Sifted spoke to VCs from Atmos Ventures, Seraphim Space, Omnes Capital, FORWARD.one and Amadeus APEX Technology Fund
Kai Nicol-Schwarz 5 min read
By 2023 standards, semiconductor startups have had a very good year.
There has been a flurry of eye-watering finance packages for chip companies rolled out over the past two years by countries looking to secure their tech sovereignty — and VCs have smelt opportunity.
Funding in the sector hit record numbers in 2023, according to Dealroom, rising to $1.4bn from $1.2bn in 2022. This at a time when, across European tech, total capital raised has fallen by nearly half compared to 2022.
French startup Aledia, which builds chips for display technology for AR and VR, and Oxford Quantum Circuits, which is developing quantum computers using superconducting circuits, have both raised nine-figure sums.
But which smaller startups are poised for big things in the semiconductor sector going into 2024? To find out, Sifted asked VCs from Atmos Ventures, Seraphim Space, Omnes Capital, FORWARD.one and Amadeus APEX Technology Fund.
Beau-Anne Chilla,
partner at FORWARD.one
FORWARD.one is a Dutch early-stage VC fund, focusing on hardware technology.
FaradaIC — Germany
FaradaIC is developing novel materials and techniques to enable miniaturised gas-sensing chips. Growing concerns over climate change are paving the way for a multitude of gas-sensing chip applications in health, energy and foodtech.
Innatera — The Netherlands
A spin-off from the Delft University of Technology, Innatera is developing neuromorphic processors — which are based on the structures of biological brains. The processors have applications like true presence detection, voice recognition and autonomous vehicles.
Skycore Semiconductors — Denmark
Skycore is building high-voltage integrated circuits for applications ranging from consumer electronics to power grids. With electrification and digitisation we are seeing increasingly higher voltages in our electronics and energy systems, and Skycore’s chips enable top performance in combination with design flexibility.
Sami Moughrabie,
managing partner at Atmos Ventures
Atmos Ventures is a US-headquartered deeptech VC fund, backing startups at seed and Series A.
Mignon — UK
Mignon is an advanced AI accelerator architecture design technology with up to 10k times improvements in on-chip inference. Pioneering on-chip training at scale, Mignon’s propositional logic-based design demands fewer resources, enabling intelligent compute on energy-efficient devices with limited internet coverage.
5x a week
~/ Sifted Daily
Setting the agenda for startup Europe and telling you why it matters.
By Sifted journalists
Black Semiconductor — Germany
Black Semiconductor is taking photonic-based microchip technology — which uses light to transfer data instead of electricity — out of research labs and into industrial, mass-producible applications. Photonic interfaces on chips can process data 100 to 1,000x faster than standard chips. This technology is crucial because there is an exponential demand for information and data processing. For example, in the case of rapidly growing AI-based systems, or hugely expanding datasets in data centres.
HIDRA — UK
HIDRA is in the process of spinning out from the UK’s National Physical Laboratory with its semiconductor wafer imaging technology. HIDRA rapidly captures high-resolution images, improving semiconductor yield — the number of chips that are produced to the maximum chip count on one wafer. The startup has secured support from ChipStart UK, a UK government-funded incubator program for semiconductor advancements.
Maia Korradi,
principal at Omnes Capital
Omnes Capital is a Paris-based multistage VC fund.
Akhetonics — Germany
Akhetonics is building a photonic general-purpose CPU/GPU. This technology promises to overcome the core limitations of electronic devices as photonic computers could be designed to be highly parallel — where a large number of computer processors can simultaneously perform a task — and energy efficient.
LIGENTEC — Switzerland
LIGENTEC manufactures photonic integrated circuits (PIC) which are used in quantum tech, ground-to-space communications and biosensors. The startup’s circuits bring advantages like small carbon footprints, high manufacturing scalability, low cost, high performance and power efficiency and low heat generation.
Maureen Haverty, Principal At Seraphim Space
Seraphim Space is a London-based spacetech VC fund.
Quobly — France
Quobly is working on quantum computing using silicon spin qubit processors — which create customisable quantum computing systems that can work for a wide range of applications. Qubits can operate at higher temperatures than many other quantum systems, significantly simplifying and reducing the cost of the cooling technology required.
Salience Labs — UK
Salience Labs is building a multi-chip processor using photonics to transmit data combined with traditional microelectronics. This combination should make Salience Labs’ chips compatible with current systems but able to tackle the growing demands of AI models.
SEMRON — Germany
SEMRON is building AI chips that can be used for things like autonomous vehicles, smart manufacturing, healthcare monitoring and smart home devices. The startup was part of accelerator Intel Ignite’s most recent cohort.
Wolfgang Neubert, General Partner At Amadeus Apex Technology Fund
Amadeus APEX Technology Fund is a partnership fund from the UK’s Amadeus Capital Partners and Austria’s APEX Ventures, backing early-stage deeptech startups.
GEMESYS — Germany
GEMESYS chips are designed to imitate the information-processing mechanisms of the human brain, a process known as neuromorphic computing. It enables AI hardware companies to distribute a novel chip that trains neural networks more efficiently than current technology, reducing cost and increasing performance. Meanwhile, its small size allows it to be embedded into nearly every device.
Synthara — Switzerland
Synthara helps chip makers to enhance their microcontroller (MCU) platforms. The startup says its product, a computational memory technology, makes MCUs 50x more efficient and is compatible with existing chip architectures and software applications. This allows developers to freely experiment with old and new neural network and signal processing algorithms.