Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/threads/defence-electronics-more-important-than-videocard-chips.24777/
        )

    [addOns] => Array
        (
            [DL6/MLTP] => 13
            [Hampel/TimeZoneDebug] => 1000070
            [SV/ChangePostDate] => 2010200
            [SemiWiki/Newsletter] => 1000010
            [SemiWiki/WPMenu] => 1000010
            [SemiWiki/XPressExtend] => 1000010
            [ThemeHouse/XLink] => 1000970
            [ThemeHouse/XPress] => 1010570
            [XF] => 2030970
            [XFI] => 1060170
        )

    [wordpress] => /var/www/html
)

Defence electronics more important than videocard chips

Paul2

Well-known member
1773845296049.png


As electronics is increasingly more talked within the military action context, I feel I have to write that PRC defence industry is far more starved off components for which they cannot procure a consumer market analogue than ones for which they can.

Any generic smartphone SoC has enough processing power nowadays, sufficient for any imaginable guidance task, including computational radar imaging – something that once required a supercomputer. An MCU will compute you a missile trajectory from any point on the earth to another, and a recent smartphone SoC will be able to do a realtime imaging radar with every computational enhancement trick.

However, PRC cannot build the said imaging radars without RF and analog chips. Shahed drones cannot get guidance without high grade RF PCBs, and MEMS sensors. TV guided drones cannot see anything without specialist image sensors. All off that stuff is being relatively low volume, and has to be 99% imported by PRC, Russia, NK, Iran.

PRC can make an average SoC you put into a TV, but not 100 Ampere VHF/UHF transistors for radars, and military radios.

RF foundries, specialist RF materials/components, niche analog nodes, advanced MEMS, non-consumer grade imaging sensors – almost all is still exclusively Western.
 
As electronics is increasingly more talked within the military action context, I feel I have to write that PRC defence industry is far more starved off components for which they cannot procure a consumer market analogue than ones for which they can.

Good question. AI is going to change that I think. There will be less central computing and more edge computing with AI which means leading edge silicon. If you look at what Tesla did for automotive, the US military will change the world and with all of the wars everyone will see it. Defense startup funding doubled to $18B+ in 2025.

Look at some of the new defense electronic companies. There are dozens of these companies. I saw this coming and wondered where the demand will come from but now with what is happening around the world.

New & Emerging Defense Electronics Companies​

United States (strongest growth)​

1. Deca Defense (founded 2024)​

  • Focus: AI + sensor fusion + embedded systems
  • Builds deployable software tightly integrated with defense electronics hardware
  • Use cases: ISR (intelligence, surveillance, reconnaissance), autonomous targeting

2. Aurum Systems (founded 2024)​

  • Focus: Edge computing + real-time analytics
  • Electronics angle: onboard processing for denied environments (no connectivity)
  • Works on ISR platforms and mission systems

3. Epirus​

  • Focus: Directed-energy systems (microwave weapons)
  • Key tech: high-power electronics to disable drones (counter-UAS)
  • Example: “Leonidas” system uses EM pulses to fry electronics

4. Modern Intelligence​

  • Focus: sensor data processing + maritime tracking
  • Electronics: integrates with radar/EO/IR systems for target detection

5. Kapta Space (startup, ~2025)​

  • Focus: electronically steerable radar (phased-array-like systems)
  • Electronics-heavy: advanced antenna + RF beam steering for satellites
  • Enables faster imaging and multi-mode radar sensing

Europe (rapidly growing defense electronics ecosystem)​

6. STARK (founded 2024, Germany)​

  • Focus: unmanned systems + loitering munitions
  • Electronics: embedded avionics, guidance, control, and mission software
  • Products include drones + mission control systems

7. Avalor AI (Netherlands)​

  • Focus: AI + sensor intelligence for defense
  • Likely applied to surveillance and battlefield awareness

8. Comand AI (France)​

  • Focus: AI-driven command & control systems
  • Integrates multiple electronic systems (sensors, communications)

Autonomous / Electronic Systems & Drones​

9. Wild West Systems (founded 2025)​

  • Focus: autonomous drones + modular hardware
  • Electronics: onboard autonomy, sensing, and control systems

10. Shield AI (scale-up but still “new gen”)​

  • Focus: AI pilots + drone autonomy
  • Electronics-heavy stack: sensors + onboard compute + autonomy software
  • Expanding via acquisitions of sensor companies
 
However, PRC cannot build the said imaging radars without RF and analog chips. Shahed drones cannot get guidance without high grade RF PCBs, and MEMS sensors. TV guided drones cannot see anything without specialist image sensors. All off that stuff is being relatively low volume, and has to be 99% imported by PRC, Russia, NK, Iran.

PRC can make an average SoC you put into a TV, but not 100 Ampere VHF/UHF transistors for radars, and military radios.

RF foundries, specialist RF materials/components, niche analog nodes, advanced MEMS, non-consumer grade imaging sensors – almost all is still exclusively Western.

Is Russia totally out of the game on producing some of this (radars/military radio)? I know they're terribly behind on IC lithography, but the Soviet Union had to have a massive industry around some of these giant analog components that probably haven't changed that much in recent decades?

I also imagine there has to be an 'unequal' response to this -- it's possible to get the effect of radar with swarms of small electronics these days. There's a few good examples of Youtube and elsewhere. Perhaps the small electronics / shorter range radios can form the foundation of a new architecture of high redundancy (due to mass # of devices) for on-ground military intelligence and communications.
 
Anyone know the training material used to green light the current adventure?

Which AI provider was used to war game the various scenarios?
 
Anyone know the training material used to green light the current adventure?

Which AI provider was used to war game the various scenarios?

Gemini writes:

The current conflict involving the U.S., Israel, and Iran (escalating significantly in late February and March 2026) has been widely characterized as a turning point for AI in military operations. Palantir has indeed played a central role, but it is part of a broader "AI stack" involving several major providers.

According to reports on Operation Epic Fury (the U.S.-led campaign), the AI operations are structured as follows:

Key AI Providers and Their Roles​

  • Palantir (Data Backbone): Palantir provides the Maven Smart System (MSS) and the AIP (Artificial Intelligence Platform). Its primary role is "data fusion"—aggregating trillions of data points from satellite imagery, drone feeds, radar signals, and intercepted communications into a single "battlefield ontology" or digital twin.
  • Anthropic (Strategic Analysis): Anthropic’s Claude models have been integrated into Palantir’s environment. While Palantir gathers the data, Claude acts as the "brain," analyzing text-based intelligence and pattern matching to identify high-value targets, such as Revolutionary Guard (IRGC) command centers.
  • OpenAI (Intelligence & Simulations): In early 2026, OpenAI reportedly secured a U.S. Department of Defense contract to deploy GPT series models on classified networks for war-gaming, translation, and intelligence analysis, filling a gap after the Pentagon briefly labeled other providers as "supply chain risks."
  • Microsoft & Google (Infrastructure): These companies provide the essential cloud and GPU infrastructure (such as Google’s Project Nimbus for the IDF) that allows these AI models to run in real-time during active combat.

Notable Developments in 2026​

  • Assassination of Leaders: There is significant speculation regarding the use of AI targeting in the strikes that killed senior Iranian figures, including Ali Larijani and Gholamreza Soleimani on March 17, 2026.
  • Retaliation Against Tech: Highlighting the strategic importance of these providers, Iran’s IRGC claimed responsibility for drone strikes against Amazon (AWS) data centers in the UAE and Bahrain in early March, specifically citing their role in hosting military AI systems.
  • Human-in-the-Loop: Official reports emphasize that while AI "leads" the identification and simulation phases, human commanders still provide the final approval for kinetic strikes.
The conflict has effectively become a real-world "stress test" for these platforms, with Palantir’s CEO, Alex Karp, recently stating that these lethal AI capabilities provide the U.S. and its allies with a "decisive advantage."
 
I suppose during the last 2-3 years, with the ever expanding drone war and expanding "kill zones", Palantir has been using the Ukraine war also as a real world R&D environment:

Gemini:

As of early 2026, the consensus among military analysts and Ukrainian officials is that Palantir has been a significant "force multiplier" in the conflict. While it is difficult to quantify exactly how much of the war’s success is attributable to one company, Palantir’s CEO Alex Karp has claimed the software is "responsible for most of the targeting in Ukraine."

Here is how the AI and data software are being utilized:

1.​

Palantir’s Gotham and MetaConstellation platforms are the backbone of Ukraine’s targeting system.

  • Data Fusion: The software ingests massive amounts of data—satellite imagery, drone feeds, intercepted radio communications, and even citizen tips from the "eVorog" (e-Enemy) chatbot.
  • Speed: It fuses this data into a single map, allowing commanders to see Russian movements in near real-time. This has compressed the "kill chain" (the time between spotting a target and striking it) from hours or days to just minutes.

2.​

In early 2026, Ukraine and Palantir launched a project called the "Dataroom."

  • AI Training: Ukraine is feeding millions of hours of real battlefield footage (including data on Russian Shahed drones) into a secure environment to train AI models.
  • The Goal: To create highly autonomous interceptor drones that can identify and take down enemy threats without needing a constant human link, which is crucial in environments where Russian electronic warfare (jamming) is intense.

3. Humanitarian and Legal Efforts​

The partnership extends beyond active combat:

  • Demining: Palantir’s AI-enabled software helps the Ministry of Economy prioritize which of the 156,000 square kilometers of contaminated land to demine first, based on proximity to hospitals, schools, and critical infrastructure.
  • War Crimes: The Prosecutor General uses the platform to organize a deluge of digital evidence—timestamped photos, satellite passes, and social media posts—to build legal cases for future international courts.



The evolution of the "kill zone" in Ukraine reflects a fundamental shift in modern warfare, where the integration of drones and advanced data analytics has made the battlefield increasingly transparent and lethal.
A "kill zone" in this context refers to a stretch of territory along the line of contact where any movement—whether by personnel or vehicles—is detected and engaged within minutes.

1. The Expanding Kill Zone

Initially, the "kill zone" was limited by the range of early FPV (First Person View) drones and the speed of human decision-making. Over the course of 2024 and 2025, several factors have significantly expanded its depth and lethality:
  • Depth of Reach: According to reports from the International Centre for Defence and Security (ICDS) and CSIS, the kill zone has deepened from an initial 5–10 km to approximately 35–40 km as of 2025–2026.
  • Tactical "Saturation": The sheer volume of drones—estimated at upwards of 10,000 per day—has created a "saturated space" where conventional fortifications (like large trenches or concentrated depots) are no longer viable and have been replaced by dispersed, small-unit positions.
  • The "Drone Line" Initiative: In early 2025, Ukraine launched a dedicated initiative to create a specialized 10–15 km deep "kill zone" where autonomous drones eliminate all enemy activity without requiring constant human pilot input.

2. The Role of Palantir

Palantir Technologies acts as the "connective tissue" or the digital kill chain that enables these expanded zones to function at scale. Their involvement has moved from high-level visualization to deep technical integration:
[table]
[TR]
[td]System/Project[/td][td]Function in the Kill Zone[/td]
[/TR]
[TR]
[td]MetaConstellation[/td][td]Aggregates data from commercial satellites, thermal sensors, and drone feeds to provide a "God's-eye view" of enemy movements in real time.[/td]
[/TR]
[TR]
[td]Brave1 Dataroom[/td][td]A 2026 initiative where Palantir provides the software infrastructure to train AI models using real-world combat data. This allows drones to "learn" how to identify and intercept targets (like Shahed drones) autonomously.[/td]
[/TR]
[TR]
[td]AIP (AI Platform)[/td][td]Compresses the time between detection and strike. By automating the identification of targets (e.g., distinguishing a tank from a decoy), it reduces the decision cycle from tens of minutes to seconds.[/td]
[/TR]
[TR]
[td]Targeting Precision[/td][td]Palantir’s tools allow commanders to prioritize targets based on available resources, essentially "suggesting" which drone or artillery unit is best positioned to neutralize a threat.[/td]
[/TR]
[/table]


3. Key Technical References

For deeper research into these developments, the following sources provide the most explicit details on the intersection of AI and the drone-enforced kill zone:
  • CSIS (2025): "Ukraine’s Future Vision and Current Capabilities for Waging AI-Enabled Autonomous Warfare" — Details the shift toward multisensor fusion and the goal of 70-80% strike success rates via AI.
  • Hudson Institute (2025): "The Impact of Drones on the Battlefield" — Discusses the "Tactical Warfare Revolution" and the transparency of the battlefield to a depth of 20km+.
  • Defense News / The Register (Jan 2026): Reports on the Palantir-Ukraine "Dataroom" agreement, which formalized Palantir’s role in training autonomous interceptor drones.
Note: Palantir CEO Alex Karp has described this integration as a "digital kill chain," a term that highlights how software now dictates the speed and efficiency of lethal operations on the ground.


Reference:
https://odessa-journal.com/the-mini...nvironment-for-training-military-ai-solutions
 
Back
Top