The EE Times Silicon 100, now in its 24th edition, highlights the vibrancy of an industry and research community with a constant drive to innovate, said Anne-Françoise Pelé, editor-in-chief of EE Times Europe and who also oversees the Silicon 100 report. The annual selection of the startups, curated by veteran technology and business journalist Peter Clarke, examines a mix of criteria in the selection of the companies. These include technology, market segment, maturity, financial position, fundraising activity and executive leadership.
These Silicon 100 startups have the potential to deliver technologies that will have a huge impact on the future of the semiconductor industry as well as on the design and development of electronic systems and devices. This year’s newcomers focus on a wide range of technology areas, including gallium nitride, battery power, indoor photovoltaics, LiDAR and smart sensors, backscatter and high-frequency RF, photonic interconnects, chiplets and processors.
This year’s edition covers 24 technology categories, ranging from materials and packaging to quantum computing and security. One of those categories is 5G & RF, highlighted by three startups: EdgeQ Inc., Forefront RF Ltd. and Picocom Ltd. These companies aim to deliver advanced technologies that further drive 5G adoption by making it easier to implement and deploy next-generation 5G features and systems.
EdgeQ, founded in 2018, aims to put 4G and 5G base stations and AI on the same system-on-chip (SoC). The company is led by executives from Qualcomm, Intel and Broadcom. They are building a converged connectivity and AI platform that is software-customizable and programmable.
EdgeQ’s 4G/5G + AI base-station-on-chip for Open RAN–based radio and distributed units delivers a combination of highly integrated silicon with production-ready 5G PHY software. This solution enables customers to deploy all key functionalities and critical algorithms of the RAN, such as beamforming, channel estimation, massive MIMO and interference cancellation out of the box, all aimed at reducing their resources and time.
At MWC Barcelona 2024, EdgeQ demonstrated a programmable, all-in-one, multi-user, 4G + 5G + AI small cell delivering over 2 Gbits/s of throughput at half the power, and the industry’s first end-to-end, 32-antenna, Open RAN massive MIMO with over 5 Gbits/s of throughput running at <50 W, according to the company.
Forefront RF, founded in January 2020, is focused on using adaptive passive cancellation of RF signals to replace BAW and SAW filters and switch banks in mobile phones and wearables across 3G to 6G. By replacing fixed-frequency filters with tunable alternatives, the benefits include reduced space, weight and cost. Forefront RF is a new addition to this year’s list.
The fabless semiconductor company believes its patented tunable duplexer technology will redefine mobile-radio front ends in smartphones, smartwatches and IoT systems by replacing fixed-frequency filters with a tunable duplexer.
The solution can reduce the number and types of components needed to support the multiple RF frequency bands by replacing them with a single frequency-agnostic module that is dynamically tunable according to available frequencies and/or services needed, according to the company. This would simplify RF architectures while reducing component count and cost.
Picocom, founded in 2018, targets Open RAN–compliant baseband SoCs and carrier-grade software products for 5G small-cell infrastructure. The company introduced the PC802 4G/5G SoC for disaggregated 5G small-cell platforms in 2021, followed by the recent introduction of the PC805 SoC, a purpose-designed PHY for 5G NR and LTE small-cell Open RAN radio units (O-RUs) that is fully compliant with Open RAN specifications.
With the launch of the PC805, the company claims the industry’s first SoC optimized for 5G small-cell O-RUs. The small-footprint, low-power 5G O-RU SoC is designed to ease the implementation of 5G NR/LTE small-cell O-RUs for use cases such as enterprise, industrial, neutral host and private networks.
In addition to 5G/RF technologies, the Silicon 100 addresses a range of emerging technologies and the key questions being asked in the industry today around topics such as generative AI, hybrid quantum computing, chiplets and RISC-V. It also looks at where entrepreneurship and innovation are happening globally.
This year’s report reveals four key technology trends: higher levels of activity around AI, a continued rise in quantum computing endeavors, a leap upward in chiplet interest and a decrease in non-processor activity. Another trend is a reduced Chinese presence on the list and an increase in European companies.
Perhaps not surprisingly, 15 companies are involved with AI in the data center, and the number of startups targeting AI at the edge inched up from eight in 2023 to 11 in 2024. Three categories cover general AI processing: photonic acceleration, GPU-to-data-center-AI and edge AI. In aggregate, across these categories, the number of startups has increased from 27 last year to 32 in this year’s Silicon 100, with edge AI startups increasing as a proportion of those numbers.