Three Offices, One Independence Day

On 9 April 2026, six complete patent specifications were filed with the Indian Patent Office. Together they cover the entire THATTE stack — from the SWCNT@MWCNT device to the THATTEOS microkernel. Application numbers were assigned, fees were paid, and a clock started running.

That clock is the Paris Convention priority window. Any patent filed internationally within twelve months of the Indian priority date can claim that April 9 date as its own. The legal deadline is 9 April 2027.

I chose not to wait for it.

The target is 15 August 2026. India’s Independence Day. The 79th.

Fifteen days from today, I intend to walk into three patent systems simultaneously — an act that, to me, carries more than legal meaning. This essay is about why that date, why those three offices, and what the arithmetic of intellectual sovereignty actually looks like from the inside.

The Treaty and Why I Rejected It

The standard path for international patent protection is the Patent Cooperation Treaty. A single PCT application, filed once, preserves rights in over 150 countries simultaneously. You get an international search report. You defer the national-phase decisions by up to thirty months. For a solo inventor with six patents, it sounds like exactly the right instrument.

I spent serious time on this calculation. The cost arithmetic was clear, and it was unsustainable. Each PCT application carries international filing fees, a search fee payable to the chosen International Searching Authority, and national fees for each contracting state. Running all six through ISA/EP — the European Patent Office as search authority, the highest-quality option for electronics and physics claims — would have required a budget I do not have, for protection in markets where enforcement, for an inventor working without institutional backing, is uncertain at best.

There was also a quality question. ISA/IN — the Indian Patent Office as search authority, the cheapest PCT option — has a search quality and turnaround time that I could not rely on for device-physics and compiler-architecture claims. The credible alternatives brought the total back to the unaffordable range.

The PCT is the right instrument when you have strong enforcement prospects across many jurisdictions simultaneously, or when a licensee or investor will absorb the cost. Neither condition applies today. The priority date is what matters now. File first. Secure the ground. Expand later.

Three Questions for Three Offices

Once the PCT was off the table, the question became simple in structure and harder in execution: for each group of patents, where are the most likely infringers and licensees, and is the patent system in that jurisdiction workable for claims of this kind?

Patents are instruments of enforcement. You do not file everywhere. You file where you can act.

Three answers emerged.

The United States for Hardware

Thatte1 through Thatte5 are hardware patents: device physics, gate library, processor architecture, memory fabric, cryptographic hardware. They describe physical structures, fabrication processes, and electrical circuits built from carbon nanotubes. There is nothing abstract about them.

The primary potential licensees and infringers for a photonic-ternary device are American companies: semiconductor manufacturers, cloud infrastructure companies, chip designers. A US patent is the only instrument that reaches them directly on their own soil. An Indian patent, against a company incorporated in California, requires bilateral enforcement machinery that does not exist. A USPTO grant does not require it.

Hardware claims grounded in physical structure and specific material choices face no serious risk under the Alice/Mayo doctrine that has invalidated thousands of US software patents since 2014. A claim describing two concentric carbon nanotubes, a specific CVD fabrication sequence, and a measured switching current is not an abstract idea. It is a physical object. The USPTO examiner has something tangible to search against.

The micro-entity fee schedule — available to individual inventors below the income threshold who have not previously filed more than five US applications — makes this genuinely accessible. Five hardware patents. Affordable. Directly enforceable in the market that matters most.

Singapore for Software

Thatte6 is different. It covers the ManiT balanced ternary programming language, its dual-target compiler, and the THATTEOS microkernel operating system. These are software inventions, and software inventions are a jurisdictional problem everywhere — but the problems are different in different places, and those differences matter.

In India, Section 3(k) of the Patents Act 1970 excludes “a computer programme per se” from patentability. The phrase “per se” leaves room for software that produces a technical effect — but that room is narrow and inconsistently applied, particularly for compiler and OS claims. Filing Thatte6 at the Indian Patent Office would mean fighting an uphill examination on claims that are genuinely novel and genuinely technical, but which an Indian examiner might dismiss on categorical grounds before engaging with their substance.

At the USPTO, the Alice/Mayo doctrine has made software patents unreliable since 2014. Claims that describe a compiler or an operating system require careful hardware anchoring to survive §101 rejections. It is doable, but it carries real prosecution risk on top of what are already dense specifications.

The Intellectual Property Office of Singapore applies a technical effect doctrine that is cleaner for software that does something measurable and non-obvious. A compiler that translates source code into a novel balanced ternary instruction set, targeting a photonic-ternary device that produces physical current states, is not an abstraction. It is a tool with a specific technical output. IPOS examiners have granted patents on claims of this kind, consistently.

Singapore also positions well for technology licensing in Asia, where the manufacturing concentration for semiconductors and advanced packaging is highest. If the THATTE stack reaches research groups or fabless chip companies in Singapore, Taiwan, South Korea, or Japan before it reaches the US — a plausible path — a Singapore patent establishes enforceability in the region without the cost and complexity of parallel filings at four different offices.

India for Additional Filings

The second batch covers additional aspects of the stack — array architecture, integration variants, bridging circuits, and application-specific processing. The details will be public after filing. None of these were in the April filing. They extend the stack in directions that became clear only as simulation and design work developed through the months after the first priority date.

These additional patents need a priority date. The fastest and most natural way to establish one, for an Indian inventor, is to file with the Indian Patent Office. Filing on 15 August creates a coherent priority date that mirrors the structure of the April 9 date.

Why Independence Day

The legal deadline for international filing is April 9, 2027. August 15 is not the deadline. It is a choice.

The practical reason is straightforward: filing seven months before the deadline means seven months of buffer. A technical defect in the specifications, an administrative error in the filing system, a change in fee schedules, a server failure at USPTO Patent Center or IPOS eServices — any of these, at the twelve-month boundary, is an unrecoverable event. Filed in August, there is time to correct errors. Filed in March, there is not.

But I will not pretend the date is only practical.

India became a free country on 15 August 1947. On that morning, a nation that had been subject to another’s economic and intellectual order stood up and said: this is ours, and we will administer it ourselves. The act was not conquest, not inheritance. It was deliberate, legal, sovereign assertion. That is what Independence Day marks.

What I am doing on 15 August 2026 is smaller in every way that matters historically. But it is structurally the same act. An Indian inventor, working on his own machines, with his own simulations and his own code, asserting: this is mine, and here are the legal instruments, in three jurisdictions, that say so.

The Paris Convention was itself a 19th-century act of international cooperation — a treaty designed precisely so that inventors without the resources of large corporations could file in one place and have time to arrange the rest. I am using it exactly as intended. India gave me the first priority date. On her Independence Day, I give back three more.

There is something fitting about the arithmetic. The first batch was six patents, filed on a quiet Thursday in April, numbers assigned and clock started. The second batch, plus the international extensions of the first six, all converging on one date with its own weight. The 79th Independence Day is not a deadline. It is a dedication.

What Comes After

After 15 August, the work continues. Through the second half of 2026, continuation filings at the USPTO and IPOS will incorporate new embodiments and expanded specifications developed since the April priority date. The patents are not frozen at their first filing; the priority date is secured, and the specifications can grow.

In early 2027, the ManiT compiler and THATTEOS kernel will be released as open source — after all filings are confirmed, not before. File first. Publish second. Always. A Zenodo preprint will be submitted alongside the open-source release, establishing a public record of the physics, the simulation results, and the design decisions. The work has been done in private because it had to be. It will be published because it should be.

April 2027 closes the Paris Convention window for Batch 1. Whatever has not been filed internationally by that date cannot claim the April 9, 2026 priority. The US and Singapore filings will be in place. If traction develops — licensing interest, infringement evidence, a collaborator willing to share the cost — the EPO, Japan, China, and other offices remain open. A USPTO or IPOS grant establishes priority for those later filings. Nothing is foreclosed. But the priority date is everything, and on 15 August, it is secured.