Memory Technology Package

Introduction

What The Package Covers

What Users Can Analyze

Why It Matters

Predicting Patent Essentiality with the LLM-Based Essentiality Score

Common Use Cases

Important Scope Note

How To Request Access

🔒

Introduction

This article explains what the Memory technology package covers in IPlytics, the data sources behind it, and how customers can use it for SEP and standards analysis.

A technology package (also called a feature package) is a pre-built, curated collection of patents covering a specific technology standard - for example, Wi-Fi 6, 5G, or HEVC video compression. Rather than constructing a patent search from scratch, users of the IPlytics platform can access these packages directly to immediately explore which companies hold relevant patents, how many patents exist in the landscape, and which may be essential to a given standard. 

The Memory package covers patent landscapes in high-performance and system memory technologies for four major memory standards: HBM, GDDR, LPDDR, and DDR. It's built for tracking patent ownership, portfolio strength, competitive position, and litigation across these technologies, all developed within the JEDEC standards ecosystem.

Unlike other technology packages built using the Semantic Essentiality Scoring Methodology, the Memory package is built using IPlytics' LLM-based Essentiality Score which is explained below.

What The Package Covers

The Memory package focuses on memory standards developed in the JEDEC ecosystem and organizes the data into four technology landscapes.

Users can filter by memory landscape and, where generation-level data exists, drill down by specific standard generation (e.g., HBM3E vs. HBM3, LPDDR5X vs. LPDDR5). 

What Users Can Analyze

• Search and filter patent families relevant to the covered memory and semiconductor standards.
• Review ownership rankings to understand leading innovators and patent owners in each landscape.
• Benchmark portfolios across companies and technology generations.
• Review litigation trends and related patent activity to support licensing, legal, and patent strategy workflows.
• Export row-level search results and use analytics views for visual breakdowns and comparison.

Why It Matters

Memory technologies such as HBM, GDDR, LPDDR, DDR are increasingly load-bearing for AI accelerators, HPC, data centers, graphics processing, and mobile devices. As performance and power requirements tighten, patenting activity around memory interfaces, controllers, packaging (e.g., TSV, interposers), and signaling schemes becomes more strategically relevant to track.

Standards bodies like JEDEC rely on self-declared essentiality companies to voluntarily disclose which patents they believe are essential to a given standard. In practice, these declarations are frequently over-broad, outdated, or simply missing, so they don't reliably show who actually holds relevant patents. The Memory package addresses this by classifying and organizing patents into structured technology landscapes (by memory type and, where available, by generation) giving users visibility into the landscape independent of what's been formally declared.

Predicting Patent Essentiality with the LLM-Based Essentiality Score

Determining whether a patent is essential to a technical standard is one of the most fundamental and challenging tasks in standard-essential patent (SEP) analysis. Traditional approaches rely on manual claim charting which is accurate, but time-consuming, expensive, and difficult to scale across large portfolios.

To address this, LexisNexis IPlytics has developed an LLM-assisted essentiality methodology, similar to our Semantic Essentiality Scoring methodology, that estimates the likelihood a patent is essential to a given technical standard. The DRAM Memory patent database covering HBM, GDDR, LPDDR, and DDR in the Memory Technology Package is the first to use this LLM-based approach. Please note that both are referred to, in the product, under Essentiality Score or ES for short. 

How the LLM-Based Essentiality Score Works

The methodology compares each independent patent claim against all relevant sections of the applicable standard, using a multi-layered framework:

1. Vector-based retrieval first identifies the most relevant sections of the standard, reducing hallucination risk and improving retrieval quality before LLM evaluation begins.
2. LLM claim analysis, guided by carefully engineered prompting, identifies individual claim elements and evaluates whether corresponding technical features exist within the standard, producing transparent reasoning for why each claim element is or isn't supported.
3. Continuous validation, including rigorous error analysis, subject-matter-expert (SME) review, patent owner feedback, error correction, bias reduction, and iterative model refinement.

This combination is what distinguishes the approach from a purely generative AI method. The LLM's output is checked, corrected, and refined rather than taken at face value. For more information on the LLM-Based scoring methodology, click here.

Common Use Cases

Important Scope Note

How To Request Access

The Memory technology package is a licensed add-on to the IPlytics Platform. Access is not included automatically in every standard platform subscription. 

To request access, contact your IPlytics account manager or reach out to the IPlytics team here.