CEO interview: Rockley’s Rickman sees silicon photonics coming back to sensing
We started by covering some background.
Bookham enjoyed a 20-year life under its own name and progressed by a series of acquisitions before eventually being US domiciled and then merged with Avanex Corp. in 2009. Under the name Oclaro, it has been one of the largest suppliers of optical components, modules and subsystems for telecommunications. Oclaro was in turn acquired by Lumentum in December 2018.
At the start of his career Rickman was looking to realize commercially what he had studied for his PhD under Professor Graham Reed, then at the University of Surrey. This was the ability to create wave guides and electronically-controlled switches in silicon.
Back then it was clear to Rickman that the nascent optoelectronics sector could make use of the economies of scale enabled by silicon microelectronics manufacturing and that it would be possible to create some reasonably complex optical circuits in silicon.
The waveguides tend to be made as ridges in the silicon so that the physical boundary of the silicon defines the channel. And by manufacturing with silicon-on-insulator wafers it is possible to define the fourth wall of the waveguide as the buried silicon-oxide layer.
“Optical fibre communications and some sensing applications were the initial possibilities,” Rickman recalls.” Approaches to sensing were usually done by allowing some parameter to deflect or affect the light path and measure the comparative time delay versus a reference path. This made for the possibility of extremely sensitive measurements in a small form factor in a wide number of areas.
Rickman remembers that sensor applications were the initial thrust of the company but it was communications where the first traction was obtained with customers. The 1990s was a time when optical fibre was being deployed in urban communications and the access market and with the adoption of the Internet came the dot-com boom, he said. This resulted in sensing being put on the back burner.
Next: Intel and Cisco
Both Intel and Cisco invested in Bookham and the company built two wafer fabs for silicon photonics with automated back-end assembly.
Although Bookham grew by supplying such things as wavelength division multiplexing (WDM) components it was also obvious to Rickman that silicon doesn’t easily make light and is missing several other electro-optic properties. “We can make wave guides but light generation and detection are difficult. Nonetheless silicon photonics remains compelling.”
“To begin it was not communications but sensors I was interested in. But in the late 1990s the concentration was on optical fiber communications and the dot-com boom. The sensors were ahead of their time.” By this time Bookham WDMs and tunable filters were designed in with the big networking and telecommunications companies of their day, Nortel and Lucent, Rickman recalls.
But after the boom came the bust and this had a negative effect on Bookham, which had grown by acquisition into a much broader optical components company that was no longer only focused on silicon photonics. Indeed, Bookham ended up licensing some of its optical intellectual property to a California company called Kotura to help sort out IP issues after Kotura started supplying variable optical attenuators (VOAs) similar to Bookham’s.
Kotura had been founded in 2004 and Rickman ended up joining that company as chairman. In a sense he transferred along with the IP he had helped to create. Kotura is now part of Mellanox Technologies Ltd., which in turn is due to be acquired by Nvidia Inc.
“First generation silicon photonics components were aimed at telecom. Second generation products were able to make more use of germanium, in silicon-germanium, and monolithic detectors. It accompanied the increased use of optical transceivers in data centers. Single-mode optical fibre also moved into data centers. There were far more optical ports in data centers than in telecommunications.”
Rickman pointed out that meanwhile the silicon photonics sector was benefitting from a virtuous cycle with increasing application volumes and the arrival of the option of foundry production. Economies of scale in manufacture in turn reduces component costs and opens up yet more higher volume applications that are cost-sensitive.
Next: silicon photonics platform
After founding Bookham and chairing Kotura, Rickman founded Rockley Photonics in 2013 to address the expanding markets in the data center and sensing with the new economic possibilities of fabless manufacturing. “We went to the development foundries and production foundries who were ready to work with us,” Rickman said.
The company has now started shipping first examples of its first product, a Transmit-Receive Optical Sub-Assembly (TROSA) for data communication that operates at 100Gbps. Rockley has spent the time up until this product launch creating a silicon photonics platform but that does not mean that every photonic operation can and should be made in silicon, Rickman said
The platform should comprise waveguides on silicon, but also III-V lasers, optical fiber self-aligning insertion structures, microelectronics and packaging. Rickman argues that the silicon photonics platform should be essentially planar, easily automated for low cost but also use the best in class manufacturing processes for different functions.
Rockley hasn’t yet revealed who is manufacturing the various parts of its photonics platform or performing the assembly operations but Rickman points out the second-tier foundries are more amenable to development collaboration.
Rickman goes further to indicate that many of the larger foundries are approaching silicon photonics with a different philosophy to Rockley’s and one that ultimately he believes will not succeed.
“The first approach is to take a CMOS process and try and adapt it for photonics. The second approach is to develop a process optimized for photonics from the ground up and which therefore has little resemblance to established processes.”
Rickman said he understood why major foundries were trying to do the former. The less they have to change their established processes the more they can leverage the money they have invested in R&D and specific manufacturing equipment. “TSMC, Globalfoundries, STMicroelectronics are in the first camp, building waveguides on adapted CMOS. In principle, it is a good vision. You can push for integration of photonics with the microelectronics.”
However, Rickman said that ultimately vision fails because of the physics of photon behaviour in silicon as opposed to that of electrons in silicon. Ultimately it comes down to the fact that the wavelength of the photon is of the order of micron while the wavelength of the electron is of the order of a nanometer.
Rickman said there are specific disadvantages to using the nanometer-scale of modern CMOS for optical integration, including increased waveguide loss, wavelength sensitivity, fiber coupling loss, polarization sensitivity. Similarly, while it may be possible to build modulation circuitry in silicon, indium-phosphide is superior and being open to III-V materials provides broader options and capabilities. Finally, Rickman pointed out that use of trailing-edge silicon manufacturing capability at the micron dimension brings reduced manufacturing cost although with some additional manufacturing steps. He added: “There will always be some packaging.”
Rickman added that it is already accepted in the microelectronics field that for many applications multi-die components that use different processes can be the optimum production method. MEMS provide a good example. And with those lessons learned in MEMS and other applications, the same packaging infrastructure can be available for the production of silicon photonics ICs.
Rockley’s first application is the 400G capable data center. In 2018 Rockley demonstrated its platform could produce the world’s first single ASIC Level 3 datacenter routing switch with integrated 100G network ports using single-mode optical fiber.
And the first customer is Hengtong Optic-Electric Suzhou, an optical fiber and cable provider, based in Suzhou, Jiangsu, China. Rockley formed a joint venture with Hengtong to make 400G DR4 transceivers using Rockley’s LightDriver 1310nm wavelength optical engines. Hengtong also participated in a $30 million investment round in Rockley. The technology is good for the new breed of datacenters that supporting cloud computing, artificial intelligence and machine learning. “It’s a great route into the Chinese market for us,” said Rickman.
Next: Sensing, China
“The next applications are consumer sensing where the optical platform can be used to create complex sensing instruments in a single chip.”
One of the opportunities is optical coherence tomography (OCT), said Rickman. This can be used to capture micrometer resolution for 2- and 3D medical and industrial imaging. At the larger scale the opportunity is lidar for autonomous vehicles.
While most automotive lidar products are amplitude-modulated systems based on pulsed laser diodes, frequency-modulated continuous wave (FMCW) lidar is an emerging technology trend. FMCW also has a reduced risk of interference from other light sources, such as the lidars of passing vehicles.
Rickman expressed the view that frequency modulated continuous wave (FMCW) would win out over the time-of-flight approach. “It uses less power and you get velocity measurement as well.” And Rickman made the point that is not just automobiles that will have some version of this technology. “These are big markets. All robotic equipment will have this. But it is not an immediate market.”
Again, rather as it was at Bookham, the communications market is where there is traction initially. “In the data center market we will be shipping volume by end of this year. For 3D imaging there is a lot of interest but the market still has a long way to go. It’s not ready for production. We have built systems for customers today. FMCW using longer wavelength will be the universal method.”
But what comes next for optical circuits?
“We see optical computing and neural networks as definite markets,” said Rickman. “Using silicon photonics; an optical neural network processor can be done. We have optical labs in Pasadena and we are reaching out to places like Tyndall Institute, VTT, Southampton University. We have sown the seeds for both ONN and optical quantum computing.”
With Rockley receiving investment from Hengtong and in a joint-venture for its first major product we asked if the international trade tensions that exist today were a cause of concern for Rickman.
“You have to be very aware. But on the other hand, China is just another part of the world with its own characteristics. You would be missing out if you ignored it. Huawei ended up being the biggest customer for both Bookham and Kotura,” said Rickman. “Since the 1990s we have embedded ourselves in China with a close relationship with the Chinese Academy of Science Institute in Shanghai. With 20 years of experience in China we feel we’ve got the right approach.”
Related links and articles: