Hero
A spintronic coprocessor
architecture using 3D hopfions.
HELIOS-3D is a staged research hypothesis for a low-energy inference coprocessor, with 2026 evidence for the underlying spin textures and a clear demonstrator path.
The thermodynamic crisis is the motivation
IEA projects AI infrastructure will consume ~600 TWh in 2026, with >4 billion m³ of water withdrawal by 2027. Embodied carbon is now 50% of AI hardware lifecycle footprint.
AI infrastructure electricity demand in 2026
Annual AI water withdrawal by 2027
Evidence structure
Highlights across the four research stages. Filter the full claims table below.
Core Architecture
MCA + BRC dual-core, hopfion-based
Candidate Materials
EuS / Bi₂Se₃ / EuS, FGT, Mn₃Sn
Performance Floor
40–60 ps write, pJ/µm² energy floor
Chiral Sensing
CISS, TOHE, peptide-crystal interface
| Claim | Tag | Source |
|---|---|---|
| Hopfion annihilation > 50kBT in EuS/Bi₂Se₃/EuS | INFERRED | Literature Review |
| Planar-first, electrical read is the minimum credible path | INFERRED | Alternative Materials |
| DISH enables sub-second 3D scaffolding with 19µm resolution | PROPOSED | Literature Review |
| Sub-Landauer computation is achievable | SPECULATIVE | Abstract |
| 40-ps write at 1.7 pJ/µm² in Mn₃Sn | DEMONSTRATED | Literature Review |
| TOHE is unambiguous 3D-hopfion electrical hallmark | INFERRED | Literature Review |
2026 evidence base
The literature that motivates and constrains the architecture.
Literature timeline 2024–2026
Where to start
A 5-document reading path. Read in order; each step builds on the previous.
- 01 AbstractEstablished
Project pitch and dual-core architecture.
- 02 Candidate MaterialsEstablished
Spin textures and the EuS/Bi₂Se₃/EuS substrate.
- 03 Literature ReviewEstablished
The 2026 evidence base.
- 04 Claims MatrixCurrent
What is and is not demonstrated.
- 05 Pitch Deck OutlineSpeculative
The full project narrative.