Part 3 of a three-part series on memory, invention, and licensing, written from MonolithIC 3D's vantage point, which has published and patented 3D integration since 2010.
What changed?
In eBay v. MercExchange (2006), the Supreme Court replaced the near-automatic injunction after a finding of infringement with a four-factor test [2]. The Court expressly noted that self-made inventors who prefer to license may well qualify, but the practical bargaining environment changed. When infringement can credibly end in exclusion, the commercial question is:
Should we take a license, or risk losing the ability to sell the product?
When the likely remedy is only money, and only after years of litigation, it can become:
Why license now rather than litigate first?
A 2023 study in the Strategic Management Journal (Ozden & Khashabi) confirmed the effect empirically: the reduced probability of injunctions reduced U.S. firms' propensity to license their technology, hitting smaller firms hardest. [7]
The America Invents Act then added the PTAB, so a patent owner may defend the same patent simultaneously in court and in a Patent Office validity proceeding. A 2024 GAO report found consequences for financing: fewer law firms take patent cases on contingency, and third-party litigation funders can require two to three times their investment back before the patent owner sees proceeds [8]. What was once:
Inventor → Patent → Manufacturer → License
became:
Inventor → Lawyers → Funding → Multiple proceedings → Litigation → Settlement.
A first-hand data point
I can add my own experience. We founded MonolithIC 3D in early 2009 with my own resources, and over the years tried to raise capital in every form the market offers: secured loans against the patents, joint ventures with semiconductor and electronics companies, venture capital. Despite building a growing portfolio in a field the industry was clearly moving toward, we never even reached serious discussions. We didn't see revenue until late 2020, more than a decade and many millions of dollars later, when we sold 21 patents to Samsung. The lesson: when exclusion is not credible, patents make weak collateral and licensing revenue is treated as speculative [9]. The inventor funds the invention, or not at all.
And our experience is not an outlier. Our own informal survey suggests that the twenty-five years before 2010 produced a steady stream of invention-based semiconductor licensing companies - MOSAID, Qualcomm, SanDisk, Rambus, Tessera, Saifun, eMemory, several of them venture-backed and publicly listed - while it is hard to name even one founded on that model in the fifteen years since. And this was exactly the wrong period for such a drought: the years before 2010 were classical Moore's Law scaling on the planar CMOS transistor, while the years since brought far more fundamental change - FinFET, gate-all-around, 3D NAND, 3D integration - the kind of foundational invention that historically spawned licensing companies. More invention should have produced more inventor companies; instead it produced almost none.
The exception that proves the rule
One venue still treats exclusion differently: Section 337 proceedings at the U.S. International Trade Commission, where the Federal Circuit held (Spansion v. ITC, 2010) that the eBay test does not apply [11]. ITC cases are not easy - infringement and a domestic industry must be proven, validity defenses remain, and investigations average about 16 months [12] - but the remedy is exclusion, and the ITC expressly recognizes complainants that rely on licensing rather than manufacturing [13].
This difference was on display this summer. Netlist won a $303 million jury verdict against Samsung in 2023, and a separate case later added a judgment of over $100 million; yet years of appeals and PTAB challenges followed, with the judgments unpaid. Netlist then filed two Section 337 complaints (the first filed on September 30, 2025). In August 2026, after about a year at the ITC and before any ITC hearing, the parties settled comprehensively: license fees of up to $750 million to Netlist over five years - with the total Samsung obligation approaching $900 million - plus a patent cross-license, a supply agreement, and an end to all pending matters between them [15, 16]. A decade of district-court litigation produced judgments; a comprehensive license followed within a year of the ITC filings. Timing does not prove motive, and the settlement resolved many proceedings at once - but as an illustration of what a credible exclusion remedy changes, the sequence is clear.
The same patent owner then repeated the demonstration. Netlist approached Micron about a license in April 2021 and was turned down; it sued in the Eastern District of Texas in 2022, and Micron answered by challenging the asserted patents at the Patent Office [21]. In May 2024 a jury found willful infringement and awarded $445 million, even though the Patent Office had already found the principal patent's claims unpatentable [22]. The two tracks ran in parallel for years. On September 2, 2026 the Federal Circuit affirmed the invalidation of five other Netlist patents asserted against Micron, and a week later it heard argument on the two patents behind the verdict, with reporting from the hearing describing the panel as skeptical of Netlist's position [23, 24]. More than five years after the license request, and more than two years after the jury spoke, nothing had been paid, and the verdict itself remained uncertain. Then, in September 2026, Netlist filed a new Section 337 complaint seeking to exclude Micron memory products from the U.S. market [25]. On October 6, 2026, weeks later, the companies announced a comprehensive resolution: a settlement of all pending proceedings and a five-year license to Netlist's entire patent portfolio for $30 million per quarter, $600 million total [25, 26]. Five years of money-only litigation produced a verdict that never became money; within weeks, an exclusion filing was followed by a signed license. Twice in two months, with two different manufacturers, the same sequence played out.
Netlist's experience is not unusual; it is the norm. A Berkeley Haas study of every patent damages award above $10 million appealed to the Federal Circuit from 2010 to 2025 found that only about a third survived intact; outside biotech, awards above $100 million held up only about one time in ten, and not one award above $1 billion survived [27]. Even the people who finance these cases agree: after a jury returned a record $5.7 billion patent verdict in September 2026, the litigation funder behind the plaintiff told its investors that "very few large patent verdicts survive the post-verdict process intact" [28]. A money-only remedy that arrives years late, and rarely intact, gives a manufacturer every reason to litigate first; a credible exclusion order gives it a reason to license now.
What history says about the stakes
This is not a niche legal debate; it is a lesson nations have learned before. England's Statute of Monopolies (1624) swept away crown-granted monopolies with one exception: exclusive rights for the true inventor of a new manufacture, for a limited term [17]. Within a few generations England became the workshop of the world, and economic historians count its enforceable patent rights among the institutions that made the Industrial Revolution British and underwrote the national wealth and power that followed. The American founders understood the lesson well enough to write the same bargain into the Constitution [18]. And if strong, enforceable patent rights could do that in an age of steam and steel, the stakes are far higher today.
Fifty years ago, the value of America's largest companies lay mostly in things one could touch: plants, equipment, inventory. Today that relationship has inverted, and most of the S&P 500's market value resides in intangible assets, designs, software, processes, and the patents that protect them [19]. The Patent Office's latest accounting finds that IP-intensive industries now generate more than 40% of U.S. GDP and pay a substantial wage premium over the rest of the economy [20]. Exclusion matters more in this economy, not less, because of the nature of the assets themselves: a modern process technology takes years and enormous sums to develop, yet once disclosed in a published patent, it can be copied at a fab anywhere in the world at almost no additional cost. What separates that disclosure from a free ride is the credible ability to exclude. In the information era, technology leadership drives national wealth beyond anything else, and the nation whose inventors are protected is the nation whose inventors keep inventing - and disclosing.
The policy question
If society wants inventors to disclose their innovations rather than keep them secret, should the patent system provide a practical path for those inventions to become commercially licensable assets again?
Policymakers are interacting with it. The bipartisan RESTORE Patent Rights Act (S. 708 / H.R. 1574), introduced and pending in committee, would restore a presumption of injunction after final judgment of infringement [3], and the Department of Justice and the USPTO have jointly emphasized that unduly limiting injunctive relief undermines incentives to innovate, and that the public interest strongly favors enforcement of valid patent rights through exclusion orders [4, 5, 6]. For semiconductor innovation, where a small technical team can contribute ideas eventually manufactured at enormous global scale, the answer matters.
- SanDisk Corp. v. Kingston Technology Co.: federal court discussion of SanDisk’s historical Intel, Toshiba, and Samsung portfolio licenses and royalty structure
- eBay Inc. v. MercExchange, L.L.C., 547 U.S. 388 (2006)
- RESTORE Patent Rights Act of 2025, S. 708 / H.R. 1574, 119th Congress
- U.S. Department of Justice and U.S. Patent and Trademark Office, Statement of Interest, Collision Communications Inc. v. Samsung Electronics Co., Ltd. (Feb. 27, 2026)
- U.S. Patent and Trademark Office, “USPTO and DOJ file statement of interest reaffirming the importance of incentives to innovate” (Feb. 27, 2026)
- U.S. Department of Justice and U.S. Patent and Trademark Office, Joint Comment on the Public Interest before the U.S. International Trade Commission (Nov. 25, 2025)
- S. Aydin Ozden & P. Khashabi, “Patent Remedies and Technology Licensing: Evidence from a Supreme Court Decision,” Strategic Management Journal 44(9): 2311-2338 (2023)
- U.S. Government Accountability Office, “Intellectual Property: Information on Third-Party Funding of Patent Litigation,” GAO-25-107214 (Dec. 5, 2024)
- M. Schultz, “The Importance of an Effective and Reliable Patent System to Investment in Critical Technologies,” Alliance of U.S. Startups and Inventors for Jobs (July 2020)
- Tessera-Micron License Agreement (SEC filing, 2006)
- Spansion, Inc. v. International Trade Commission, 629 F.3d 1331 (Fed. Cir. 2010)
- U.S. International Trade Commission, Section 337 statistics: average length of investigations
- U.S. International Trade Commission, Section 337 statistics: complainants that are non-practicing entities
- Federal Trade Commission, In the Matter of Rambus Inc., Docket No. 9302
- Sterne Kessler, “Client Netlist Negotiates Five-Year Patent Alliance With Samsung” (Aug 10, 2026)
- ip fray, “Netlist, Samsung settle multi-venue patent dispute, sign five-year memory patent cross-licensing deal” (Aug 5, 2026)
- Statute of Monopolies 1623 (21 Jac. 1, c. 3), legislation.gov.uk
- U.S. Constitution, Article I, Section 8, Clause 8 (National Archives)
(Research and drafting assisted by AI tools. The analysis and conclusions are presented for discussion and do not address the merits of any pending or contemplated legal proceeding.)
This article is for general informational purposes only. It does not constitute legal advice or a statement regarding infringement, validity, liability, remedy, or any pending proceeding.







RSS Feed