Rockley Photonics and the University of Southampton: A Prosperity Partnership

Lead Research Organisation: University of Southampton
Department Name: Optoelectronics Research Centre (ORC)

Abstract

According to the Cisco's 2016 Global Cloud Index, hyperscale data centers will grow from 259 in number at the end of 2015 to 485 by 2020 and will represent 47 percent of all installed data center servers by 2020. This extraordinary growth prediction will exasperate an industry already struggling to cope with the costs and power requirements of existing IT infrastructure technology. Rockley Photonics is developing a highly scalable optical fibre networking and packet switch solution for cloud datacenters. The heart of the solution will be the company's TopangaTM optical fiber interfaced packet switch application-specific integrated circuit (ASIC). With the TopangaTM ASIC, one can network up to hundreds of thousands of servers together in cloud datacenters at a fraction of the current capital expenditure (CAPEX) and operating expenditure (OPEX - mainly power), achieving greatly superior server utilization and lower communication latency required for emerging virtual reality and machine learning applications. Rockley Photonics TopangaTM for the first time in the industry uses silicon photonics chips incorporating wavelength division multiplexing (WDM) integrated with the switch ASIC to directly bring high bandwidths of data on and off the ASIC. These chips use micro-scale silicon photonic opto-electronic elements (e.g. modulators) to perform the electrical-to-opical conversions with new levels of compactness and high densities pitch-matched to the ASICs to reduce electrical connection length almost to zero.

This proposed Prosperity Partnership brings together the UK's premier industry and academic partners in the field of Silicon Photonics, Rockley Photonics and The Silicon Photonics Group at the Optoelectronics Research Centre (ORC), University of Southampton. The management team at Rockley have already been involved in 2 Silicon Photonics start-up companies and now Rockley Photonics is pioneering the 3rd generation of this emerging technology. The ORC team have demonstrated numerous world firsts in the field, and are known around the globe as a pioneering team in the field of Silicon Photonics. Together these teams will form a formidable Prosperity Partnership that will work together to transform the way in which data centre architectures handle vast quantities of data by developing novel photonic solutions to the modulation and distribution of optical signals, and the overall switching architectures. We anticipate significant impact from the funding, should we be successful.

The Silicon Photonics Group at the University of Southampton are well known internationally for pioneering work in the field since 1989. In 2012 the group moved to Southampton University where the head count has since grown by a factor of x3 to more than 40 researchers in total. Not only do the group have a large number of key collaborators within the Southampton environment, but also play a key role in running and using the clean room complex, putting them in a unique position worldwide in having design, fabrication, prototyping and testing facilities/expertise. The investment of more than £120 Million in the cleanroom complex has recently been enhanced by a £3million EPSRC investment in a photolithography scanning capability which enables fast prototyping, ideal for facilitating disruptive optical device and optical circuit research.

Together the combined teams will develop the devices and photonic circuits necessary for future generations of Rockley Photonics products.

Planned Impact

The Prosperity Partnership is an ideal scheme to encourage impact. The most obvious pathway to impact for the devices and photonic circuits developed within the programme, together with the associated new patent filings within the Prosperity Partnership will be via Rockley Photonics exploiting the IP directly, but if this is not appropriate we will seek other exploitation with other companies via the University Research and Innovation (RIS) Office. RIS at Southampton will set up and coordinate contract agreements in order to maximize the exploitation of the generated IP. Rockley Photonics will receive priority in licensing opportunities, and then our wider portfolio of contacts will be approached for potential exploitation. Establishing a spin-out company based on existing and future IP also remains an optional pathway to impact. The investigators have established networks in academia, industry and government agencies. We will take advantage of such networks and also of our membership on various conference committees and our invited talks to create new strategic links and to increase prospects for leveraging our findings in new funding opportunities and commercial applications.

Societal impact will be via the manner in which the Rockley Photonics product will transform the architecture and power consumption of datacenters. This will in turn support the continued growth of data for business and for individuals, transforming the way in which the daily lives of advanced nations advances. Of course the pathway to the future generations of Rockley Photonics products is via the research that will be conducted within this Prosperity Partnership.

The students and researchers trained by the project and the knowledge gained through the collaborations are valuable future resources for the UK. Our programme will provide further training and stimulation to retain staff in photonics and electronics. The University of Southampton offers a range of personal development courses, including enhancing leadership qualities and effectiveness, the research team will be able to develop their technical and leadership skills. We will also utilise short visits to Rockley Photonics to give them the opportunity to learn about new equipment, processes and techniques, and to expand their own professional networks for future collaborations and to bring new knowledge back to the team.

To ensure that our research is disseminated to a broad academic audience, we will target high impact (e.g. Nature group), open access (e.g. Optics Express, IEEE Photonics) and non-photonics specific (e.g. IEEE Solid State Circuits) journals. We will also present our results at high profile international conferences, preferentially those connected to major photonics trade shows such as the CLEO conferences (USA and Europe), Photonics West (including sub-symposia on optoelectronics and green photonics), and Photon in the UK. The investigators are often symposium chairs and/or invited speakers at these major conferences. We will organise a special workshop at one of major conferences on silicon photonics for data centres.

Dissemination to the wider public will be carried out by publishing articles in scientific magazines (e.g. New Scientist and Laser Focus World) which are targeted at non-specialist readers, online platforms such as "the Conversation", and, when key breakthroughs are made, by liaising with journalists. Outreach activities will include participation in the Southampton's National Science and Engineering Week exhibition and poster presentations at the House of Commons SET for Britain. Opportunities to exhibit our research at the high profile annual Royal Society Summer Science Exhibition will be considered as results emerge. Team members will be encouraged to highlight recent results at UK and EU meetings and workshops (e.g. EPIC); we will also propose a silicon photonics for data centres topic to the Royal Society discussion meetings panel.
 
Description 1st Quarter: The programme has focussed on two key elements of work to date. The first maps to work package 1 of the work and surrounds a new design of optical modulator, the so-called MOSCAP modulator. MOSCAP modulators themselves are not new, but we are developing a process to facilitate the fabrication of a new design of such a device. We are developing fabrication processes in order to allow the device to have better quality silicon on both sides of an oxide junction rather than just on one side (as has been reported in the literature to date). This is a non trivial task, and whilst we are making progress, 3 months is innsufficient to report significant progress due to the detailed and complex nature of the task. We are also working with Caltech (see collaborators) to design and fabricate modulator drivers compatible with the modulators that will be developed at Southampton. This requires us to carry out detailed modelling based upon preliminary experimental data, in order that the drivers can interface effectively with the modulators, and drive the device efficiently at high speed, and can be physically integrated effectively. The second area of work maps to work package 2 and surrounds the development and integration of mid index material such as silicon nitride for low temperature sensitivity multiplexing and efficient optical electro absorption modulators (EAM) based on the so-called QCSE effect (quantum confined stark effect) and Franz Keldysh effect. We are currently developing the waveguide structure design and integration processes for the Mid Index based multiplexors and EAM Optical modulators. The development and integration of two material systems is complex and non-trivial and whilst we are making progress, similarly to WP1, 3 months is insufficient to report significant progress due to the complexity of the material integration and processing.
2nd Quarter. The work in WP1 and WP2 continues, and we have initiated work on the revised WP3. A detailed model of defect mediated detectors has been produced and calibrated against the literature. One conference presentation on MOSCAP results also made.

3rd Quarter. WP1 and WP2 works continues well, although we have had to slightly reorganise the sequencing of the work as some work elements were more complex than anticipated. 2 additional publications this quarter with Lorenzo Mastronardi winning the best studnent paper at a major international conference, CLEO Pacific Rim.
WP3 work was rearrange in the last quarter because recruitment has proved difficult. We wll update the Gantt chart within the next quarter

4th quarter: Recruitment continued to be a problem, but we reached a position where we had finally recruited a full complement of researchers. However, we then had a resignation from one RA role within the project. However, ths was actually a success story as the individual moved to a more senior position within the Southampton organisation.

The second patent was filed in Q5, and together with the first, is potentially of interest to Rockley Photonics for licencing purposes.

In the 6th and 7th quarters, research progressed well, albeit with some delays. Several milestones were achieved as reported elsewhere
Overall year two was successful, and has resulted in additional patents being filed. We have now filed 3 patents and a further 2 are in preparation
Exploitation Route 1st Quarter: Rockley Photonics are planning to incororate the findings of the work intop future iterations of their products. It is too early to determine axactly how this will be carried forward, but it will inevitably require some transfer of IP which may be in the form of know how, patents or other forms of IP.
2nd quarter: Comments from q1 remain relevant
3rd Quarter: Ist patent filing underway, so licencing will be a possibility once the patent is granted.
4th Quarter: 2nd patent application is being drafted.
5th Quarter: 2nd patent application filed
6th & 7th Quarter: Further discussions on additional potential patents
8th quarter: One more patent filed and 2 in preparation
Sectors Digital/Communication/Information Technologies (including Software),Electronics,Energy

URL http://rockleyphotonics.siliconphotonics.co.uk
 
Description A further 5 potential patents are under discussion, and may be filed in the future. Patent applications mentioned previously (below) are now filed. They are not listed as outputs yet, as they are not published yet. Two patent applications from the Prosperity Partnership are being drafted, but are not yet filed. In the previous reporting session, only 1 patent application was in preparation.
First Year Of Impact 2018
Sector Digital/Communication/Information Technologies (including Software)
Impact Types Economic

 
Description N/a
Amount £95,221 (GBP)
Organisation Rockley Photonics 
Sector Private
Country United States
Start 12/2017 
End 02/2019
 
Description Prosperity Partnership Partner 
Organisation Rockley Photonics
Country United States 
Sector Private 
PI Contribution Our team provide technical expertise in terms of device design, modelling, fabrication and testing
Collaborator Contribution Rockley Photonics bring technical expertise, device specifications, and managment expertise. This relationship is the basis of the Prosperity Partnership. Spend of the Rockley Photonics contribution this quarter (Q3 - 01/11/17 - 31/12/17) is £26,634.58. Predicted spend for the next quarter (Q4 - 01/01/2018-31/03/2018) is £234,241.52. Finally, the forecast spend in the financial year 01/04/2017-31/03/2018 is £207,606.94.
Impact None yet.
Start Year 2017
 
Description Technical collaboration on optical modulators 
Organisation California Institute of Technology
Country United States 
Sector Academic/University 
PI Contribution We have expertise in modulator design, so we provide details of the modulator characteristics.
Collaborator Contribution Caltech have expertise in modulator driver design, so together we can optimise the electronic-photonic interaction
Impact None yet. Very early days.
Start Year 2017
 
Description University Contribution 
Organisation University of Southampton
Department Optoelectronics Research Centre
Country United Kingdom 
Sector Academic/University 
PI Contribution Expertise in device design, fabrication and testing
Collaborator Contribution This is to capture the University contribution over and above the 20% of FEC. The University is contributing 3.25 man years of salary of a level 4 PDRA plus on-costs in cash. In addition it is contributing the overheads associated with this RA plus the additional overheads associated with a further 0.75 FTE of this PDRA. It is further contributing 445 cleanroom days associated with students who will work on the project, 5% of the time of Professor Mashanovich and 10% of the time of Dr Gardes.
Impact Too early
Start Year 2017