Most people think AI’s big constraint is computing power. Syenta thinks they’re looking at the wrong bottleneck.

The Sydney-based semiconductor startup is focused on the connections between chips: the tiny pathways that move data from memory to processor. It’s a part of the AI supply chain that rarely makes headlines, but one that’s increasingly setting the pace on AI performance globally.

What Syenta does

Syenta has developed a proprietary manufacturing technique called Localised Electrochemical Manufacturing (LEM). In plain terms: it builds extremely fine metal interconnects: pathways up to 100 times thinner than a human hair: between chips, allowing data to move much faster than existing methods allow.

The problem they’re solving is known as the “memory wall.” Processor speeds have improved roughly 60,000 times over the past 20 years. Memory bandwidth: the speed at which data moves: has improved only about 100 times. The connections between components have improved just 30 times. That gap is now one of the primary constraints on AI performance.

Syenta’s LEM technology targets that gap directly: faster chip-to-chip connections mean AI systems can process data more efficiently without waiting on memory bottlenecks.

Company background

Syenta was spun out of the Australian National University (ANU) in 2022. The founding team is PhD-led:

Dr Jekaterina Viktorova. Founder and CEO
Professor Luke Connal. Co-founder, materials science
Ben Wilkinson. Co-founder and CTO, photovoltaics and applied physics
Zachary Dowse. Co-founder and COO, neuroscience background

The company is headquartered in Sydney, with a second office in Tempe, Arizona: positioning it to work with US chip designers and manufacturers including potential licensees like TSMC, AMD, and Nvidia supply chain partners.

Funding and investors

In April 2026, Syenta raised US$26 million in Series A funding. The round was led by Playground Global and included a AU$10.1 million investment from the National Reconstruction Fund Corporation (NRFC): the Australian Government’s industrial investment arm, which cited the strategic importance of anchoring semiconductor manufacturing capability in Australia.

Former Intel CEO Pat Gelsinger joined Syenta’s board as part of the raise: a notable endorsement from one of the most experienced figures in global semiconductor manufacturing.

Prior to the Series A, Syenta had raised A$8.8 million in earlier rounds.

What it means for Australian AI

Syenta sits at an unusual intersection: a deep tech hardware company, founded in Australia, working on infrastructure that underpins global AI development. It’s not building AI products for end users: it’s building the physical infrastructure that makes AI faster and more efficient at scale.

The NRFC investment signals that the Australian Government views semiconductor capability as strategically important: consistent with the Anthropic MOU and the broader National AI agenda announced in early 2026. The goal is to ensure Australia isn’t purely a consumer of AI infrastructure built offshore.

Syenta is targeting high-volume production by 2028. If the technology scales as intended, Australian-developed chip packaging could eventually be licensed to major global manufacturers for use in next-generation AI and quantum chips.