Case Study: IP Trend Analysis Case Study: Timing a Semiconductor R&D Bet

This IP trend analysis case study follows a mid-size fabless semiconductor company deciding where to concentrate a fixed R&D and patent budget for the next design cycle, and shows how reading multi-year filing curves across the USPTO, EPO and WIPO pointed the board to the technology cluster gaining momentum rather than the one that had already peaked. It is a representative scenario: every hard figure below is a cited public statistic, and every model output is framed as what the method produces, not a disclosed client result.

The Challenge

The client competed in a market that had just posted its best year on record. Global semiconductor sales reached $627.6 billion in 2024, up 19.1% on 2023, and the industry association was already projecting double-digit growth again for 2025. A rising tide, however, is not a strategy. The board had budget for exactly one serious platform bet – advanced packaging and chiplets, wide-bandgap power (GaN and SiC), or a dedicated AI-accelerator block – and could not fund all three at the depth needed to build a defensible patent position.

The internal debate had stalled because each camp argued from revenue that already existed. Revenue is a lagging signal: by the time a segment shows up in the sales mix, the filing race that protects it is usually two to three years old and the best claim space is taken. The founders needed a leading indicator – evidence of where R&D capital across the industry was being committed now, before it converted into product – and they needed it segmented, not as an aggregate industry headline.

They also needed to know who they would be filing against. Committing a multi-year budget to a cluster already saturated by a handful of prolific assignees is a different risk from entering one where filing is rising but ownership is still fragmented. The question the board put to us was narrow and answerable: which of the three clusters is accelerating fastest in the patent record, and where is there still room to own claims?

Our Approach

We ran the engagement as an IP trend analysis – a time-series read of filing behaviour – rather than a one-off landscape snapshot, sequenced so each step fed the next:

• Cluster definition from classification. We defined the three candidate clusters against patent classification (CPC/IPC) subclasses rather than marketing labels, so every boundary could be traced to a code an examiner recognises and every count was reproducible year over year.
• Multi-year filing curves per office. For each cluster we pulled annual filing and grant counts from the USPTO, the EPO and the WIPO PCT system separately, because the three offices lead and lag each other and a trend confirmed in all three is far more reliable than one office’s noise.
• Momentum and inflection detection. We measured not just the level of filing but its rate of change – where a curve was bending upward, flattening, or rolling over – to separate a cluster still accelerating from one that had already had its filing peak.
• Assignee trajectory benchmarking. For each cluster we tracked the top applicants over time to see whether ownership was concentrating into a few hands or staying fragmented, and whether the client’s likely competitors were increasing or easing their filing cadence.
• Policy and priority overlay. We flagged external accelerants – CHIPS Act fab investment and the USPTO Semiconductor Technology Pilot Program that grants special (expedited) status to qualifying semiconductor applications – because they change how fast a competitor can convert a filing into an enforceable grant.

The discipline that made the readout credible was triangulation. A single office can be distorted by a fee change, a classification reshuffle or one applicant’s filing spree. By requiring a trend to appear in USPTO grants, EPO applications and WIPO PCT publications before we called it real, we gave the board a signal it could act on rather than a chart that flattered whichever camp commissioned it.

What the Research Found

The office-level data confirmed how central semiconductors have become to the whole patent system. At the USPTO, semiconductor technology was the single largest field for granted patents for the third consecutive year, rising from 49,831 grants in 2021 to 67,118 in 2024 – a 34.6% increase over three years, against a backdrop of total US patent applications hitting an all-time high of 430,625 in 2024. This is a field pulling ahead of the average, not merely riding it.

The international picture pointed the same way. WIPO recorded roughly 273,900 PCT applications in 2024, with semiconductors and digital communication the leading published-application fields – a signal that the cross-border filing appetite in chips is broad, not confined to one home market. At the EPO, computer technology became the top field for the first time in 2024 at 16,815 applications, with AI-related filings up 10.6%, underscoring that the fastest-moving demand sits at the intersection of chips and AI compute.

The assignee trajectories decided the bet. The prolific chip patentees – Samsung, which stayed the top US patent recipient overall with 6,377 grants in 2024, and TSMC, which climbed to second with 3,989 (up 8%) – had concentrated their acceleration in advanced logic and packaging, the most contested cluster in the record. The wide-bandgap power cluster, by contrast, showed a filing curve still bending upward but with ownership far more fragmented across power-electronics and automotive suppliers – rising momentum, open claim space.

Put together, the trend analysis reframed the choice. Advanced packaging was growing but crowded and expensive to enter behind Samsung and TSMC; a bare AI-accelerator block put the client into direct claim conflict with far larger balance sheets; wide-bandgap power was accelerating in all three offices while its patent ownership was still up for grabs. The fastest-growing defensible slice was not the loudest one in the boardroom.

The Outcome

The board redirected the platform budget toward the wide-bandgap power cluster and scoped a filing program timed to the momentum the curves revealed – moving before the ownership fragmented into fewer hands. Rather than chasing advanced packaging into a field the two largest assignees were already accelerating in, the company committed where the trend was rising and the claim space was still open.

Crucially, the decision was wired to the patent calendar as well as the technology. Because we had flagged the USPTO Semiconductor Technology Pilot Program, the client sequenced its earliest filings to use expedited examination, compressing the time to an enforceable grant in exactly the cluster where being first to a defensible claim mattered most. The policy overlay turned a technology choice into a timing advantage.

Most durably, the company kept the trend model as a living dashboard. Because every cluster was defined by classification and every curve was tied to a public office release, refreshing the analysis each time new USPTO, EPO and WIPO data landed was a data update, not a rebuild – so the R&D and patent budget could be re-steered on evidence as the next cycle unfolded.

What This Means for Similar Matters

• Revenue is a lagging signal; filing curves lead it. An IP trend analysis shows where R&D capital is committed two to three years before it appears in the sales mix.
• Triangulate across offices. A trend confirmed in USPTO grants, EPO applications and WIPO PCT publications is actionable; one office alone can be distorted by fees, reclassification or a single filer.
• Measure the rate of change, not just the level. The cluster with the steepest still-rising curve, not the biggest absolute count, is where the next defensible position sits.
• Read assignee trajectories. A rising cluster with fragmented ownership is a far better entry than a rising cluster already concentrating into two balance sheets.
• Overlay policy. CHIPS Act investment and the USPTO semiconductor pilot change how fast a rival can convert filings into grants – and can be used to your own advantage on timing.

What This IP Trend Analysis Case Study Shows

The deliverable was not a chart of who leads the industry today – it was a directional read of where the industry is heading and how fast. That distinction is the whole point of an IP trend analysis case study: a landscape snapshot tells you the current standings, while a trend read tells you which standings are about to change.

By defining each cluster against classification codes and pulling annual counts from three offices, we produced a model whose every data point is reproducible and whose every inflection can be re-tested when the next year of data lands. A board can start at the recommendation and walk back to the specific filing curve that justified it without hitting a single unsupported leap.

Why Filing Curves Lead the Chip Cycle

Semiconductor companies patent before they tape out, and a patent family is typically published around 18 months after its priority date – so a rising family count is a paid-for commitment to a technology direction that has not yet reached the market. When wide-bandgap power filings climb across the USPTO, EPO and WIPO while advanced-packaging ownership concentrates into two dominant assignees, the record is telling you where conviction sits and where the door is closing, well ahead of the revenue mix. Reading those curves per cluster, not in aggregate, is what surfaces the divergence a chipmaker needs to time its next bet – and it is why the same method that sizes a market can also steer an R&D budget.

Reading the Trend Against a Record-Breaking Market

Trend analysis only earns its keep when the market underneath it is moving, and in semiconductors it is moving fast. After 2024’s $627.6 billion in global sales, the industry closed 2025 at a record $791.7 billion, up 25.6% year on year – the steepest annual jump in more than two decades, driven overwhelmingly by demand for AI accelerators and the high-bandwidth memory that feeds them. A market compounding at that rate does not reward a company that waits for the sales data to confirm a direction; by then the filing race is over.

That is why the patent signal has to be read forward. The EPO’s shift, with computer technology overtaking every other field and AI-related filings up 10.6% in 2024, is the same story the sales figures tell – but visible a year or two earlier and broken down by exactly the sub-technologies a board has to choose between. The client’s three candidate clusters all sat inside this surge; the trend analysis simply asked which one was accelerating fastest in the record and still had room for a newcomer to own claims.

The policy layer sharpened the timing further. The CHIPS Act channelled billions into new US fabrication – TSMC alone drew $6.6 billion in direct funding for its Arizona fabs – and the USPTO’s semiconductor pilot shortened the path from filing to enforceable grant for qualifying applications. For a company committing a multi-year budget, those two forces mean a competitor can convert intent into protection faster than the historical average, so the window to establish a defensible position in a rising cluster is narrower than the raw filing curve alone would suggest.

Data Sources

The market and patent data referenced above comes from:

• Semiconductor Industry Association – 2024 Global Sales — Global semiconductor sales reached $627.6B in 2024, up 19.1% on 2023's $526.8B, with double-digit growth projected for 2025.
• Anaqua – Analysis of USPTO Patenting Statistics 2024 — Semiconductor technology the top US patent field for a third straight year, up from 49,831 grants in 2021 to 67,118 in 2024; leading recipients Samsung, TSMC, Intel, BOE, LG.
• WIPO – PCT Yearly Review 2025 — About 273,900 PCT applications in 2024 (+0.5%); semiconductors and digital communication the leading published-application fields; top filers Huawei, Samsung, Qualcomm.
• EPO – Patent Index 2024 — 199,264 European applications in 2024; computer technology the top field for the first time at 16,815 (+3.3%); AI-related filings up 10.6%.
• USPTO – Semiconductor Technology Pilot Program — Grants special (expedited) status to qualifying semiconductor patent applications, supporting the CHIPS Act push to accelerate domestic chip innovation.

Discuss an IP Trend Analysis

Deciding where to concentrate R&D or patent budget? We read the filing curves across offices and tie them to the position you can defend.

Frequently Asked Questions