Ninety-one startups have raised $2.6 billion in funding.
Artificial intelligence attracted more investors into the chip industry in the second quarter. Four AI-focused chip startups received over $100 million in financing, targeting transformer data center ASICs, highly flexible platforms for embedded edge, dataflow processors, and mixed-signal neuromorphic chips. Memory computing also contributes to the advancement of artificial intelligence, with three companies either integrating it into chips or providing specialized IP.
Methods of connecting small chips, chips, and server racks are another area of activity. Germany invested over $200 million in graphene-based co-integrated optical systems to achieve large-scale parallel connections. Other optical interconnection methods have also garnered attention, as well as a method for manufacturing ultra-high-density interconnections using liquid metal ink.
This report covers 91 startups in the second quarter of 2024, which collectively raised $2.6 billion.
Chips
Artificial Intelligence Hardware
Electronic Design Automation (EDA)Manufacturing and Equipment
Testing, Measurement, and Inspection
Materials
Memory and Storage
Photonics and Optics
Sensors
Security
Previous Reports
(Note: The " " in the original text is an HTML entity used to represent a non-breaking space. It has no direct English translation and is not typically included in plain text translations.)MetisX has raised $44 million in Series A funding, with investors including SV Investment, STIC Ventures, LB Investment, and the Korea Development Bank, as well as Mirae Asset Venture Investment, Mirae Asset Capital, IMM Investment, SBI Investment, Tony Investment, and Wonik Investment Partners. MetisX is a fabless company developing CXL-based memory-centric data domain-specific architectures to overcome performance bottlenecks associated with frequent memory access. The startup claims its intelligent memory architecture can accelerate large-scale data processing applications, such as vector databases, big data analysis, AI data preprocessing, and DNA analysis. It plans to produce its first chip by early next year. Founded in 2022 and headquartered in Seongnam, South Korea, the company is poised to make significant advancements in the chip industry.
Flow Computing stands out with a €4 million (approximately $4.3 million) pre-seed funding round led by Butterfly Ventures, with participation from FOV Ventures, Sarsia, Stephen Industries, Superhero Capital, and Business Finland. Flow Computing offers what it calls a Parallel Processing Unit (PPU), which can be integrated into any CPU design architecture, instruction set, or process geometry with minimal modifications to the CPU. The startup claims that the PPU IP can increase CPU performance by 100 times by hiding memory reference latency by executing other threads during memory access, reducing the number of thread synchronizations needed, and overlapping them with execution, organizing functional units in a chain where units can use the results of their predecessors as operands. It can be customized according to specific requirements, including the number of PPU cores, the types and quantities of functional units, and the size of on-chip memory resources. A spin-off from the VTT Technical Research Centre of Finland, the company was founded in 2024 and is headquartered in Helsinki, Finland.
Morphing Machines has secured $2.8 million in seed funding, led by Speciale Invest, with IvyCap Ventures, Golden Sparrow, Navam Capital, CIIE Initiatives, and DeVC participating. Morphing Machines is developing a multicore, massively parallel SoC platform that can scale from 16 cores to 4K cores and is reconfigurable at runtime, enabling the instantiation of domain-specific architectures for mixed-criticality application tasks based on event demands. It targets applications in avionics, automotive, and 5G/6G telecommunications. Founded in 2006 and headquartered in Bangalore, India, the company is set to revolutionize the chip industry with its innovative approach.
Artificial Intelligence Hardware
Rivos secured over $250 million in Series A-3 funding, led by Matrix Capital Management, with new investors including Intel Capital, MediaTek, Cambium Capital, CIDC Consultants, Capital TEN, and Hotung Venture Group, as well as existing investors Walden Catalyst, Dell Technologies Capital, Koch Disruptive Technologies, and VentureTech Alliance. Rivos designs software-defined hardware solutions for accelerating data center workloads, particularly data analytics and generative AI. Its chips combine high-performance server-grade RISC-V CPUs and general-purpose GPUs (GPGPUs), optimized for LLMs and data analytics, with a unified memory shared between DDR DRAM and HBM. It also offers an open software stack from firmware to applications. The funds will be used to launch its first silicon product and expand manufacturing operations, platform hardware, software engineering, and support functions. Founded in 2021 and headquartered in Santa Clara, California, the company is at the forefront of AI hardware innovation.
Etched raised $120 million in Series A funding, led by Primary Venture Partners and Positive Sum Ventures, with participation from Two Sigma Ventures, Skybox Datacenters, Hummingbird Ventures, Oceans Ventures, Fundomo, Velvet Sea Ventures, Fontinalis Partners, Galaxy Digital, Earthshot Ventures, Max Ventures, Lightscape Partners, and individual investors. Etched is building an application-specific integrated circuit (ASIC) specifically for Transformer model inference. Transformers are currently the primary architecture for generative AI. The startup claims that by running a single algorithm, it can eliminate the vast majority of control flow logic, resulting in more math blocks and an FLOPS utilization rate of over 90%. Etched claims that a server containing eight Sohu chips can run over 500,000 Llama 70B tokens per second. Founded in 2022 and headquartered in Cupertino, California, the company is poised to make significant contributions to the AI hardware industry.
Hailo completed a Series C funding round, raising $120 million, led by Delek Motors, DCLBA, Vasuki, OurCrowd, Talcar, Comasco, Automotive Equipment Group, Poalim Equity, and individual investors. Hailo manufactures AI processors for edge devices. The company has developed a domain-specific structure-defined data flow processor architecture that allows software to allocate compute, memory, and control blocks for different neural network layers as needed. The startup recently launched a new AI accelerator specifically for processing low-power large language models (LLMs) for generative AI in personal computers and automotive infotainment systems. The company also offers general-purpose AI processors and a range of AI vision processors for smart cameras, available in the form of SoCs and modules, complete with a full software stack. Founded in 2017 and headquartered in Tel Aviv, Israel, the company has raised over $340 million to date.
Blaize secured $106 million in funding from existing investors such as Bess Ventures, Franklin Templeton, DENSO, Mercedes Benz, and Temasek, as well as new investors like Rizvi Traverse, Ava Investors, and BurTech LP. Blaize provides a full-stack programmable graphics stream processor architecture suite and a low-code/no-code software platform for edge AI across industries such as automotive, mobile, retail, security, industrial automation, and healthcare. The company's graphics stream processors form the basis of a range of products, including system-level modules, M.2 cards, PCIe accelerator cards, and embedded kits. Blaize is going public on NASDAQ through a merger with SPAC BurTech Acquisition Corp. Founded in 2010 and headquartered in El Dorado Hills, California, the company is a leader in AI hardware solutions.
DeepX secured $80.5 million in Series C funding, led by SkyLake Equity Partners, with participation from BNW Investments, AJU IB Investment, and TimeFolio Asset Management. DeepX is developing device NPU IP and AI SoCs for applications in physical security, robotics, and mobility. The startup claims that its NPU SoC can achieve a computational efficiency of over 10 TOPS/W with a power consumption of less than 3 watts by compressing FP32 to Int8 without compromising accuracy and reducing DRAM access. It also offers NPU design services, AI modules, and software design kits. The funds will be used for the mass production of its first product line and the development and launch of its next-generation LLM device solutions. Founded in 2018 and headquartered in Seongnam, South Korea, the company is at the forefront of AI hardware innovation.SiMa.ai has secured $70 million in a new round of financing, led by Maverick Capital, with participation from Point72 and Jericho Capital, as well as existing investors Amplify Partners, Dell Technologies Capital, Fidelity Management & Research Company, Lip-Bu Tan, and others. SiMa.ai builds hardware and software for machine learning workloads with low power consumption. The startup claims its platform can adapt to any framework, network, model, sensor, or pattern. Currently, it provides computer vision inference chips for embedded edge in industries such as manufacturing, retail, aerospace, defense, agriculture, and healthcare. Its software-centric MLSoC platform aims to be a simple method for deploying and scaling edge ML, including computer vision processors, video encoders/decoders, ML accelerators, and application processors. Its software supports static scheduling and double buffering, allowing for proactive data prefetching in a tiered manner before computation time, which SiMa.ai claims enables it to run models of any size. The platform is available in the form of SoC and PCIe production boards, which can be used both as standalone edge system controllers and as ML offloading accelerators for processors and ASICs. The startup is currently developing its second-generation edge multimodal generative AI chip, targeting a launch in early 2025. Founded in 2018 and headquartered in San Jose, California, the company has raised $270 million to date.
Axelera AI has raised $68 million in Series B funding, including Invest-NL, the European Innovation Council Fund, and Samsung Catalyst Fund, as well as existing investors Verve Ventures, Innovation Industries, Fractionelera, and CDP Venture Capital SGR. Axelera AI provides an edge AI acceleration platform for generative AI and computer vision inference, which utilizes a dataflow architecture based on SRAM-based digital memory computing and RISC-V control to minimize data movement between memory and computing elements. It also offers quantization technology and mapping tools that reduce AI computational load and improve energy efficiency while minimizing accuracy loss across various networks. The startup expects its AI processing unit to be in full production in the second half of 2024. The funds will be used to expand into new markets, including automotive and high-performance computing. Founded in 2021 and headquartered in Eindhoven, the Netherlands, the company has raised $120 million to date.
Expedera has raised $20 million in Series B funding led by an independent semiconductor company. Expedera provides high-performance, energy-efficient neural processing unit (NPU) IP for on-device AI. Its NPU IP targets applications such as automotive, AR/VR/MR, and smartphones, scaling up to 128 TOPS with a single engine and requiring minimal external memory bandwidth. It uses a packet-based architecture, dividing each layer into independently schedulable executable segments, enabling parallel execution across multiple layers, better resource utilization, and deterministic performance. The platform includes a software stack based on TVM, allowing the import of trained networks, offering various quantization options, auto-tuning, compilation, estimators, and analysis tools. The funds will be used to advance its IP and services and expand its global engineering footprint. Founded in 2018 and headquartered in Santa Clara, California, the company has raised $20 million to date.
Innatera has raised an additional $5 million in Series A funding, led by Innavest and Invest-NL, bringing the total funding in this round to $21 million. Innatera develops neuromorphic processors based on an analog mixed-signal computing architecture that mimics the brain's mechanism for processing sensory data. Innatera's processors use spiking neural networks, featuring ultra-low power consumption and short response latency, enabling always-on pattern recognition capabilities in sensor-edge applications. The funds will be used to accelerate mass production. Founded in 2018 as a subsidiary of Delft University of Technology, the company is headquartered in Rijswijk, the Netherlands.
RaiderChip has secured €1 million (approximately $1.1 million) in seed funding. RaiderChip is developing a hardware IP core to accelerate generative AI inference on FPGA platforms. The company claims that its approach, which employs large-scale floating-point parallelism and optimized memory bandwidth utilization, enables small to large language models to run natively on low-cost standalone or embedded devices. The funds will be used for marketing. Founded in 2024 as a subsidiary of Visengi, the company is headquartered in Solàres, Spain.
EDA
Itda Semiconductor has received ₩3 billion (approximately $2.2 million) in pre-Series A funding from We Ventures and L&S Venture Capital. Itda Semiconductor offers a tool for code-free design and optimization of power and clock systems in complex SoCs. Founded in 2022 and headquartered in Hwaseong, South Korea.
Manufacturing and Equipment
Eyco has secured €16 million (approximately $17.4 million) in funding from companies such as Bpifrance, Région Sud Investissement, and CappCreation. Eyco manufactures flexible circuits on ultra-thin film substrates with a thickness of just a few micrometers. The company has developed a roll-to-roll processing machine with features such as multi-layer and multi-material co-lamination, mechanical cutting, screen printing, etching, metal surface treatment, and electrification of insulating materials and microvias. The company focuses on equipment for the medical, payment, security, and telecommunications industries. The funds will be used for industrialization. Founded in 2020 and headquartered in Trélex, France.
Axus Technology has received $12.5 million in capital funding from IntrinSiC Investment. Axus Technology manufactures chemical mechanical planarization (CMP) equipment for semiconductor and compound semiconductor wafer polishing, thinning, and cleaning. Its surface processing solutions cover semiconductors, including silicon carbide, MEMS device processing, device packaging, and TSV/3D integration. The funds will be used to pursue and fulfill high-volume orders. Founded in 2002 and headquartered in Chandler, Arizona.Testing, Measurement, and Inspection
Liquid Instruments secured $12 million in funding, led by Breakthrough Victoria, with participation from Lockheed Martin Ventures, Acorn Capital, and Powerhouse Ventures. Liquid Instruments manufactures reconfigurable test and measurement equipment, combining multiple instruments into one using FPGAs. Its platform includes oscilloscopes, programmable power supplies, PID controllers, digital logic analyzers, arbitrary waveform generators, data loggers, spectrum analyzers, and more. The company's product range spans from portable devices for engineering education to more advanced devices for research and professional engineering, the latter allowing users to run and connect multiple instruments simultaneously to build custom test systems. The company targets a range of application fields from optics and photonics to aerospace and defense. The funds will be used to expand its manufacturing operations, establish new offices, and broaden its global business. Founded in 2014 and headquartered in Canberra, Australia.
Materials
Thintronics raised $23 million in Series A funding led by Maverick Capital and Translink Capital. Thintronics manufactures high-performance interconnect dielectrics for AI data centers, networks, and RF/millimeter wave (mmW) applications. The startup claims its ultra-thin dielectric layers can be used for chips, substrates, and PCBs in computing, switches, and networking, enabling interconnects up to 224 Gbps while reducing channel loss and thinning by 50%. Founded in 2019 and headquartered in Berkeley, California, USA.
LQDX received $10 million in equity and government funding. LQDX developed a suite of liquid metal metallization chemistries that can apply nanoscale layers of pure palladium, gold, or copper atom-by-atom to various substrate materials (including 3D surfaces) for advanced packaging interconnects in chips, chiplets, and heterogeneous integration. The company claims its technology can produce denser circuits than traditional methods, suitable for ultra-high-density interconnect substrates, complex multi-layer interposers, and advanced wafer-level fan-out technologies. Founded in 2007 as Averatek and headquartered in Santa Clara, California, USA.
Memory and Storage
Synthara raised $11.4 million in equity financing and grants, led by Vsquared Ventures, with participation from OTB Ventures, Onsight Ventures, Deeptech Labs, and existing investors such as High-Tech Gründerfonds, DeepIE Ventures, Excellis, and Zürcher Kantonal Bank. Synthara offers SRAM memory alternatives with integrated computing capabilities. It requires no foundry exemptions, is compatible with any CMOS process and ISA, and includes an SDK with optimized implementations of linear algebra subroutines, signal processing functions, and neural network layers. The memory computing technology targets embedded microcontrollers, and the startup claims it can increase energy efficiency and latency by 100 times by offloading up to 99% of model operations to memory, eliminating the need for dedicated AI accelerators in wearable devices, robotics, and smart sensing applications. Founded in 2019 as a spin-off from the Institute of Neuroinformatics at ETH Zurich and the University of Zurich, headquartered in Zurich, Switzerland.
Rain AI secured $8.1 million in Series A funding from Epic Venture Partners. Rain AI offers digital memory computing IP, specifically targeting device AI workloads that require ultra-low latency and high energy efficiency. It supports training and inference and uses quantization algorithms co-designed at the circuit level to maintain FP32 precision. The startup also developed proprietary interconnects between RISC-V and its D-IMC core, enabling AI developers to implement any operator and compile any model. Its first chip is expected to launch in 2025 and will target applications such as drones, VR, smartphones, robotics, and wearable devices. Founded in 2017 and headquartered in San Francisco, California, USA.
ZeroPoint Technologies raised €5 million (about $5.5 million) in Series A funding led by Matterwave Ventures, with participation from Industrifonden, Climentum Capital, and Chalmers Ventures. ZeroPoint develops memory compression IP. The company claims its technology can reduce data center energy consumption by combining ultra-fast data compression with real-time data compression and transparent memory management into an IP block integrated with existing industry-standard on-chip bus protocols. The technology targets data center servers and smart devices, compressing data across the entire memory hierarchy and is independent of data load, processor type, architecture, memory technology, and processing node. The funds will be used to expand sales, bring more hardware-accelerated memory products to market, and hire staff. Founded in 2016 as a spin-off from Chalmers University of Technology, headquartered in Gothenburg, Sweden.
RAAAM Memory Technologies received $4 million in seed funding from companies such as Serpentine Ventures, J-Ventures, HackCapital, Silicon Catalyst Angels, and Claves Investments. RAAAM developed an on-chip memory technology called Gain Cell Random Access Memory (GCRAM) that can store one bit of data with only three transistors. GCRAM bit cells utilize decoupled write and read ports, providing native dual-port operation. It is proposed as an embedded SRAM alternative with smaller area and power consumption, manufacturable using standard CMOS processes. Founded in 2021 and headquartered in Petah Tikva, Israel.Photonics and Optics
Black Semiconductor has secured €254.4 million (approximately $273.9 million) in its Series A funding round, led by Porsche Ventures and Project A, with participation from East Hill Equity, Capnamic, TechVision Fonds, and NRW.BANK, as well as public funding under the European IPCEI ME/CT program. Black Semiconductor has developed a graphene-based co-integrated optical system that acts as a transducer, converting electronic signals into optical signals and vice versa. The startup claims that its approach enables large-scale parallel optical chip connections over distances of several kilometers. The funds will be used to advance product development and the establishment of a 300mm wafer pilot production facility, with a long-term goal of mass production by 2031. Founded in 2019 and headquartered in Aachen, Germany, the company aims to revolutionize the field of integrated photonics.
Wave Photonics has received £4.5 million (approximately $5.8 million) in seed funding, led by the UK Innovation & Science Seed Fund and Cambridge Enterprise Ventures, with participation from Redstone QAI Quantum Fund, Kyra Ventures, Parkwalk Advisors, and Deep Tech Labs. Wave Photonics utilizes computational techniques to create tools and intellectual property (IP) for integrated photonics design. Its approach combines simulation, statistical modeling, and optimization techniques. The startup offers a photonic device kit (PDK) and photonic chip packaging services for prototyping and research and development. The funding will be used to transition its technology from a research production line to a commercial foundry, with a particular focus on solutions for applications in quantum technology and biosensing. Founded in 2021 and headquartered in Cambridge, UK, the company is poised to make significant contributions to the field of photonics.
NcodiN has secured €3.5 million (approximately $3.8 million) in funding from Elaia Partners, Earlybird Venture Capital, and OVNI Capital. NcodiN is developing optical interposer technology for chip integration. By integrating ultra-miniature III-V semiconductor nano-lasers and nano-detectors on-chip, the startup claims that its interposer can achieve an ultra-high integration density of over 10,000 components per square centimeter while enhancing data transfer speeds and efficiency. The startup is particularly focused on chip-based disaggregate processors and memory-centric computing architectures for high-performance computing and AI workloads. The funds will be used to accelerate the development of integrated optical link prototypes and company expansion. Founded in 2023 and headquartered in Palaiseau, France, the company is set to make a significant impact in the field of integrated photonics.
Ranovus has received a repayable investment of CAD 4.8 million (approximately $3.5 million) from FedDev Ontario. Ranovus develops monolithic silicon photonic interconnects for AI/ML workloads, high-performance computing, and hyperscale data centers. The startup's technology utilizes quantum dot multi-wavelength lasers (QD MWL), which can provide multiple wavelengths simultaneously from a single device, with each wavelength selectable and usable as a light source for data transmission. It is capable of generating up to 96 wavelengths in the C-band. The silicon photonic micro-ring resonator architecture enables high-speed modulation with data rates of up to 100Gbs/lambda. These components are combined with RF drivers, TIA, and control logic using wafer-level laser connection packaging processes. The funds will be used to expand in-house production capabilities. Founded in 2012 and headquartered in Ottawa, Ontario, Canada, the company is at the forefront of silicon photonics.
Lightium has received a grant of CHF 2.7 million (approximately $3 million) from InnoSuisse. Lightium has developed a thin-film lithium niobate (TFLN) process for the manufacturing and packaging of photonic integrated circuits. The startup, which intends to provide foundry services, claims that its TFLN platform can support data transfer speeds of over 1.6 Tb/s. Initially targeting designs in the telecommunications and data center sectors, the platform is said to have applications in satellite communication, quantum computing, and LiDAR markets. The funds will be used to fine-tune its PDK and transfer its manufacturing process to a commercial CMOS fab for mass production. Lightium expects to open beta access to its TFLN platform by the end of 2024. Founded in 2023 and headquartered in Zurich, Switzerland, the company is poised to make significant contributions to the field of photonic integrated circuits.
Sensors
Bitsensing has secured $25 million in Series B funding, with investors including the Korea Development Bank, HL Mando Corporation, the Korea Industrial Bank, Aju Capital, Life Asset Management, and SCL Investment. Bitsensing produces a range of radar systems, including a 79GHz 4D imaging radar for automotive applications with a range coverage of over 300 meters, as well as a 24GHz radar with integrated AI perception for smart traffic management. The company has expanded into the health market with the launch of a radar that can monitor sleep quality, apnea events, limb movement, and provide data analysis. The funds will be used to scale operations and R&D. Founded in 2018 and headquartered in Seoul, South Korea, the company is set to make a significant impact in the field of radar technology.
Security
TXOne Networks has raised $51 million in Series B funding from existing investors TGVest Capital, Cathay Financial Holdings, Cathay-Crea, as well as new investors Taishan Capital and Applied Ventures ITIC Innovation Fund. TXOne Networks provides cybersecurity solutions that ensure the reliability and security of industrial control systems (ICS) and operational technology (OT) environments through a zero-trust approach. It offers network-based and endpoint-based products that use real-time, in-depth defense methods to protect OT networks and mission-critical equipment across a range of industries, including semiconductor facilities. Founded in 2019 as a joint venture between Trend Micro and Moxa Technologies, the company is headquartered in Taiwan, China.PQShield has secured a $370 million Series B funding round led by Addition, with participation from Chevron Technology Ventures, Legal and General, Braavos Capital, and Oxford Science Enterprises. PQShield provides hardware components, subsystems, and accelerators for post-quantum cryptographic algorithms and side-channel analysis protection. It also offers software implementations of post-quantum and classical cryptographic primitives. Founded in 2018 and headquartered in Oxford, UK, the company specializes in these areas.
Niobium Microsystems has received $5.5 million in seed funding. Niobium manufactures System-on-Chips (SoCs) and hardware accelerator cards for Fully Homomorphic Encryption (FHE), combining massively parallel computing circuits with compilers optimized for large-scale memory efficiency to maximize data reuse during computation. The funding will be used for the development of FHE acceleration, recruitment, and product optimization for commercial applications. Established in 2021 and spun out from Galois, the company is based in Dayton, Ohio, USA.
In other technologies, Frore Systems has raised $80 million in Series C funding led by Fidelity Management & Research Company, with participation from existing investors including Prosperity7, Mayfield Fund, Clear Ventures, Addition, Qualcomm Ventures, MVP Ventures, Stepstone Group, and Alumni Ventures. Frore Systems manufactures solid-state active cooling modules that sit on top of chips and dissipate heat without the need for fans. The piezoelectric MEMS devices contain tiny membranes that vibrate at ultrasonic frequencies to draw in airflow, which then becomes a high-velocity pulsating airflow that carries heat away from the heatsink. The startup claims these modules can remove more heat than fans while remaining silent, vibration-free, and dust-proof. It targets a range of systems, including laptops and other consumer devices, as well as connected edge AI gateways and data centers. Founded in 2018 and headquartered in San Jose, California, USA, the company has raised $196 million to date.
In terms of funds and investors, imec.xpand has launched a new €300 million (approximately $320.9 million) fund that will invest in semiconductor and nanotechnology startups with the potential to drive semiconductor innovation beyond traditional applications and propel the next generation of technology.
Silicon Catalyst Ventures aims to invest between $10 million and $20 million in its first round of funding this year. The company will primarily support early-stage semiconductor startups in North America, the UK, the EU, and Israel, focusing on artificial intelligence, communications, photonics, MEMS, sensors, intellectual property, materials, and life sciences.
PhotonVentures has added €15 million (approximately $16 million) in its second round, bringing its total funds to €75 million ($79.8 million) for investing in startups and scale-ups developing photonic chip solutions.
Intel Ignite has selected several startups from the US, UK, Europe, and Israel to participate in its accelerator program for early-stage deep-tech startups. The selected startups include those working on EDA, AI chips, advanced packaging photonic interconnects, memory compression IP, holographic data storage, and conformal imprint lithography technologies.
China will invest 344 billion yuan (approximately $47.5 billion) in the third phase of the National Integrated Circuit Industry Investment Fund (also known as the Big Fund) to support its semiconductor industry and supply chain.*Disclaimer: This article is an original work by the author. The content of the article represents the author's personal views. Our reposting is solely for the purpose of sharing and discussion, and does not represent our endorsement or agreement. If there are any objections, please contact the backend.
Comment