Scalable System on a Chip One Processor Design That Grows From Phone to Desktop

This patent describes a system on a chip design that can be scaled up or down, so the same building blocks serve a small mobile device or a powerful desktop. It comes from Apple Inc., the maker of the iPhone, iPad, and Mac.

The Problem

A modern system on a chip, or SoC, packs the processor cores, graphics, memory controllers, and communication pathways for a whole computer onto a single piece of silicon. Designing one is enormously expensive, so a company would ideally reuse the same design across many products. The catch is that a phone, a laptop, and a high-end desktop have very different needs: more processing cores, more graphics power, and far more memory bandwidth at the top end. Traditionally, spanning that range means designing several separate chips from scratch. What manufacturers want is a single architecture whose components can be replicated and stitched together in different quantities, so one design effort covers an entire product line without redrawing the plumbing each time.

What This Invention Does

The patent describes an SoC built for scalability. The chip is organized around a communication fabric that links processor clusters, graphics, and memory controllers, and that fabric is designed so the blocks it connects can be duplicated. Multiple instances of the chip, or multiple copies of its internal blocks, can be joined so that software sees them as a single unified system rather than separate parts. The design addresses how memory requests are routed and kept coherent across the copies, how interrupts are distributed, and how the parts present a consistent address space so programs run unchanged regardless of how many blocks are present. The result is an architecture where the same engineering can be configured into a modest chip for a compact device or a much larger one for a workstation, with the communication fabric handling the growth in scale.

Key Features

• Scalable fabric. A communication fabric connects the on-chip blocks and is built so those blocks can be replicated.
• Unified view. Multiple chips or blocks are joined so software treats them as one coherent system.
• Coherent memory. Memory requests are routed and kept consistent across the replicated parts.
• Distributed interrupts. Interrupt handling is spread across the copies so the combined system behaves as a whole.
• One design, many products. The same architecture configures into small mobile chips or large desktop-class ones.

Who Is Behind It

The applicant is Apple Inc., the Cupertino-based technology company whose in-house Apple silicon chips now power its iPhone, iPad, and Mac lines. The filing lists a large team of processor architects, including Per H. Hammarlund, Lior Zimet, and Gerard R. Williams III among many others. This filing is a divisional application from an earlier related Australian application.

Why It Matters

Reusing one chip architecture across a whole product family is one of the biggest levers a device maker has for both performance and cost. A design that scales cleanly from a phone to a desktop lets a company move the same proven technology up and down its range and reach the very high memory bandwidth that demanding professional software needs. Protecting the architecture in Australia supports the company’s products in a significant consumer and professional market.

Related Concepts

• System on a chip – the single-chip computer design at the heart of this patent.
• Network on a chip – the on-chip communication fabric that links the blocks.
• Cache coherence – keeping memory consistent across replicated parts.
• Apple silicon – the company’s family of custom processors.
• Multiprocessing – running many processor cores together as one system.

---

AU 2026201978 was published in the Australian Official Journal of Patents on 2 April 2026 and is open for public inspection. Patent applications represent inventions that are sought to be protected and do not necessarily reflect commercially available products.