Cambridge GaN Devices

Cambridge GaN Devices (CGD) is a deeptech spin-out from Cambridge University that is working in the innovation of gallium nitride (GaN) power devices. It is a fabless semiconductor company that develops a range of energy-efficient GaN-based power devices in order to make greener electronics possible.

The startup is focused on delivering highly efficient and reliable transistors that outperform traditional silicon-based components. GaN-based devices represent a breakthrough in power electronics, offering faster switching speeds, lower energy consumption, and more compact designs than their silicon-based counterparts. CGD's ICeGaN technology offers enhancement- mode GaN transistors that operate like conventional silicon MOSFETs without requiring additional gate drivers or circuitry.

 This advancement enables easier integration, improved efficiency, and lower power consumption for applications in data centres, industrial power systems, and automotive sectors. The technology simplifies the implementation of GaN into existing and progressive designs, delivers efficiency levels exceeding 99%, and enables energy savings of up to 50% in a wide range of high- power applications including electric vehicles and data centre power supplies.

As the demand for high-performance, energy-efficient electronics increases, CGD is well-positioned in the industry to drive innovation in power semiconductors. By leveraging GaN’s superior properties, the company is shaping the future of power electronics and contributing to a more sustainable technological landscape.

GEMESYS

German startup GEMESYS is developing an analog AI chip design based on the information processing mechanism of the human brain. By integrating memristors, electronic components that mimic synaptic behaviour, GEMESYS aims to revolutionise AI processing, and make it significantly more energy efficient.

Unlike traditional digital processors, GEMESYS’ chips operate using analog computation, which allows for more natural and efficient neural network processing. The startup claims that its technology could make AI training up to 10,000 times more energy-efficient than current digital alternatives. This efficiency could enable powerful AI models to run directly on Edge devices, such as smartphones and hearing implants, which would reduce the need for cloud-based processing, and enhances data privacy.

GEMESYS’ approach aligns with a growing industry shift towards Edge AI, where computation happens closer to the user, rather than in remote data centres elsewhere. By eliminating the need to transmit raw data to the Cloud, GEMESYS’ technology would be able to reduce latency and enhance security while also lowering the environmental footprint of AI operations, which is becoming an ever-growing concern.

AllFocal Optics

Cambridge-based AllFocal Optics is pioneering a new approach to augmented reality (AR) and virtual reality (VR) through its nanophotonic lens technology. By replicating the way the human eye focuses, the company aims to eliminate the discomfort, nausea, and visual strain that often comes with extended AR and VR use.

At the heart of AllFocal’s innovation is its advanced nanophotonic lenses, which dynamically adjust focus in real time, providing a seamless transition between virtual and real-world environments. This technology addresses one of the key challenges in AR and VR: the disconnect between traditional optics and the way the human eye naturally perceives depth and focus.

If the company’s upcoming trials prove successful, its nanotech could integrate into a range of everyday applications. Car windshields could display sat-nav directions directly onto the glass, allowing drivers to navigate without glancing away from the road. Surgeons and doctors might use smart glasses to access critical patient data hands-free during procedures. Meanwhile, gamers could experience immersive virtual worlds without the dizziness that typically comes with extended headset use.

The startup has attracted major industry players, with Dr Ash Salusbury, former Microsoft Vice President of Technology and Meta’s AR programme leader, joining as chair. It is also collaborating with Jaguar Land Rover, with technology trials scheduled to take place this year.

SemiQon

SemiQon is a Finnish startup that specialises in silicon-based quantum processors, aiming to make quantum computers more scalable, affordable, and sustainable. The company’s mission is to develop silicon-based quantum processors that enable the scaling up of quantum computing. By leveraging existing semiconductor manufacturing infrastructure, SemiQon aims to produce quantum processors that are commercially competitive and suitable for mass production.

SemiQon was founded in 2022 as part of VTT’s Launchpad incubator programme, and began independent operations in early 2023. By the time February 2023 rolled around, SemiQon had secured €2 million in funding. Most recently in February 2025, the startup secured €17.5 million in funding, including €2.5 million in non-dilutive grants, and €15 million in equity-based funding from the European Innovation Council.

SemiQon’s approach enables quantum computers to operate at warmer temperatures compared to traditional quantum systems, significantly reducing the cooling requirements. Its technology is designed to make quantum computing more accessible by integrating seamlessly with established semiconductor manufacturing processes.

This article originally appeared in the March/April 2025 issue of Startups Magazine. Click here to subscribe