Patent Protection Of Photonic Computing ChIPs.
1. What Are Photonic Computing Chips?
Photonic computing chips (also called optical processors) use light instead of electrons to perform computations. Key characteristics:
- Optical components: waveguides, modulators, lasers
- Integration with electronics: hybrid silicon-photonics chips
- Applications: AI acceleration, high-speed data processing, optical communication
Patenting these systems involves protecting hardware designs, integrated circuits, and photonic architectures, as well as software/algorithms specifically enabling optical computation.
2. Patentability Requirements
For photonic computing chips, patent offices typically require:
- Novelty – The chip design or method must be new worldwide.
- Inventive Step (Non-Obviousness) – It cannot be obvious to a skilled person in photonics/electronics.
- Industrial Applicability / Utility – The invention must be practically implementable.
- Technical Character – Pure algorithms are not patentable unless tied to physical hardware or technical effect.
Key challenges:
- Software or AI for optical computation may be rejected as “computer programs as such” unless integrated into hardware.
- Hybrid photonic-electronic designs often strengthen patent claims because they have clear technical effect.
3. Relevant Case Laws
Here are more than five landmark cases that shape how photonic computing chips can be patented:
1. Diamond v. Diehr
Facts:
A rubber-curing process used a computer algorithm to calculate curing time.
Judgment:
- Algorithms alone are not patentable.
- A process using an algorithm with a technical effect (curing rubber) is patentable.
Relevance:
- AI or algorithms controlling optical chip behavior (modulation, routing, photonic logic) → patentable.
- Key principle: software + hardware integration creates patent eligibility.
2. Alice Corp. v. CLS Bank International
Facts:
Patent for computerized financial transactions.
Judgment:
- Abstract ideas implemented on a computer → not patentable.
- Must include “inventive concept” beyond abstract idea.
Relevance:
- Algorithms for photonic computation alone → may be abstract.
- Must demonstrate specific hardware integration or optical process.
3. Enfish, LLC v. Microsoft Corp.
Facts:
Software for a self-referential database.
Judgment:
- Software can be patentable if it improves computer functionality or technical process.
Relevance:
- Photonic chip architectures improving computation speed or power efficiency → technical effect.
- Example: a waveguide network increasing processing throughput is patentable.
4. State Street Bank v. Signature Financial Group
Facts:
Software producing “useful, concrete, and tangible result”.
Judgment:
- Tangible results make software patentable.
Relevance:
- Photonic computing chips producing measurable computation improvements (e.g., terabit data throughput) → strengthens patentability.
5. KSR International Co. v. Teleflex Inc.
Facts:
Combination of known mechanical components (pedals + sensor).
Judgment:
- Obvious combinations → not patentable.
Relevance:
- Photonic chips combining existing optical modulators and waveguides → must show non-obvious improvement (e.g., reduced latency or energy consumption).
- Demonstrating unexpected technical effect is key.
6. T 1173/97 (Computer program product / IBM)
Facts:
Software controlling a technical system.
Judgment:
- Software with technical effect on hardware → patentable.
- Pure business logic → not patentable.
Relevance:
- Algorithms controlling optical switching or photon routing in chips → patentable under EPC.
7. Microsoft Corp. v. i4i Limited Partnership
Judgment:
- Patent invalidity must be proven with clear and convincing evidence.
Relevance:
- Once photonic computing chip patents are granted → enforceable and strong against infringement.
8. Apple Inc. v. Samsung Electronics Co.
Facts:
Patent dispute over smartphone hardware/software integration.
Relevance:
- Demonstrates how integration of hardware + algorithm + physical effect strengthens enforceability.
- Applies to photonic chip architectures with unique optical layouts or hybrid designs.
4. Application to Photonic Computing Chips
To successfully patent a photonic computing chip:
Must Include:
- Physical components (waveguides, modulators, optical interconnects)
- Integrated algorithms controlling or optimizing light flow
- Measurable improvement (speed, energy efficiency, computational power)
Avoid:
- Claiming only pure optical algorithms
- Abstract computational methods without physical effect
Example Claim (Simplified):
“A photonic computing chip comprising:
- an array of optical waveguides,
- modulators controlling light signals based on a control algorithm,
- an optical interconnect routing data between photonic cores,
wherein the arrangement reduces latency and increases computation throughput.”
5. Key Takeaways
- Software alone → not patentable; hardware integration is critical.
- Technical effect on physical computation is essential.
- Prior US and EU case law strongly influence patent examination globally.
- Hybrid photonic-electronic designs strengthen inventive step claims.
- Once granted, patents have strong enforceability if technical contributions are clearly documented.
RELATED Blog
comments