The Commonwealth announced another $150 million in funding for semiconductor research and manufacturing capability last week. The goal: build Australia’s sovereign chip manufacturing capacity and reduce dependence on Asian supply chains.

The timing is interesting. Global chip demand has softened from the pandemic highs. Major fabs in Taiwan and South Korea are already planning capacity expansions. And the capital requirements for competitive semiconductor manufacturing continue to escalate.

The Strategic Case

There’s a legitimate national security argument here. Australia’s defence systems rely on semiconductors manufactured overseas, primarily in Taiwan and South Korea. If geopolitical tensions in the region escalate, supply chains could be disrupted.

Research from the Australian Strategic Policy Institute suggests that semiconductor supply chain resilience is a genuine capability gap. During the pandemic-era chip shortage, Australian manufacturers faced 6-12 month delays for components that were previously available within weeks.

The question isn’t whether semiconductor supply chain risk is real—it is. The question is whether local manufacturing is the right response.

The Economics Problem

Building a competitive semiconductor fabrication facility requires tens of billions in capital investment. TSMC’s latest fab in Arizona is costing over $40 billion. Samsung’s new facility in Texas carries a similar price tag.

Australia’s $150 million—even as part of a larger strategy—isn’t going to build cutting-edge chip manufacturing. We’re talking about trailing-edge processes, niche applications, and research capability.

That might still be worthwhile, but we should be clear about what we’re actually building. This isn’t sovereign capability to manufacture the chips that go into smartphones or data centre processors. It’s capability to manufacture simpler chips for specific applications.

What Actually Makes Sense

Several Australian companies are pursuing semiconductor design rather than manufacturing. This is smarter economically. Design requires talent and IP, but not tens of billions in capital equipment that becomes obsolete every few years.

Silicon Quantum Computing in Sydney is working on quantum computing chips. Several startups are designing specialised processors for AI and edge computing applications. These efforts could generate genuine IP value without requiring massive fab infrastructure.

There’s also opportunity in the packaging and testing segment. You don’t need cutting-edge fabrication capability to excel at chip packaging, but you do need precision manufacturing expertise—something Australia has in other industries.

The Skills Pipeline

Even if we built the fabs, we’d struggle to staff them. Australia graduates roughly 100 people annually with postgraduate qualifications in semiconductor engineering, according to Engineers Australia data.

Taiwan graduates thousands. South Korea graduates thousands. China graduates tens of thousands.

You can’t build a semiconductor industry without the talent pipeline to support it. That means decades of investment in university programs, vocational training, and research infrastructure. The $150 million in immediate funding needs to be paired with sustained commitment to education—which has historically been where these initiatives fall apart.

The Trade-Off

Money spent on semiconductor manufacturing is money not spent on other priorities. Could that $150 million generate more economic or strategic value if directed toward software expertise, cybersecurity capabilities, or biotechnology research?

Those fields have lower capital intensity and align better with Australia’s existing strengths. We’ve got strong universities producing software engineers and biotech researchers. The incremental investment to excel in those areas is lower than the investment to build semiconductor capability from scratch.

This doesn’t mean semiconductors are the wrong choice—just that the opportunity cost deserves consideration.

Regional Context

It’s worth noting that other nations are making similar moves. The US CHIPS Act allocated over $50 billion. The EU has committed similar amounts. Japan and South Korea are heavily subsidising their semiconductor industries.

Australia’s investment looks modest by comparison, which raises questions about whether we’re investing enough to matter or just enough to say we’re doing something.

What Could Work

If Australia is serious about semiconductor capability, a few approaches might be viable:

Focus on compound semiconductors (gallium nitride, silicon carbide) rather than competing in silicon CMOS where we’ll always be behind. These materials are important for power electronics, RF applications, and emerging technologies—and the industry leaders are less entrenched.

Build capability in specific defense applications where we need sovereign supply. Not consumer chips, not data centre processors, but the specialised components that go into military systems where supply chain security genuinely matters.

Invest heavily in semiconductor design and IP rather than manufacturing. Partner with existing fabs for production. This is how many successful semiconductor companies operate globally—they don’t own fabs, they design chips and contract manufacturing.

Create centres of excellence in specific aspects of semiconductor technology. Maybe we excel at chip packaging, or at testing methodologies, or at design tools. Being world-class at one piece of the value chain is more realistic than competing across the entire stack.

The Long Game

Semiconductor industry development is measured in decades, not years. Taiwan didn’t become a chip manufacturing powerhouse overnight—it was the result of sustained government support, private investment, and strategic focus over 40+ years.

If Australia is going to pursue semiconductor capability, it needs to be a sustained commitment spanning multiple government cycles. The history of Australian industrial policy suggests this is challenging. Initiatives get announced with fanfare, receive a few years of funding, then fade when priorities shift.

Measuring Success

How do we know if this investment is working? Metrics matter.

Number of semiconductor jobs created. Value of chip-related exports. Number of startups in the semiconductor design space. Patents filed. Graduates with semiconductor engineering qualifications.

Without clear metrics and public reporting, these initiatives become expensive press releases that quietly wind down when nobody’s paying attention anymore.

The Realistic Outcome

Most likely, Australia ends up with some research capability, a few small-scale manufacturing facilities handling niche applications, and hopefully a stronger ecosystem of semiconductor design startups.

That’s not failure—it’s just not the “sovereign chip manufacturing capability” that the press releases imply. If we’re clear-eyed about what’s achievable, we can make strategic choices about where to focus within the semiconductor value chain.

The worst outcome would be diffuse investment across too many priorities that collectively don’t add up to competitive capability in anything. Focus matters. Sustained commitment matters. Clear metrics matter.

We’ll see over the next few years whether Australia has the stomach for the long-term investment that semiconductor industry development actually requires.