Public Funding for Vector Photonics Limited

Artificial Intelligence (AI) is set to revolutionise modern life, by enabling autonomous vehicles, real time medical diagnosis, financial modelling, power grid management, etc. However, the growth in the compute power necessary to develop future AI systems will run up against a data transmission bottle neck in compute power. This bottleneck is caused by delays in electrical signal transmission. One solution is to use light to send signals, rather than the electricity. Photonic chips will lead to higher performance because light produces less heat than electricity, and crucially travel faster. Many companies are looking to develop these chips. In order to create such a photonic chip, many channels of data are needed to be running simultaneously. And each channel needs ~50mW of optical power. The current cutting-edge lasers can only just meet this specification. Which means for each channel there must be a single laser, and each laser is large, power consuming, and integration is challenging. Vector is developing high end lasers using highly innovative photonic crystal design expertise and sees an opportunity to extend this work to enable world-leading photonic chip performance. Vectors photonic crystal technology is capable of delivering a laser with significantly higher optical power (up to 1W), enabling a single laser to drive multiple channels. This results in a system with significantly reduced energy consumption, a much lower component count, a much simpler system architecture and reduced cost. A system where you have "multiple channels per laser" rather than "multiple lasers per channel". This project will develop a 1W laser capable of supporting 20 channels on a photonics chip -- an ambitious objective which if achieved could revolutionise the field. The photonics crystal surface emitting laser (PCSEL) is a paradigm shift in laser functionality and breaks the performance vs cost trade-off present in current semiconductor lasers. PCSELs for datacentre optical interconnects are smaller, cheaper, more efficient and easier to integrate into systems than current solutions. Vector photonics will enable system integrators to differentiate themselves against people who use legacy technologies and better meet the needs of their customers.

Artificial Intelligence (AI) is set to revolutionise modern life, by enabling autonomous vehicles, real time medical diagnosis, financial modelling, power grid management, etc. However, the growth in the compute power necessary to develop future AI systems will run up against a data transmission bottle neck in compute power. This bottleneck is caused by delays in electrical signal transmission. One solution is to use light to send signals, rather than the electricity. Photonic chips will lead to higher performance because light produces less heat than electricity, and crucially travel faster. Many companies are looking to develop these chips. In order to create such a photonic chip, many channels of data are needed to be running simultaneously. And each channel needs ~50mW of optical power. The current cutting-edge lasers can only just meet this specification. Which means for each channel there must be a single laser, and each laser is large, power consuming, and integration is challenging. Vector is developing high end lasers using highly innovative photonic crystal design expertise and sees an opportunity to extend this work to enable world-leading photonic chip performance. Vectors photonic crystal technology is capable of delivering a laser with significantly higher optical power (up to 1W), enabling a single laser to drive multiple channels. This results in a system with significantly reduced energy consumption, a much lower component count, a much simpler system architecture and reduced cost. A system where you have "multiple channels per laser" rather than "multiple lasers per channel". This project will develop a 1W laser capable of supporting 20 channels on a photonics chip -- an ambitious objective which if achieved could revolutionise the field. The photonics crystal surface emitting laser (PCSEL) is a paradigm shift in laser functionality and breaks the performance vs cost trade-off present in current semiconductor lasers. PCSELs for datacentre optical interconnects are smaller, cheaper, more efficient and easier to integrate into systems than current solutions. Vector photonics will enable system integrators to differentiate themselves against people who use legacy technologies and better meet the needs of their customers.

Quantum computing is a rapidly emerging technology offering transformative changes to society as a whole by providing vast improvements in computational capability that will solve complex many-body problems that are currently intractable. It will potentially deliver advancements in diverse fields such as finance, climate change, infrastructure planning, drug discovery, secure communications and material science. Trapped-ion Quantum Computing (TIQC) systems are one of the most advanced and promising quantum computing platforms in which an oscillating electric field is used to confine ions which serve as the qubits used to encode quantum information. This approach offers a route towards scaling up the number of qubits and thereby delivering the increase in computing power that is ultimately desired, allowing the technology to emerge from small scale lab-based experimental environments to integrated user-friendly systems for everyday use. Laser sources are a key requirement in TIQC, performing essential system functions including ionisation, cooling, repumping and spectroscopy. Typically these different requirements are served by a wide range of laser sources, each with different wavelengths and performance requirements. The ADRENALIN project will develop a novel type of laser, a Photonic Crystal Surface Emitting Lasers (PCSEL) for use in QT applications. PCSELs employ photonic crystals to produce 2nd order out-of-plane diffraction and enable vertical, single-frequency emission. This novel device architecture provides excellent beam quality compared to other laser diodes and significantly reduces manufacturing costs. PCSELs can also be configured in 2D arrays with steerable individually addressable output, enabling different lattice sites to be addressed simultaneously. In addition these devices can be manufactured in most III-V semiconductors, allowing most of the wavelength range used in QT applications to be addressed. The many advantages of the PCSEL device will help facilitate scaling in next generation TIQC systems to accommodate larger numbers of qubits thereby enabling exponential increases in computational power and more widespread utilisation of the technology. The PCSEL will also help drive miniaturisation in QT applications -- this is important in TIQC but is also a key driver in the development of miniature atomic clocks for portable high-precision time-keeping, enabling a more widespread adoption of the technology and providing the potential to significantly advance improvements in transportation, defence and communication sectors. In both applications, PCSELs will ultimately displace incumbent light sources which typically rely on relatively bulky, expensive and complicated external cavity lasers and will become essential components in future QT systems.

Project QUDITS is a feasibility study which aims to develop a demonstrator platform to showcase the feasibility of developing quantum communication systems using qudits based on orbital angular momentum (OAM). By using using commercially available novel photonics technologies from the UK supply chain, photonic crystal surface-emitting lasers (PCSELs) and low-noise Avalanche Photo-Diodes (ALDs), able to operate at optical communications wavelengths. Quantum information is shaped around the use of qubits, the quantum analogy to the standard bit. This is a two-level, binary system, which is well known and has been used for many years. All quantum technologies currently being commercialised are based on qubits as the building block of quantum information. However, a two-level system inherently limits the density of information that can be carried in a quantum system. Higher dimensional Hilbert states of quantum information exist, known as qudits, and have more than two discrete states and can carry more information. The QUDITS project is developing a new area of quantum technologies for a potentially disruptive future communication system that will greatly enhance the state-of-the-art. It will demonstrate the feasibility of generating and detecting qudits from commercially available components from the UK supply chain. Qudits are a natural scale up technology for communication systems, enabling more data to reside on one quantum state, instead of having to send more qubits.

Additive manufacturing (AM), or 3D printing, is an enabling technology and is the future of manufacturing in a digital, local, low turnaround time, personalised world. Industry 4.0 aims to update manufacturing practice bringing it into the digital world through creativity, intelligence and connectivity. Direct digital manufacturing by allowing user access to 3D printing machines, is a key driver for industry 4.0, providing custom parts with rapid turn-around and reduced environmental impact. Designing, ordering, and delivery is now possible within 24 hours. This expanding field has the potential to revolutionise many aspects of human life. The plastic and metal parts which can be manufactured through 3D printing can significantly reduce waste during the manufacturing process, as well as reduce the weight of parts shipped (reducing carbon emission for key sectors such as aerospace). These systems can also produce individually tailored parts for medical applications, such as orthopaedic implants, giving improved quality of life over a one size fits all approach. Selective laser melting (SLM), is a leading AM process for making metal parts. Current SLM systems repeatedly deposit a layer of metal powder on a bed, then a high-power laser is scanned over the surface, melting only areas which require metal deposition on the layer below, akin to writing with a pen. This is repeated layer by layer, building a complete 3D structure. The SLM systems are expensive, and \\\>60% of the cost of the parts manufactured in this way are time dependant, holding back the deployment of these systems. We will develop a rapid 3D printing technology using individually addressable direct diode (iADD) semiconductor laser arrays. The novelty in this work is the development of new laser systems based on photonic (light controlling) structures pattered on the nanoscale. These new laser arrays can allow an entire layer to be written in one go instead of one spot at a time. The full 3D parts can then be built between 4 to 10 times faster using our 1D array SLM system and between 10 and 100 times faster using our 2D array SLM system. The lasers that produce the power to melt the powder require extensive lifetesting, which is essential to guarantee that the 3D printer is reliable and the parts made are robust. We predict that this time and cost reduction will do for additive manufacturing what the printing press did for publishing.

Additive manufacturing (AM) is an enabling technology and is the future of manufacturing in a digital, local, low turnaround time, personalised world. This new field has the potential to revolutionise many aspects of human life. The materials which can be manufactured using this process can significantly reduce waste during the manufacturing process, as well as reduce the weight of parts shipped (reducing carbon emission for key sectors such as aerospace). These systems can also produce bespoke parts for medical applications, giving improved quality of life. Direct digital manufacturing, as part of industry 4.0, will become a future norm, providing custom parts in a timely manner with reduced environmental impact. Selective laser sintering (SLS), is a leading AM process. Current SLM systems repeatedly deposit a layer of material powder on a bed, then a high-power laser is scanned over the surface, melting only areas which require material deposition on the layer below, akin to writing with a pen. This is repeated layer by layer, building a complete 3D structure. The SLM systems are expensive, and \>60% of the cost of the parts manufactured in this way are time dependant, holding back the deployment of these systems. We have developed a method to massively increase the write speed of these systems (increasing the speed by 10x), which reduces the cost of parts manufactured by 4x. We propose the development of a rapid 3D printing technology based on semiconductor laser technology. The novelty in this work is the development of new laser systems which will allow an entire layer to be written in with a speed increase of 10x, which equates to a reduction in cost of part production of 4x. This will do for additive manufacturing what the printing press did for publishing.

Additive manufacturing (AM), or 3D printing, is an enabling technology and is the future of manufacturing in a digital, local, low turnaround time, personalised world. Industry 4.0 aims to update manufacturing practice bringing it into the digital world through creativity, intelligence and connectivity. Direct digital manufacturing by allowing user access to 3D printing machines, is a key driver for industry 4.0, providing custom parts with rapid turn-around and reduced environmental impact. Designing, ordering, and delivery is now possible within 24 hours. This expanding field has the potential to revolutionise many aspects of human life. The plastic and metal parts which can be manufactured through 3D printing can significantly reduce waste during the manufacturing process, as well as reduce the weight of parts shipped (reducing carbon emission for key sectors such as aerospace). These systems can also produce individually tailored parts for medical applications, such as orthopaedic implants, giving improved quality of life over a one size fits all approach. Selective laser melting (SLM), is a leading AM process for making metal parts. Current SLM systems repeatedly deposit a layer of metal powder on a bed, then a high-power laser is scanned over the surface, melting only areas which require metal deposition on the layer below, akin to writing with a pen. This is repeated layer by layer, building a complete 3D structure. The SLM systems are expensive, and \>60% of the cost of the parts manufactured in this way are time dependant, holding back the deployment of these systems. We will develop a rapid 3D printing technology using individually addressable direct diode (iADD) semiconductor laser arrays. The novelty in this work is the development of new laser systems based on photonic (light controlling) structures pattered on the nanoscale. These new laser arrays can allow an entire layer to be written in one go instead of one spot at a time. The full 3D parts can then be built between 4 to 10 times faster using our 1D array SLM system and between 10 and 100 times faster using our 2D array SLM system. We predict that this time and cost reduction will do for additive manufacturing what the printing press did for publishing.

The COVID-19 crisis has placed an unprecedented demand on communication networks. The OECD recommends that "Network operators should anticipate increased demand and prevent congestion by upgrading their interconnection capacity" (http://www.oecd.org/coronavirus/policy-responses/keeping-the-internet-up-and-running-in-times-of-crisis-4017c4c9/) Datacentres require low latency, ultra-high capacity optical data connections. At present data centres are large scale and interconnected by long-haul fibre-optic connections. Within it a data centre has many short range (100-300m), low cost optical links, driven by low cost vertical cavity surface emitting lasers (VCSELs). The long-haul fibre connections (100's-1000's km) are driven by high cost and high-performance edge emitting lasers (EELs). Future internet of things (IoT) and 5G cellular roll-out will result in sustained growth in the volume of data storage and the volume of data movement which will require changes in the data centre landscape. Initially, an increase in the size of datacentres, requiring longer link lengths is expected. Additionally, low latency and problems in situating centres in urban areas, we expect distributed datacentres, requiring a step up in link-lengths approaching those of current metro networks. This change in system architecture cannot be supported by the present low cost VCSEL technologies as low powers limit the link length to a few hundred metres. Current EEL technologies can offer the required performance but not at the required cost. Following on from the commoditisation of active optical cables for datacentres is another opportunity in future domestic data cables. 8K TV and high speed USB are fighting the limits of copper connectivity, and a switch to optical cabling is now on the horizon. Cost is the ultimate driver for domestic markets, and our establishment in data centre cabling should put us in an ideal position to pursue this large market in the future. We have developed a new class of laser that due to its unique design allows performance that is better than that of both existing laser types; EELs and VCSELs. It comes with added wavelength agility (i.e. it can be applied to almost any emission wavelength) and has major cost advantage. The cost advantage from surface emission allows for; on-wafer testing, the packaging of only known good die, and the opportunity of on-wafer burn-in. A symmetric low divergence beam further reduces packaging costs beyond the VCSEL. A key element to the high-speed operation of these devices, allowing high data-rate transmission lies in size scalability to reduce the laser volume. The technology we have developed will allow the UK to be at the heart of the global photonics market. And with Governments increasingly moving to cloud data storage, access to this supply chain is a key strategic capability.