Tiny Chips, Big Myths: How Military Systems Actually Use Semiconductors
The Geopolitics of Chips: Why the Military Isn’t Chasing Consumer Tech
Is the “chip war” really about national security and military equipment or is it about consumer goods?
TL;DR:
The “chip war” isn’t primarily about military systems relying on cutting-edge, sub-5nm chips. Most military applications use larger, proven node sizes (45nm–250nm) for durability and reliability. While smaller chips are emerging in niche areas like AI-driven systems, the race for advanced semiconductors is more about future defense use and current geopolitical control and economic influence than immediate defense needs.
Introduction:
In the escalating competition for semiconductor supremacy, the “chip war” is often framed as a matter of national security, with arguments highlighting the dependency of military systems on cutting-edge, low-node chips. However, the reality of semiconductor usage in military applications is far more nuanced. While consumer electronics increasingly rely on chips as small as 3nm, military technologies predominantly utilize larger nodes for reasons such as durability, reliability, and resistance to radiation in extreme environments.
This analysis aims to demystify the relationship between semiconductor technology and modern military systems, exploring the actual node sizes employed in key applications and examining whether the race for ever-smaller chips is truly a matter of defense necessity—or if it serves broader geopolitical and economic objectives. By addressing the intersection of chip sizes, military technology, and national strategy, this study sheds light on the myths and realities surrounding semiconductor use in defense systems.
Introduction:
One of the confounding arguments that I frequently hear about the need to “win” the chip war is that it is necessary for national intelligence because all of our weapon systems depend on low node (now nanometer) chips. I would like to do a little myth busting on that notion. First we have to start with definitions perhaps. I consider advanced chips to be anything under 5nm because lower than that you need EUV or NAEUV lithography and technology and cannot depend on older DUV and "self-aligned quadruple patterning (SAQP)" technology while accepting very low yield. Just for comparison’s purposes consider that an iPhone 16 uses a 3nm chip.
While semiconductor chips are pivotal in modern military technology, influencing both capability and strategy across different nations, you might be surprised to learn that these systems do not depend on these cutting edge super low nanometer size nodes. This study aims to elucidate the relationship between chip technology and military strength by examining specific applications and their geopolitical ramifications.
Semiconductor Chip Node Sizes in Modern Military Equipment:
The size of a semiconductor node, measured in nanometers (nm), indicates the level of miniaturization and thus the performance of the chip. In modern military applications. Smaller node chips, particularly those below 28nm, are crucial for enhanced performance, advanced capabilities, miniaturization, and to provide a strategic advantage. However as we will see a lot of the well know systems do not use chips below 28 and certainly not down to the 3nm level like an iPhone 16.
US Military Systems and Node Sizes
The semiconductor node size used in most U.S. military systems is typically larger than those found in cutting-edge commercial consumer electronics, for several reasons. Larger node sizes are less susceptible to single-event upsets (SEUs) and other radiation-induced errors. Military systems often have design cycles lasting decades, so they use technologies proven over time. Older nodes ensure a steady supply chain as these fabs remain operational longer than cutting-edge semiconductor facilities.
Many military systems rely on semiconductor nodes between 90nm and 45nm or above, as these technologies are mature and well-understood. Larger nodes are often used because they are more resistant to radiation (important for space and high-altitude applications) and provide increased reliability over time. Legacy avionics in fighter jets like older F-16 variants or early F-35 systems typically use around 90nm chips. Upgraded mission computers in platforms like the F/A-18 and F-22 can use 65nm node chips. Computational systems in upgraded fighter jets (e.g., newer F-16 variants or early F-35 software upgrades) can use down to 45nm chips. Chips in secure communication devices for tactical operations may be down to 45nm node chips.
Chips used in aerospace and military systems often require radiation-hardened (Rad-Hard) capabilities, which are typically manufactured on larger nodes like 130nm or even 250nm. These nodes prioritize durability and reliability under extreme conditions, such as high radiation environments. Radiation-hardened electronics for spacecraft and ICBM guidance systems are eve higher nm nodes than 90nm. Satellites such as the US GPS III system use 250nm node chips.
Modern military systems may utilize smaller nodes (e.g., 28nm or 14nm) for specific applications requiring higher computational power or energy efficiency, such as advanced radar processing, AI applications, or next-generation avionics. NXP's 28nm RFCMOS Radar One-Chip Family NXP Semiconductors has developed the SAF85xx series, a 28nm RFCMOS radar system-on-chip (SoC) designed for short, medium, and long-range radar applications. These chips enable high-performance 4D sensing and are designed to operate from 76 to 81 GHz, covering the full automotive radar frequency band. Xilinx's Kintex UltraScale family, built on a 20nm process node, offers high-performance, low-power programmable solutions. These FPGAs are utilized in military avionics for applications requiring significant processing capabilities and flexibility.
There are exceptions for cutting-edge technologies in niche areas. In advanced systems like 5th and 6th-generation fighter jets, missile defense systems, and AI-based tools, newer nodes like 14nm, 10nm, or even smaller may be used for specific subsystems requiring significant processing power. The U.S. military works closely with trusted foundries like GlobalFoundries, Intel, and others under programs such as the Trusted Foundry Program to ensure secure and reliable chip manufacturing. These foundries often cater to specialized military-grade requirements.
AI Based Military Systems
The integration of artificial intelligence (AI) into military operations is becoming increasingly significant, with semiconductor technology playing a critical role in enhancing AI performance. Chips smaller than 5nm offer unprecedented computational power, which is vital for sophisticated AI applications in defense. Smaller node chips enable faster processing for AI algorithms, critical for real-time decision-making and autonomous operations. These sub 5 nm chips allow AI systems to operate with less power, vital for field-deployable units or long-endurance missions. Smaller chips can lead to more compact AI hardware, which is advantageous for integration into various military platforms. The ability to deploy AI systems with such advanced chips could lead to superiority in information warfare, autonomous systems, and cyber operations.
(Pictured above is Nvidia's next-generation graphics processor for artificial intelligence, called Blackwell, will cost between $30,000 and $40,000 per unit. The GB200 combines two Blackwell GPUs and one Grace CPU. It's part of the NVIDIA GB200 NVL72, which is an exascale computer in a single rack.)
Although not explicitly confirmed to use sub-5nm chips, the US's AI initiatives, including Project Maven for enhancing drone operations with computer vision, would benefit significantly from such technology. The project's continued development suggests a possible future integration of chips below 5nm to handle the vast amount of data for real-time image analysis. The US has been investing in R&D for chips that might go beyond current nodes, particularly for AI applications in defense. Companies like NVIDIA, with their focus on AI, and partnerships with the Department of Defense, might be exploring or already implementing sub-5nm chips in experimental or classified projects. Future iterations of autonomous vehicles, drones, and weapon systems could leverage sub-5nm chips for enhanced AI capabilities, although current deployments are likely using larger nodes for operational systems.
Lower Nodes as Future Proofing
Lower node chips and AI can significantly enhance Intelligence, Surveillance, and Reconnaissance (ISR) operations in the military and national security sectors in the many ways. Lower node chips (like 5nm and below) provide higher computational speeds and efficiency, allowing for real-time analysis of vast amounts of data collected from ISR assets. This rapid processing is crucial for timely decision-making in dynamic battle environments.These lower node chips consume less power, which is vital for extending the operational life of unmanned systems or for mobile command centers where energy resources are limited.
AI algorithms running on advanced chips can identify patterns, anomalies, or threats in imagery or signals intelligence (SIGINT) faster than human analysts. For instance, AI can detect changes in terrain or infrastructure that might indicate enemy activity. AI can manage and interpret feeds from multiple sensors (e.g., EO/IR, SAR, acoustic) simultaneously, providing a comprehensive situational awareness that would be overwhelming for human operators alone. AI, supported by high-performance computing chips, can fuse data from various sources (satellites, drones, ground sensors) to create a more accurate and actionable picture of the operational environment. This fusion can lead to better target identification and tracking. Using AI to analyze historical and real-time data can predict potential adversary movements or intentions, enhancing proactive ISR strategies.
Lower node chips enable more sophisticated AI algorithms to be embedded in unmanned aerial vehicles (UAVs), allowing for autonomous or semi-autonomous flight paths, target acquisition, and threat response, thus reducing human operator workload. With AI processing at the edge (on the device rather than relying solely on cloud computing), decisions can be made faster in environments where communication back to a central command might be compromised or delayed.
Smaller node chips with AI can analyze and react to complex electronic signatures in real-time, aiding in jamming, spoofing, or countering enemy electronic systems. AI on these chips can enhance cyber ISR by quickly identifying patterns in network traffic or data breaches, offering real-time insights into cyber threats or breaches. The computational power of lower node chips allows for more complex encryption algorithms, ensuring secure communication channels for ISR data transmission. AI can sift through intercepted communications to identify key information or detect encrypted messages, enhancing SIGINT capabilities.
Smaller chips mean that sophisticated ISR systems can be deployed in smaller, more discreet packages, useful for covert operations or in situations where size, weight, and power (SWaP) are critical. AI and low-node chips can power advanced wearable devices for soldiers, providing them with real-time ISR insights directly to their field of vision or personal communication devices. AI can generate complex scenarios for training based on real ISR data, improving the readiness of personnel by simulating various tactical environments with high fidelity.
Conclusion:
The semiconductor “chip war” is often portrayed as a race critical to national security, with a focus on developing ever-smaller node chips. However, a closer analysis reveals that military applications are not uniformly dependent on the cutting-edge technology dominating consumer electronics. Instead, military systems often prioritize reliability, radiation resistance, and proven performance over the sheer computational power offered by chips below 5nm. Larger node chips—ranging from 45nm to over 250nm—remain critical for many defense systems, including avionics, secure communications, and radiation-hardened aerospace technologies.
While there are niche applications where smaller node chips could enhance military capabilities, particularly in AI-driven systems and advanced ISR operations, these are not yet the dominant use cases. The true importance of the chip war lies less in the immediate defense applications of sub-5nm nodes and more in the geopolitical implications of semiconductor supply chain control. As nations vie for technological dominance, the debate over military versus consumer priorities underscores the broader strategic stakes of semiconductor innovation and manufacturing.
In the end, the chip war is as much about global influence and economic leverage as it is about military readiness. Understanding the nuanced role of semiconductor technology in defense applications helps dispel myths and refocus the conversation on the strategic realities driving this competition.
Sources:
https://www.cnbc.com/2023/04/17/us-components-found-in-russian-iranian-military-tech.html
https://www.asiafinancial.com/china-providing-90-of-chips-used-in-russia-despite-sanctions
https://carnegieendowment.org/politika/90250
https://www.nytimes.com/2023/12/29/opinion/chips-semiconductor-china-russia-military.html
https://www.rferl.org/amp/us-semiconductor-firms-russia-imports-evading-sanctions/32840928.html
https://foreignpolicy.com/2024/02/22/russia-sanctions-weapons-ukraine-war-military-semiconductors/
https://www.eetimes.com/experts-u-s-military-chip-supply-is-dangerously-low/
https://www.csis.org/analysis/semiconductors-and-national-defense-what-are-stakes
https://www.hpcwire.com/2020/06/01/10nm-7nm-5nm-should-the-chip-nanometer-metric-be-replaced/
https://breakingdefense.com/2021/06/dod-seeks-2-3b-to-bolster-chip-making/
https://asiatimes.com/2022/08/missing-military-pieces-in-the-chips-act/
https://www.howtogeek.com/755895/what-is-a-5nm-chip-and-why-is-5nm-so-important/
https://asiatimes.com/2024/04/china-to-make-5nm-chips-with-saqp-process/
https://www.forbes.com/sites/linleygwennap/2020/10/12/apple-moores-law-is-running-out/
https://citylabs.net/military-semiconductor-applications/
https://semiengineering.com/5nm-vs-3nm/
https://interestingengineering.com/military/russia-china-team-military-use-of-ai
https://thediplomat.com/2024/04/ai-chips-for-china-face-additional-us-restrictions/
https://www.semianalysis.com/p/china-ai-and-semiconductors-rise
https://www.csis.org/analysis/blocking-chinas-access-ai-chips-matters-us-national-security






