National

What the 2026 Semiconductor Shortage Means for Australian Manufacturers

Shivangi

Jul 31, 2026 5:54 PM IST

Category National

“

Synopsis

Australian manufacturers are once again facing semiconductor supply challenges in 2026 as AI-driven demand for memory chips tightens global availability. Here’s how the shortage is impacting production costs, delivery timelines, and supply chains, and the steps businesses can take to stay ahead.

Australian manufacturers are bracing for a new round of chip shortages in 2026, but this time it is not through factories closing. Memory chips are being diverted away from consumer products and put into AI data centres. The shortfall is already raising prices and may soon extend delivery times for manufacturers dependent on imported parts, just as the pandemic chip crunch did.

Why There Will Be A Shortage In 2026 

Chip makers such as Samsung, SK Hynix and Micron are moving their manufacturing lines to produce memory that is more likely based on AI data centres because that is where the greatest return on investment now comes. This reduces the amount of headroom for regular memory chips that regular manufacturing relies upon. Several chip makers have already stated publicly that they lack visibility into when memory supply will return to balance with demand, and a handful of them indicated that relief is unlikely until 2028.

S&P Global and other analysts believe DRAM prices can increase by 70 to 100 per cent during 2026 compared with 2025. More industry data show increases of a similar magnitude, with some reports indicating a rise in memory contract prices exceeding 50 per cent within just one quarter. It is also not only about semiconductor factories. The aforementioned memory crisis is merely a small part of it, but raw materials including helium, bromine and copper used in chip manufacture are also tight.

For perspective on how big a change it was, demand for data centres accounted for only about one-fifth of all global DRAM consumption in 2025, up considerably from around a third as recently as five years earlier. And that share will rise in tandem with the growing scale of AI systems, which means the chips being yanked out from underneath manufacturers aren’t a short-term blip in the supply chain, they’re being funnelled into long-term supply deals between chip makers and some of the biggest tech firms.

Why Is This Shortage Different From The Pandemic One?

This chart shows how broad and deep the shortage was through 2020 to 2021, fueled by factory shutdowns, shipping delays and a surprising spike in consumer demand for electronics as people stayed home. It sent almost every kind of chip into scarcity at the same time and faded away as factories caught up and shipping routes reopened.

This shortage is more targeted. It is focused on memory chips and advanced packaging, and it is being powered by long-term supply contracts in which AI firms secure production capacity for years (not months), nowhere near a “disruption” of the sort that might normally numb demand. All of the world’s DRAM output is in effect controlled by just three companies, and they have all been steadily moving production out of generic memory chips and into specialised memory used in AI systems.

How long this is likely to last makes a difference. Chip makers have said they do not see a serious reprieve until new memory production capacity is brought to market, with industry estimates placing this no sooner than 2027 and some even warning that the pressure could last well into 2028, While the pandemic shortage was more connected to global logistics returning to normal, this one is linked to a real structural change in what chip makers build most profitably.

What This Could Mean For Australian Manufacturers

The bulk of semiconductor supply for Australian manufacturers, even those in automotive, electronics and industrial equipment, is imported with precious little local buffer available should global supply start to tighten or be diverted to other areas. In the last shortage, local makers saw lead times stretch from weeks to six or twelve months, and such delays are a real risk of repeating themselves this time around.

Modern cars are particularly vulnerable, as just one vehicle can have thousands of chips controlling everything from electric seats to advanced driver-assistance systems that brake or steer the car away from danger. Memory prices go up, and that eventually makes its way to the top-line price of the car or chunk of machinery; in some previous shortages, producers have needed to leave features out rather than postpone whole product families, biting off functionality instead of delivery.

Electronics and industrial equipment makers feel a similar squeeze, because memory chips are packed into a lot more products than most realise. Overseas some manufacturers are already telling me that memory is now a meaningfully larger fraction of their total production costs than even just one quarter earlier, and there is no reason Australian manufacturers directly importing similar components would be immune from the same cost pressure.

Where LED semiconductor investment really is

Due to the unique composition of the semiconductor supply chain (which does not include Fabs in Australia), Australia is very reliant on overseas trade and currently has no Fabs, so it could never manufacture its own chips at scale regardless of how this shortage develops. The vast majority of local capability resides within university research and CSIRO and is a long way off from being enough to allow for production manufacturing.

Federal funding for local capability is in the National Reconstruction Fund, roughly $1 billion of which has been earmarked on critical technologies with semiconductors a priority area. So far this has included a $60 million investment in Silicon Quantum Computing, and $10.1 million (later expanded to $37 million) in the Sydney-based chip connectivity firm Syenta. South Australia, for example, has also topped up its own funds with a semiconductor research hub.

This, however, is real but small when compared with the hundreds of billions of dollars agreed by the US, EU, Japan and South Korea to support their domestic semiconductor industries. At the end of the day, Australia’s best prospects don’t lie in competitive mass chip production. They’re in chip design, packaging and testing; they’re in compound semiconductors, gallium nitride, silicon carbide, fields where local competency can play at a level competitive with nations that also need fabrication plants.

What Manufacturers Should Be Watching Out For Next

Component price hikes are expected to continue spilling through to Australian manufacturers, and ultimately consumers well into 2026 as automakers and electronics producers externalise rising input costs. Easing isn’t expected until fab capacity from new memory starts coming online from 2027 onwards, so those pushing out components should plan on a runway longer than what manufacturers are used to for standard supply timelines.

We recommend keeping an eye on our design, packaging and testing sector as this is probably the only part of the industry which could sustain local growth versus higher-volume manufacturing relatively easily dominated by a handful of countries with significantly deeper pockets. Given the external sourcing of components, businesses will want to lock in longer-term bilateral supply paths rather than wait until scarcity bites harder and options reduce.

And that may be worth watching to see whether government funding announcements become a real, long-term industrial strategy, as serious ability in chip design and packaging takes years of cumulative spending to build, not one budget cycle. The Australian Strategic Policy Institute’s critical technology tracker is a useful way to monitor what happened as this plays out, and how Australia compares internationally.

The Bottom Line

Rather, this isn’t a one-off shock instead it appears to be a structural reallocation of the world’s chip-manufacturing capacity. There is no one bottleneck to clear or shipping lane to reopen, unlike the pandemic-era crunch. The source of the pressure is that AI companies are booking years’ worth of supply upfront and market cycles in economies usually tend to fix this and it did indeed fix the last shortage but this time it won’t be fixed by a typical market cycle.

That means manufacturing input costs and inputs to Australian manufacturers with longer lead times are likely to stay here for an extended period, probably well into 2027. Companies that incorporate this into their planning now, whether locking in supply agreements earlier or redesigning products to mitigate chip dependency, or just accounting for costlier components in forecasts. will be better off than those waiting on supply chains to return to the way they were after COVID eventually did.