to be transmitted over a single fiber, increasing bandwidth density.


Optical Modulators in Silicon

Optical modulators are devices that detect electrical signals and modulate a light beam that propagates either in free space or in an optical waveguide. This device can alter different beam parameters and therefore optical modulators can be categorized as amplitude, phase, or polarisation modulators. Modulators have been improved dramatically in recent years. Most notably the bandwidth has increased from the MHz to the multi GHz regime in little more than half a decade.


Ideally, optical modulators are required that have high modulation speed, large bandwidth, and small footprint, as well as low loss and ultra-low power consumption. They also need to be CMOS compatible. For any interconnect solution, optical modulation is one of the main functionalities required in a photonic circuit.


Ring Modulators

Ring resonators have played a very prevalent role in the field of silicon modulators. In general, a ring resonator consists of a closed loop optical waveguide, which creates a resonant condition for wavelengths which are a whole number of the optical path length. A ring resonator is only useful when light can be coupled in and out of the resonant cavity. The most common method of coupling is codirectional evanescent coupling between the ring and an adjacent waveguide

A resonator is a passive device but a ring can also be used as an active device such as a modulator. In a ring modulator the resonator is made to align the operating wavelength with the resonance peak. In this way, modulating the optical path length of the ring shifts the resonance peak.  High speed modulation has been demonstrated many times over with ring resonators via shifting the carrier density of the material, changing the refractive index, and therefore the resonance wavelength through the application of an external voltage to the device’s PN junction.


Conventional silicon modulators take up too much area and are costly to place on IC packages. By developing micro-ring modulators, Intel has miniaturized the modulator by a factor of more than 1000, thereby eliminating a key barrier to integrating silicon photonics onto a computing package. Existing silicon modulators utilize half p-type and half n-type doped silicon, meaning they have had impurities added to them through a standard process used in transistor manufacturing. The MIT researchers’ device is similar, except that the center of the modulator — including the waveguide that runs along its top — is undoped. When a voltage is applied, the free carriers don’t collect in the center of the device; instead, they build up at the boundary between the n-type silicon and the undoped silicon.


“Silicon has had a huge renaissance within the optical communication space for a variety of applications,” says Jason Orcutt, a researcher in the Physical Sciences Department at IBM’s Thomas J. Watson Research Center. “However, there are still remaining application spaces — from microwave photonics to quantum optics — where the lack of second-order nonlinear effects in silicon has prevented progress. This is an important step towards addressing a wider range of applications within the mature silicon-photonics platforms around the world.”


Light Detectors

The innovations in CMOS processing by which alloys of silicon and germanium can be integrated without any harm to transistors have led to the integration of detectors on silicon.  These alloys can absorb the light signal that passes freely through silicon, liberating electrons that can be detected electrically.


Ge photodetectors offer an enormous responsivity to laser wavelengths near 1.55μm at high frequencies to 40GHz, and they can be easily fabricated as part of a planar silicon processing schedule. At the same time, germanium has enormous potential for enabling 1.55 micron lasers on silicon and for enhancing the performance of silicon modulators.


All-silicon photodetector: For decades, the industry has believed silicon has virtually no light detection capability. Intel showcased research that proves otherwise. Lower cost is one of the main benefits of this breakthrough.


Intel micro rings serve as a light modulator, and as a silicon photodetector. Until now, the industry believed that silicon had virtually no light detection capability in the 1.3–1.6um wavelength range, seen as a fundamental limitation for silicon photonics technology. That was proved wrong. On the receiver side, we have been working on all silicon ring-based photodetectors which combine wavelength selection and photodetection functions in a single device. Earlier this year we published light detection capability with an all-silicon photodetector, and demonstrated this photodetector operating at a data rate of 112Gb/s, says Jeff Hockert. These ring photodetectors can be assembled with CMOS trans-impedance amplifiers to build low-cost optical receivers. A major advantage of this approach is processing and material cost reduction.


Japanese Scientists develop world’s first O-E-O ‘transistor.’

A team of scientists sponsored by the Nippon Telegraph and Telephone Corporation, a Japanese telecom, have made a titanic breakthrough with photonic technology. They were able to build an Electro-Optic Modulator (E-O) that runs at 40 Gbps that uses just 42 attojoules per bit, meaning it consumes over an order of magnitude less power than the best of the previous experiments. It outperforms them, too, with about half the capacitance at less than a femtofarad.


They then constructed a photoreceiver (O-E) based on the same technologies, and that was able to run at 10 Gbps using two orders of magnitude less power than other optical systems at just 1.6 femtojoules per bit. It’s also the first to not require an amplifier (which saves power) and have a low capacitance at just a few femtofarads.


Combining the two, they demonstrated the world’s first O-E-O ‘transistor.’ It can function as an all-optical switch, a wavelength converter, and a repeater. The incredible versatility makes it the first device that provides benefits over electronic hardware at chip-scale. The researchers suggest it could be used for inter-core communication and to sustain cache coherency.


The scientists were able to make this breakthrough by developing a new type of photonic crystal (a term meaning a synthetic insulating material that controls light), and it’s a piece of silicon with a bunch of holes drilled in it. The holes are arranged such that if the light goes through them it interferes with itself causing it to cancel out. If a line of holes is blocked, then the light goes follows the path and is funneled into light-absorbing material that converts it into a current. The same system also works in reverse.


It’s hard to understate just how exciting this breakthrough is. Up until now, the only role photonics has played in the data center is long-range communication, targeting distances from 500m to 10km. Recent announcements like Intel’s 400G have shrunk that distance to room-scale, with board-scale known to be in the works. But bringing photonics down to chip-scale makes the technology consumer accessible and has the potential to rewrite the rulebook when it comes to performance. After all, light is faster than electronics.


The main challenges scientists have faced are the power requirements, which can exceed a thousand times the requirements of electronic processing, and speed, because each time the light is absorbed it must go into a capacitor. That capacitor must fill up and discharge fully to pass the signal on, but up until now, it’s been very challenging to build a capacitor small enough for that to happen quickly. The research team made leaps and bounds and finally matched silicon hardware in terms of performance and power requirements.

Silicon Germanium

Germanium on silicon processes are being developed has many advantages to enable higher performance designs that can be better incorporated into integrated photonic circuits (IPC). However, the growth of germanium for photodetection and III–V materials for light generation is technologically challenging on a silicon substrate due to mismatched lattice constants and thermal expansion coefficients.


Sicoya, a spinout of Technischen Universität Berlin, entered the market in 2017 with a 100G silicon photonics transceiver that uses a silicon-germanium (SiGe) BiCMOS process for the optical components. Most silicon photonics companies build their components on a silicon-on-insulator (SOI) process, but Sicoya CEO Sven Otte called it “a natural fit to start with the BiCMOS electronics and add the photonics.”


“It makes sense because the electronics part of most optical interconnects uses BiCMOS anyway,” Otte said. “The transimpedance amplifiers, the clock and data recovery, and the laser drivers [are in SiGe BiCMOS] … CMOS is used only where you need signal processing.”


Today, said Otte, “Sicoya’s electronics and photonics are on one chip, and our transistors are faster than CMOS transistors. SiGe is a super-efficient technology that not only gives us advantages from the performance perspective [but lets us] eliminate some of the assembly processes.” It’s very difficult, if not impossible, to build optical components such as laser diodes and photodetectors (PIN diodes) effectively in silicon, so most silicon photonic solutions put those on a separate die.



The ultimate physical manifestation of the silicon photonic device would be as part of an optoelectronic integrated circuit (OEIC) formed monolithically in silicon, combining the photonic functionality and the electronic intelligence in seamless integration. This requires realization of integrated photonic circuits that combine various silicon photonic devices with excellent properties, such as laser sources, modulators, low-loss waveguides, wavelength filters, optical receivers and photonic switches.


One solution to this problem is to grow materials on a compatible substrate and then transfer them onto silicon.  Two-dimensional (2D) materials are a class of crystals that naturally lend themselves to this type of transfer process. Because these materials are covalently bonded in-plane and held together out-of-plane by van der Waals forces, individual atomic planes can be mechanically separated from the bulk crystal and placed onto arbitrary substrates.  These materials have many unique electrical and optical properties and can be transferred to an arbitrary substrate without lattice matching requirements. Owing to strong quantum confinement out-of-plane, 2D materials have many unique properties that are uncommon in their 3D counterparts, which make them particularly attractive for optoelectronic applications. Some of the promising two-dimensional materials, graphene, black phosphorus and  transition metal dichalcogenides.


Intel Labs Day 2020 conference in Dec 2020, Intel highlighted key advances in the fundamental technology building blocks that are a linchpin to the firm’s integrated photonics research. These building blocks include light generation, amplification, detection, modulation, complementary metal-oxide-semiconductor (CMOS), all of which are essential to achieve integrated photonics.


Among the first noteworthy updates, Intel showed off a prototype that featured tight coupling of photonics and CMOS technologies. This served as a proof-of-concept of future full integration of optical photonics with core compute silicon. Intel also highlighted micro-ring modulators that are 1000x smaller than contemporary components found in electronic devices today. This is particularly significant as the size and cost of conventional silicon modulators have been a substantial barrier to bringing optical technology onto server packages, which require the integration of hundreds of these devices.


Ayar labs developing one terabit per second electro-optical I/O chip

Ayar Labs, a silicon photonics startup based in Emeryville, California, is getting set to tape out its electro-optical I/O chip, which will become the basis of its first commercial product. Known as TeraPHY, it’s designed to enable chip-to-chip communication at lightning speed. The company is promising bandwidth in excess of one terabit per second, while drawing just a tenth the power of conventional electrically-driven copper pins.


According to the company’s website, the initial TeraPHY device will be available as a 1.6 Tb/sec optical transceiver, comprised of four 400 Gb/sec transceivers per module. All the componentry except for the light source (which is supplied by a separate 256-channel laser module, called SuperNova) has been integrated into the device. That includes the electrical interfaces, the optical modulators, the photodetectors, and the dense wavelength division multiplexing (DWDM) wavelength multiplexer/demultiplexer, as well as all the driver and control circuitry.


Wright-Gladstein says the solution not only delivered a high performance electro-optical device, but was able to do so in an area 1/100 the size of a typical long-haul optical transceiver. “Because of that 100X size difference, you’re now crossing that threshold set by electrical SerDes, and you’re making an optical I/O that’s smaller than your electrical I/O,” she says.


Forgoing the more exotic designs of other silicon photonics solutions required some extra tinkering. The design uses optical “micro-ring resonators” implemented in CMOS to achieve the extreme density of the TeraPHY. These resonators can be “finicky” due to thermal issues and size, but according to Wright-Gladstein, they’ve implemented a patented thermal tuning technology that stabilizes the resonators and makes them very reliable.


Physically, TeraPHY is in the form of an “chiplet,” a chunk of silicon that is meant to be integrated into the kind of multi-chip modules that are becoming more commonplace in high-end processor packages. The TeraPHY tape out is slated for the end of the current quarter (Q1 2019), with the first integrated products from Ayar’s silicon technology partners due to hit the street in 2020.


Research project integrates laser directly on silicon with a modulator

IRT Nanoelec, an R&D consortium have announced the first co-integration of a III-V/silicon laser and silicon Mach Zehnder modulator demonstrating 25 Gbps transmission on a single channel. This transmission rate usually is achieved using an external source, over a 10 km single-mode fiber.


”To achieve these recent results, silicon photonics circuits integrating the modulator were processed first on a 200mm SOI wafer, although 300mm wafers also could be used in the near future. Then, a two-inch wafer of III-V material was directly bonded on the wafer. In the third step, the hybrid wafer was processed using conventional semiconductor and/or MEMS process steps to produce an integrated modulator-and-laser transmitter.


“Jointly obtained by STMicroelectronics and Leti in the frame of the IRT Nanoelec cooperation, these results, especially fabricating the laser directly on silicon, demonstrate IRT Nanoelec’s worldwide leadership in III/V-on-silicon integration to achieve high-data-rate fiber-optic modules,” said Stéphane Bernabé, project manager.


Integrating photonics with silicon nanoelectronics for the next generation of systems on a chip

Researchers from MIT, the University of California at Berkeley, and Boston University announced earlier that  they’d managed to build a microprocessor that combined electronic and optical components using existing manufacturing processes. They built their device on “silicon-on-insulator” wafers, which feature a layer of silicon dioxide glass beneath the top silicon layer. This is the material typically used for silicon photonics and for some high performance electronic chips, but it is much more expensive than the bulk silicon used for most microchips.


In the latest issue of Nature, a team of 18 researchers, led by the same MIT, Berkeley, and BU groups, reports another breakthrough: a technique for assembling on-chip optics and electronic separately, which enables the use of more modern transistor technologies. Again, the technique requires only existing manufacturing processes.



Silicon — which is the basis of most modern computer chips — must be fabricated on top of a layer of glass to yield useful optical components. The difference between the refractive indices of the silicon and the glass — the degrees to which the materials bend light — is what confines light to the silicon optical components.


The earlier work on integrated photonics, which was also led by Ram, Stojanovic, and Popovic, involved a process called wafer bonding, in which a single, large crystal of silicon is fused to a layer of glass deposited atop a separate chip. The new work, in enabling the direct deposition of silicon — with varying thickness — on top of glass, must make do with so-called polysilicon, which consists of many small crystals of silicon. “Here we introduce photonics into bulk silicon complementary metal–oxide–semiconductor (CMOS) chips using a layer of polycrystalline silicon deposited on silicon oxide (glass) islands fabricated alongside transistors.”


In addition to millions of transistors for executing computations, the researchers’ new chip includes all the components necessary for optical communication: modulators; waveguides, which steer light across the chip; resonators, which separate out different wavelengths of light, each of which can carry different data; and photodetectors, which translate incoming light signals back into electrical signals.


We integrated this photonic platform with a 65-nanometre-transistor bulk CMOS process technology inside a 300-millimetre-diameter-wafer microelectronics foundry. We then implemented integrated high-speed optical transceivers in this platform that operate at ten gigabits per second, composed of millions of transistors, and arrayed on a single optical bus for wavelength division multiplexing, to address the demand for high-bandwidth optical interconnects in data centres and high-performance computing”


By decoupling the formation of photonic devices from that of transistors, this integration approach can achieve many of the goals of multi-chip solutions, but with the performance, complexity and scalability of ‘systems on a chip’. As transistors smaller than ten nanometres across become commercially available, and as new nanotechnologies emerge, this approach could provide a way to integrate photonics with state-of-the-art nanoelectronics.

Silicon photonics to meet communication needs

A recently launched project, funded by the European Union’s Horizon 2020 programme, hopes to develop silicon photonics based transceivers that will meet these requirements. Called COSMICC, the project intends to combine CMOS electronics and silicon photonics with innovative high-throughput fibre attachment techniques.


“While there are already some commercial products that support 100Gbit/s communications, we need to prepare for the next generation – for example, devices that can support 400Gbit/s and 1Tbit/s, as well as developing technology that can aggregate data rates beyond 1Tbit/s.” In particular, the COSMICC project is looking to develop technology with a cost/bit that cannot be achieved using the current wavelength division multiplexing approach, said Dr Ségolène Olivier acting as project leader.


In its roadmap, the COSMICC project is looking to meet a cost target of €0.15/Gbit/s, while consuming just 2pJ/bit. By contrast, today’s technology costs something like €20/Gbit/s, while consuming 35pJ/bit. By combining CMOS electronics and silicon photonics with high throughput fibre attachment techniques, the COSMICC project believes it will be able to develop solutions that can scale to meet the future requirements of data centres and supercomputers.




References  and Resources also include:



Cite This Article

International Defense Security & Technology (February 8, 2023) Silicon Photonics breakthroughs enable large data centers, Real time cloud-computing, Exascale supercomputers and cognitive computing. Retrieved from
"Silicon Photonics breakthroughs enable large data centers, Real time cloud-computing, Exascale supercomputers and cognitive computing." International Defense Security & Technology - February 8, 2023,
International Defense Security & Technology November 9, 2021 Silicon Photonics breakthroughs enable large data centers, Real time cloud-computing, Exascale supercomputers and cognitive computing., viewed February 8, 2023,<>
International Defense Security & Technology - Silicon Photonics breakthroughs enable large data centers, Real time cloud-computing, Exascale supercomputers and cognitive computing. [Internet]. [Accessed February 8, 2023]. Available from:
"Silicon Photonics breakthroughs enable large data centers, Real time cloud-computing, Exascale supercomputers and cognitive computing." International Defense Security & Technology - Accessed February 8, 2023.
"Silicon Photonics breakthroughs enable large data centers, Real time cloud-computing, Exascale supercomputers and cognitive computing." International Defense Security & Technology [Online]. Available: [Accessed: February 8, 2023]

About Rajesh Uppal

Check Also

Optical and Laser Illusions

Optical illusions (or visual illusions) are tricks that your mind plays on you based on what …

error: Content is protected !!