How Next-Generation Chips Are Powering Future Artificial Intelligence and Space Missions
Munich, Thursday, 1 October 2026.
With over 500,000 chips already operating in space, advanced gallium nitride semiconductors are rapidly replacing traditional silicon to power massive modern artificial intelligence datacenters and autonomous robotics.
EPC Announces Electronica 2026 Showcase
Efficient Power Conversion Corporation, Inc. (EPC) has confirmed plans to exhibit its latest Gallium Nitride (GaN) power conversion solutions at Electronica 2026 in Munich, Germany [1][2]. The event is scheduled to take place from November 10 to November 13, 2026, positioning it as a future event relative to the current date of October 1, 2026 [1][2]. EPC will be located at Messe München in Hall C6, Booth 221, where it intends to demonstrate hardware scaling for high-density artificial intelligence datacenters and autonomous vehicles [1][2]. This showcase underscores the company’s strategy to replace traditional silicon semiconductors with GaN technology to achieve higher efficiency and power density [1].
Next-Generation Product Portfolio
During the exhibition, EPC will demonstrate specific power solutions including the EPC91123 6 kW 800 V-to-12.5 V ISOP LLC converter [1][2]. The company is also highlighting a new 6 kW 800 V-to-6 V converter designed for next-generation AI processors [2]. For robotics and drone propulsion, the showcase includes the EPC33110, a 100 V three-phase integrated ePower™ Stage IC, alongside the preview-stage EPC33111 IC [1][2]. Additionally, mass-produced 100 V integrated ePower™ Stage IC families such as the EPC23108/09 and EPC23110/11 will be featured to support current demands in motion control [2].
Strategic Market Positioning in AI and Space
The demand for GaN semiconductors is driven by the need for improved power distribution in AI datacenters and space missions [1][4]. EPC Space estimates it has captured more than 35 percent of rad-hard power FET sockets, indicating significant market penetration in specialized sectors [2]. Over 500,000 devices are currently in orbit across LEO, MEO, and GEO configurations, validating the technology’s reliability in extreme environments [2]. This infrastructure supports the broader 800 VDC backbone architecture increasingly adopted for AI server power supply units [4].
Radiation Hardening and Reliability
Technical discussions will address the stability of spaceborne eGaN HEMT-based LDO regulators under Total Ionizing Dose (TID) environments [3]. Analytical models demonstrate how these regulators stabilize control loops and eliminate radiation-induced drift, which is critical for long-duration missions [3]. EPC Space offers radiation-hardened ICs like the EPC7011 and EPC7009 to meet these rigorous standards [2]. The long-term reliability of GaN devices under AI data center mission profiles is also cited as critical to widespread adoption [2].
Technical Workshops Schedule
EPC has scheduled five technical sessions at their booth to educate attendees on specific applications and reliability [1]. On November 10, 2026, at 2:00 p.m., CEO Alex Lidow will present “Where is GaN Going?” to discuss the technology’s evolution over the last 17 years [1][2]. On November 11, 2026, sessions include “Radiation-Hardened GaN: From Niche to Mainstream” at 11:00 a.m. and “New 100 V Half-Bridge and Three-Phase GaN ICs for Robots and Drones” at 2:00 p.m. [1][2]. Further sessions on November 12, 2026, will cover high-voltage applications and mission profile reliability for AI data centers [1][2].
Executive Insights on Integration
Alex Lidow, CEO of EPC, stated that GaN is enabling new levels of system integration across power, motion, and computing architectures [1]. He noted that from 800 V power distribution for AI to high-performance motor drives, the technology is applying across next-generation electronic systems [1]. Bel Lazar, CEO of EPC Space, highlighted that nearly every new power-management design is either using or evaluating their rad-hard GaN FETs, ICs, or modules [2]. Michael De Rooij, GaN Applications Fellow at EPC, emphasized that power supply demands require innovative solutions for small size and high efficiency [2].
Infrastructure Trends
The technical framework for AI data center power distribution is being reconstructed across five segments to optimize transmission and voltage reduction [4]. Key industry collaborations driving the 800VDC architecture shift include NVIDIA and GE Vernova’s “AI Factory” design and ABB and NVIDIA’s 800VDC data center architecture [4]. Power conversion solutions involve multiple semiconductor players including Infineon, Texas Instruments, and ST, alongside EPC’s GaN multi-level totem-pole PFC [4]. This evolution shifts AI server power supply units from traditional configurations toward 400V/800V high-voltage bus generation [4].
Industry Collaboration and Future Path
Analysts plan to conduct comprehensive architectural interpretations by analyzing typical AI power supply architecture schemes from leading companies [4]. The goal is to map the full chain from the grid entrance all the way to GPU core power supply [4]. Case 拆解 will target companies like NVIDIA, Delta, Vertiv, Infineon, TI, ADI, Navitas, Empower, and Renesas to analyze voltage conversion and system integration technology [4]. This comprehensive view supports the understanding of the Grid-to-Core energy path required for modern AI infrastructure [4].