Business · General Industry
Silicon platform
A reusable base of architecture, IP, software and methodology from which a family of related chips is derived. Investing once in a platform converts later products into derivatives rather than new developments.
- Also known as
- Platform approach, Chip platform
- Type
- Business
- Domain
- General Industry, Industry structure
Notes
The platform argument is also a software argument, since customers value a stable toolchain and driver stack across a family more than any single part.
Overview
A silicon platform is a reusable foundation of chip architecture, pre-integrated IP blocks, software and design methods from which a company derives a family of related chips, rather than designing each one from scratch. The first product on a platform carries most of the cost and risk; later products become derivatives that reuse the validated core and change only what differs. The term is used both by chip companies for their own product lines and by IP suppliers that sell ready-made subsystems for customers to build on.
How it works
A platform typically fixes the parts that rarely need to change: the processor cores and interconnect, memory and I/O subsystems, security and power management, the boot and driver software, and the verification environment. Product teams then add or remove peripherals, scale the number of cores or memory interfaces, or attach a custom accelerator. The idea follows platform-based design, a methodology built on systematic reuse of a library of compatible hardware and software components to reduce risk, cost and time to market.
Platforms take several forms. Microcontroller and processor families share cores, peripherals and software so that customers can move between parts; NXP, for example, presents its i.MX applications processors as a scalable family spanning several series. IP vendors now offer platforms too: Arm's Neoverse Compute Subsystems are validated compute platforms that partners customise, and Arm states they can cut time to silicon by up to a year.
The trade-off is flexibility. A platform makes derivatives cheaper and faster, but decisions baked into it, such as the interconnect, the process node or the security architecture, can be costly to change later, and a product that needs something very different may fit poorly. Companies therefore plan platform generations carefully and decide when a new one is justified.
Examples
- A microcontroller family in which many part numbers share the same core, peripherals and software libraries but differ in memory size and pin count.
- Arm Neoverse CSS N4 or V3, compute subsystems that partners customise into server and infrastructure chips.
- NXP's i.MX applications processor family, presented as a scalable line backed by a product longevity programme.
- A company deriving a lower-cost automotive chip from its flagship design by removing cores and interfaces rather than starting again.
Why it matters
Related terms
Part of the General Industry glossary, with every term in the topic and the lifecycles, tools and standards that use them.
Where it appears on the Hub
Used on 12 pages of the Hub. In each group, the pages that mention it most come first.
Lifecycle stages and gates
Roles
Suggest an improvement
Suggest an improvement to this entry
A clearer definition, an example, a correction or a source an editor can check. Editors review every suggestion before the entry changes.
Last reviewed . Written by The Semiconductor Hub editorial team. The full glossary