Radiation-Hardened AI Portfolio
SYNAPLEX is a polymer substrate for neuromorphic computing that enables AI in radiation-intensive environments. ISS-validated (Experiment ID 8071) with greater than 500 kGy tolerance where silicon fails at less than 100 kGy.
Defence applications represent 70% of SYNAPLEX revenue projection, driven by satellite mega-constellations requiring edge AI and tactical systems operating in contested electromagnetic environments.
Our ITAR-free UK origin enables direct export to allied nations without State Department involvement, providing significant time and cost advantages for multinational programmes.
Why Silicon Fails in Space
Van Allen Belt radiation degrades conventional silicon-based neuromorphic chips
Radiation Environment
Van Allen radiation belts expose LEO satellites to 10-100 kGy total ionizing dose over 5-year mission life. Silicon neuromorphic chips experience threshold voltage shifts and increased leakage current, degrading synaptic weight precision.
50,000+ LEO satellites planned by 2030 (Starlink, OneWeb, Kuiper) require edge AI for autonomous collision avoidance and payload processing without ground station latency.
Silicon Limitations
Intel Loihi, IBM NorthPole, and BrainChip Akida all use CMOS silicon substrates that fail at less than 100 kGy. Radiation-hardened variants reduce performance by 3-5 generations and increase cost 10-15x.
SYNAPLEX polymer substrate maintains integrity beyond 500 kGy without performance degradation, enabling commercial-off-the-shelf neuromorphic integration for space applications.
Space Autonomy
Edge AI substrate for autonomous satellite operations, collision avoidance, and payload processing in mega-constellations.
- Starlink Gen2 — 30,000 satellites
- OneWeb Phase 2 — 6,372 satellites
- Kuiper — 3,236 satellites
- Military Constellations — SDA Layer 1
Tactical Counter-UAS
Edge AI for drone swarm engagement in EMP-contested environments with less than 10 fJ per synaptic event power efficiency.
- Swarm Classification — Real-time threat ID
- Engagement Prioritization
- EMP-Hardened Operation
- Network-Degraded Mode
Missile Guidance
Active seeker processing for imaging infrared and millimeter-wave guidance in terminal phase engagement.
- IIR Seeker Processing
- MMW Target Classification
- Countermeasure Rejection
Nuclear Monitoring
IAEA safeguards and power plant monitoring with sustained operation in high-radiation environments.
- IAEA Verification
- Reactor Monitoring
- Waste Processing
Satellite Processing
On-orbit Earth observation and SAR processing with edge AI for immediate ground delivery.
- EO Classification
- SAR Target Detection
- Hyperspectral Analysis
Battlefield AI
Distributed sensor networks for tactical situational awareness in network-degraded environments.
- Acoustic Classification
- EO/IR Tracking
- Edge Fusion
Competitive Comparison
Silicon neuromorphic chips fail at less than 100 kGy radiation exposure
| Platform | Type | Rad Tolerance | Power/Event | Space-Ready |
|---|---|---|---|---|
| SYNAPLEX | Polymer Substrate | >500 kGy | <10 fJ | Yes |
| Intel Loihi 2 | Silicon Chip | <100 kGy | 23 fJ | No |
| IBM NorthPole | Silicon Chip | <100 kGy | ~15 fJ | No |
| BrainChip Akida | Silicon Chip | <100 kGy | ~20 fJ | No |
Discuss Defence Integration
Contact our defence programmes team to explore SYNAPLEX integration for radiation-hardened AI applications.
Contact Defence Team Back to SYNAPLEX Overview