EnSilica plc logo

EnSilica plc: Custom Chips for the AI World

EnSilica designs custom-made computer chips for industries including satellites, cars, factories and data centres. It earns a development for each design, then years of further income for supplying the finished product, benefiting from the growing demand for chips built for specific AI tasks.

LSE:ENSI
$76.40-3.29%
Updated: Aug 13, 2026
Technology
microuk

Bull & Bear Case

An overview of the main reasons to invest and the key risks involved.

Bull Case

Initial development fees become years of revenue

Each funded design can become a decade-long supply contract, so revenue compounds.

Reusable chip designs sold to multiple customers

Owning the design lets EnSilica sell the same chip to several customers at once.

Growing demand for custom-designed chips

More products need chips built for one job, widening the pool of potential customers.

Bear Case

Chip programmes can slip or be cancelled

Customers can delay or drop projects, pushing the supply revenue years further out.

Small company funding big design costs

Designing chips costs cash upfront, and slippage strains a small balance sheet.

Depends on outside foundries and a few large customers

Factories and key accounts sit outside its control, so disruption hits revenue directly.

Executive Summary

About EnSilica

EnSilica is a chip design house: companies come to it with a product idea, and EnSilica designs a custom chip to run it, then keeps supplying that chip in volume for years. Its customers build satellites, cars, factory equipment and data centre hardware. They pay engineering fees during the design phase, followed by repeat income per chip shipped. EnSilica owns no factories, relying on outside foundries such as TSMC, and employs over 200 people, mostly engineers.

The case rests on that shift from one-off design fees to long-life chip supply, helped by reusable building blocks it can sell to several customers. The debate is timing and scale: chip programmes are slow, cash-hungry and prone to delay, and EnSilica is a small company funding development in a market of giants.

Investment Thesis

Overview of buy and sell case of the business.

Why Invest?

Key pieces of information about the business that you need to know about.

Initial development fees become years of revenue

A customer pays EnSilica to design a chip for their specific product, which covers much of the engineering cost upfront. Once that chip is proven in silicon, EnSilica supplies it for as long as the customer builds the product. Published contract examples include a chassis controller for cars worth around $40m over six years and a timing controller worth around $30m over ten years. One design win therefore creates a long, predictable tail of repeat orders rather than a single payday.

Reusable chip designs sold to multiple customers

Alongside chips built for a single client, EnSilica invests in its own designs, called ASSPs, meaning standard parts aimed at a whole industry rather than one buyer. Selling the same silicon to many customers spreads the development cost and carries better margins than bespoke work. Each project also adds circuit blocks to a shared library, cutting the time and cost of the next design. Larger fabless peers such as Broadcom built on the same reuse logic.

Growing demand for custom-designed chips

More and more companies want a chip built for their own product rather than an off-the-shelf part, because a custom design can be smaller, cheaper to run and harder for rivals to copy. Few of those companies employ the specialists needed to do it, which is the gap EnSilica fills. Its particular strength is mixing analogue and digital functions on a single piece of silicon, one of the harder jobs in the industry, and it sells that same skill into satellite communications, vehicles and industrial equipment. Customers named in company materials include Siemens, Visteon and AST SpaceMobile. Serving unrelated end markets means a slowdown in one, such as car production, does not halt the work coming from the others, which matters for a company whose revenue arrives in large, lumpy contracts.

Catalysts

The key events that could drive investment opportunities and shift markets.

Near term
  • Cash Generation Milestone: The next stretch is about designs already funded turning into chips shipped in volume. The board has said it expects to reach positive monthly operational cash generation by the end of calendar 2026, which would reduce reliance on outside funding.

  • Terminal Chips Sampling: Four satellite terminal chips are with customers for testing, according to the July 2026 corporate presentation. Successful evaluation is the step that turns a sample into a production order and recurring supply revenue.

Medium term
  • Payload Study Conversions: Two further satellite payload engagements sit at funded study stage. EnSilica says these are expected to convert into funded contracts carrying development fees plus follow-on supply or chip-as-a-service income.

  • New Design Wins: Management targets three to four new design and supply contracts each year. Each win adds development revenue now and, if the chip reaches production, a supply stream that can run for the best part of a decade.

Long term
  • Terminal Volume Ramp: EnSilica's own analysis puts the addressable market for satellite terminal chipsets at roughly five times its 2025 size by 2030 as constellations move to mass adoption. Volume orders from terminal makers would shift the mix towards higher-margin repeat supply.

  • Secure Processor Rollout: A £5m UK Contract for Innovation funds a processor built to resist future quantum-computer code-breaking, aimed at critical national infrastructure. Tighter security rules such as the EU Cyber Resilience Act could widen demand for that intellectual property.

Key Risks

Key pieces of information about the business risks that you need to know about.

Chip programmes can slip or be cancelled

Getting from a design win to volume shipments takes years, and the profitable supply phase only starts once the chip is proven and the customer's own product sells. If a customer delays, redesigns or abandons a programme, EnSilica keeps the engineering fees but loses the long supply tail the case depends on. Pipeline figures the company publishes are opportunities and lifetime estimates, not committed orders.

Small company funding big design costs

EnSilica spends real cash each year capitalising chip development and its own intellectual property, and much of that spend lands well before the resulting silicon earns anything. The company carries external loans and has raised money since listing on London's AIM market in 2022. If programmes slip or customer milestone payments arrive late, a business of this size has limited room to absorb the gap without further funding.

Depends on outside foundries and a few large customers

Being fabless means the actual manufacturing sits with partners such as TSMC and GlobalFoundries, so capacity, pricing, export rules and geopolitics around chip production are outside EnSilica's control. Revenue is also concentrated in a modest number of large programmes and named customers, so one lost account or one supply disruption matters far more than it would at a broadly based competitor.

Follow the Experts

Quickly navigate key insights from industry experts and leverage their knowledge and market intelligence.

Paul Schell profile

Paul Schell

Senior Analyst, AI & Machine Learning, ABI Research

1k Followers audience

Expert Insights

"Edge AI silicon must be more specialized for complex, device-dense industrial environments"
Alexander Harrowell profile

Alexander Harrowell

Principal Analyst, Advanced Computing for AI, Omdia

1k Followers audience

Expert Insights

Omdia forecasts application-specific AI silicon taking share in edge devices "largely at the expense of GPUs"

Investor Materials

Access the most recent investor updates published by the company.

Investor Presentation

Team

Meet the experienced professionals leading our organization

Ian Lankshear - undefined

Ian Lankshear

Kristoff Rademan - undefined

Kristoff Rademan

What the Pros are asking

Here are the questions that professional investors are asking before making an investment decision.

How does EnSilica actually make money, in plain terms?

Money arrives in two waves. First, a customer pays engineering fees to have a chip designed for their product, which the industry calls non-recurring engineering because it is paid once per design. Second, if the chip goes into production, EnSilica supplies the finished parts for as long as the customer builds the product, which can be six to ten years. It also sells consultancy engineering time and licenses some of its own circuit designs to other chip companies.

If it does not own factories, who makes the chips?

Outside manufacturers do, and that is what fabless means: EnSilica designs and sells the chip, and specialist foundries build it. Company materials name TSMC and GlobalFoundries as foundry partners, with packaging and testing also outsourced. The advantage is that EnSilica never has to fund a multi-billion-pound plant and can use the latest production processes. The trade-off is dependence on those partners for capacity, cost and delivery, none of which it controls directly.

Why does the space and satellite work get so much attention?

Satellite communications is where EnSilica sees the biggest growth, and it sits on both sides of the link. It designs beamforming chips that steer signals inside satellites, proven through delivery to AST SpaceMobile, and chipsets for the ground terminals people and businesses use to connect. It has secured funding from the European Space Agency and the UK Space Agency for the terminal chipset work. New constellations from operators including Eutelsat, ViaSat and Amazon are what drive that demand.

Who else does this, and how big is EnSilica next to them?

The comparison set is other fabless chip designers and design-and-supply houses, from giants such as Broadcom, Nvidia and Marvell down to specialists like Taiwan's Global Unichip. EnSilica sits at the very small end of that group, employing over 200 people from Oxford with engineering teams in the UK, Germany, Hungary, India and Brazil. It competes by specialising in complex chips that blend analogue and digital functions for safety-critical, long-life products rather than chasing high-volume consumer parts.

What has to go right for the business model to work?

Three things, and they are linked. Chips already designed have to reach volume production, because supply revenue is where the margin and the repeat income sit. New design wins have to keep arriving at the company's stated target of three to four a year to refill the pipeline. And the cash spent on development has to stay within what the business can fund. Management has said it expects positive monthly operational cash generation by the end of calendar 2026.