Canada has a world-class, compound semiconductor fab that could be a key part of advanced technology supply chains but it has been underinvested in for two decades.
AI Data centres build outs will drive explosive demand for exactly the kind of components this fab creates.
Bold investment now based on the model of the Netherlands and Taiwan can turn this underused asset into a global champion, creating jobs and securing prosperity.
Summary
The demand for compound semiconductors is expected to grow dramatically in the coming years as a result of massive investments in data centres, modern telecommunication networks, and the growing use of AI.
Canada could be a world leader in cutting-edge semiconductor technology. Previous investments by the Department of Defence and market leadership by NORTEL have left the country with a strategic asset The Canadian Photonics Fabrication Centre. This fabrication center (fab) is one of the world leaders in producing compound semiconductors. With the right focus it could be turned into an essential part of the global supply chain for advanced technology. But we have failed to capitalize on this opportunity for over two decades.
If Canada invests in this facility—following the model of the Netherlands and Taiwan—it could not only build a world-class fabrication facility but also catalyze an entire industry of component designers and manufacturers. This would come with massive economic and security benefits and transform the country’s position within the global semiconductor industry.
What is Canada’s Opportunity?
Semiconductors are the backbone of the modern electronics industry. Most of the time we hear about modern “mega-fabs” from companies like TSMC which manufacture large quantities of silicon semiconductors or about chip design firms like NVIDIA or Intel that design integrated circuits using multiple silicon semiconductors to create the CPUs and GPUs that are the backbone of personal computing as well as large scale datacentres.
However, there is another kind of semiconductor compound semiconductors – so called because unlike silicon semiconductors they are made out of a combination of elements like Indium Phosphide (InP) or Gallium arsenide (GaAs) – that are equally as important to modern advanced technology. Compound semiconductors have different properties that make them unsuitable for general-purpose computing units like CPUs and GPUs. However, they are crucial for rapid information transfer, underpinning all telecommunications and datacommunications, as well as applications such as high-speed signals in 5G, satellite comms and specialized defence radar systems.
Modern data centres require silicon semiconductors to process large quantities of data within server racks and compound semiconductors to enable high-bandwidth data transfer between racks. As AI data centers grow in size and complexity, the need for high speed communications between racks is critical and where photonics built with compound semiconductors shine.
Right now the world is experiencing a boom in AI data center construction. Over the next 5 years to service this new compute capacity (alongside advancements in all other areas involving photonics) production of compound semiconductors is going to need to expand rapidly.
Luckily, Canada is uniquely positioned to benefit from this growth. In the 1980s and 1990s, the Canadian Department of National Defence backed Nortel’s pioneering work in compound semiconductors to support NATO radar and optical telecom systems. Nortel’s components – many manufactured in Canada – became part of the global internet backbone. After Nortel’s collapse in the early 2000s, there was a risk that much of this investment would be lost. But, the National Research Council wisely decided to preserve the expertise and unique know-how by creating the Canadian Photonics Fabrication Centre in Ottawa. Today this remains North America’s only publicly accessible end-to-end Indium Phosphide wafer fabrication facility.
But, instead of being a strategic advantage the Canadian Photonics Fabrication Centre increasingly looks more like a missed opportunity. It operates mainly as a research and prototyping facility producing advanced designs but without commercial production scale. The facility has the tools, workforce, and intellectual property to support a thriving industry, but without investment and scale it cannot capture any of the growing data centre market..
Think of it like having one of the world’s only airplane factories capable of achieving supersonic flight but only ever building prototypes, never selling jets. Canada has wisely maintained the tools and the talent to create cutting-edge compound semiconductors, but without investment and commercialization it is sitting wasted, producing far less value than it could.
By learning from the approach of top electronics supply chain component manufacturers we can leverage smart government policy to transform this asset into a global champion with huge benefits to the country. A detailed economic analysis indicates that a commercial-scale compound semiconductor fab could generate 10,000 new jobs, $10–15 billion in GDP over ten years, and create a 4–5x multiplier effect where every dollar invested produces four to five across the economy1. It would also supply critical components for Canadian defence, aerospace, and communications, while anchoring new startups in AI and quantum computing.
Currently there is no clear world leader in compound semiconductor manufacturing. And, while creating a modern silicon semiconductor mega-fab costs tens of billions of dollars, the top compound semiconductor fabrication facilities cost on the order of $500 million. It will take bold action but Canada has a unique opportunity to become part of the global photonics supply chain and benefit from the investments and expertise won in the late 1990s and early 2000s.
Global Leadership Examples
The Netherlands became a global semiconductor powerhouse through ASML, the sole producer of extreme ultraviolet lithography (EUV) machines. The Dutch government and universities partnered with Philips in the 1980s, co-investing in research and guaranteeing long-term public funding2. Over time ASML attracted global suppliers and talent, creating an innovation cluster that cannot be replicated elsewhere. Today, every advanced chip on Earth depends on ASML’s technology, giving the Netherlands extraordinary influence in the global supply chain.
Likewise Taiwan transformed itself into an essential part of the electronics supply chain through a strategic bet on Taiwan Semiconductor Manufacturing Company (TSMC). In 1987, the government invested nearly half of TSMC’s $220 million founding capital, provided land, and created favorable industrial policies3. TSMC pioneered the pure-play foundry model, where companies like Apple and NVIDIA design chips and TSMC fabricates them. Creating not only a global champion but an ecosystem of companies and subsidiaries that enriches the nation. The success created hundreds of thousands of jobs and made TSMC one of the most valuable semiconductor companies in the world.
What Must Be Done
Canada can learn from these examples to transform the know-how and capital locked up in CPFC into a world class company and ecosystem of compound semiconductor development. A shining opportunity to create a new batch of future global champions. The first key step is bringing in private partners and recapitalizing the fab following the successful examples with CN rail, and Petro-Canada4.
Commercialize the Canadian Photonics Fabrication Centre
To ensure that the CPFC has the resources and incentive structures to compete effectively with comparable fabs around the world it should be moved from public management to a private-public partnership. This means transferring authority from the NRC to a new commercial entity with private shareholders following the approach used to successfully commercialize CN Rail, Air Canada and other crown corporations. By bringing in additional investors the fab will have access to the private markets and be incentivized to expand operations. To safeguard national interests, the federal government should retain a stake and provide clear governance rules to prevent foreign control and ensure Canadian strategic autonomy.
Upfront government capital investment
As part of ensuring the fab is ready for full commercialization the government should provide an additional $350 million in capital to modernize and expand the facility extending the $115 million already committed5. This includes accelerating the transition to 4-inch wafer processing, finishing construction on the new 8,000 square-foot expansion, and upgrading to industrial-scale production. Funding should follow the model used for helping to develop EV manufacturing facilities, recapitalize CN rail at the point of privatization6, or as was used in the build out of the telecoms network in the 90s.
Encourage a complete ecosystem
The CPFC can be the anchor for a wider ecosystem connecting academia, industry, and government. This requires aligning NSERC, ISED, and Defence R&D programs to prioritize compound semiconductors, simplifying permitting for new photonics firms, and setting up a dedicated office within ISED for semiconductor coordination. Universities should receive new grants to train photonics engineers, and tax credits should be extended for Canadian firms using CPFC capacity.
Common Questions
Is privatizing CPFC just selling off a national asset?
No. The aim of commercializing the fab is to create a competitive fab facility that can help Canada become an essential part of the global photonics supply chain. The government would retain a stake and strategic veto powers to ensure strategic autonomy. The goal is to give CPFC commercial freedom to grow and become the backbone of the Canadian compound semiconductor industry while keeping control in Canadian hands.
Won’t this cost too much when budgets are tight?
The process of commercializing CPFC will bring with it new government revenue to offset costs. More importantly, a $350 million investment is small compared to the $10–15 billion GDP return projected over ten years not including the potential knock-on effects that could come from spurring more developments in the photonics ecosystem in Canada.
Isn’t the global market already dominated by others?
Canada doesn’t need to compete with silicon giants like TSMC. CPFC is unique in Indium Phosphide and photonics, technologies critical for AI, data centres, and defence. The unique history of Nortel and the lack of comparable global companies means this is a market niche where Canada can lead, not follow.
Conclusion
Canada has the talent, infrastructure, and history to lead in compound semiconductors. The problem is that for two decades, the CPFC has been left underfunded and underused while demand exploded globally. The path forward is clear: privatize the CPFC, invest upfront capital, and build a full ecosystem. With these steps, Canada can secure prosperity, create thousands of jobs, and become the trusted North American hub for photonics and compound semiconductors.
