)
After developing a novel way to build physical connections inside semiconductor chips, allowing for denser and more precise wiring that helps processors keep up with the size and complexity of AI computing, startup Syenta has raised $8.8 million in pre-Series A funding to commercialise its technology.
The round was led by Investible, with participation from Blackbird Ventures, In-Q-Tel, SGInnovate, OIF Ventures, Salus Ventures, Jelix Ventures, Wollemi Capital Group and Brindabella Capital.
Spun out from the Australian National University (ANU), Syenta has developed a transformative approach to semiconductor manufacturing with its proprietary Localised Electrochemical Manufacturing (LEM) technology, co-invented by its CEO and co-founder Dr Jekaterina Viktorova.
The novel LEM approach enables micron-scale resolution in advanced semiconductor packaging, offering a scalable solution to the pressing memory bandwidth bottlenecks in AI computing whereby more data needs to be moved between memory and processors.
Memory bandwidth relates to the speed at which data can move, and it hasn't kept up with how fast processors have become - an issue that impacts AI performance.
In the last 20 years, processor speeds have improved about 60,000 times, but memory bandwidth has only improved 100 times, and the connections (called interconnects) between components have only improved 30 times.
This is where the opportunity lies for Syenta as its LEM technology enables memory and processors to sit closer together, exchanging data more quickly and efficiently.
"LEM is the foundation for a new generation of chip packaging," says Viktorova.
"It offers the scale, performance and manufacturability needed to overcome the critical ‘memory wall’ throttling AI systems today."
What makes LEM different is that it doesn’t use the traditional, expensive methods like photolithography which the company claims are too slow and complex for today’s needs.
Syenta's process uses 12 manufacturing steps instead of the 23 used in the current semi-additive process, and the company notes it cuts down on waste and emissions.
Originally developed in Australia, Syenta’s technical credibility was recognised on a global scale in 2024 when it was selected for ASTRA, the flagship accelerator by Applied Materials. The selection followed a successful feasibility project with Applied Materials, the world’s largest semiconductor company.
"Being selected for ASTRA is a major milestone for Syenta - it puts us shoulder-to-shoulder with the industry’s most disruptive innovators and gives us direct access to leaders in the electroplating world, global foundries, OSATs (Outsourced Semiconductor Assembly and Test) and materials giants," Viktorova explains.
"It validates LEM as a serious platform for global semiconductor manufacturing."
The company reports significant traction with its LEM process, and is engaged with early access partners across the semiconductor supply chain.
"Syenta is a key technology enabler for the next generation of AI infrastructure and high-performance computing (HPC) chips, particularly as semiconductor chip fabrication levels are unable to keep up with current demand for generative AI infrastructure," says Nicholas Ooi, lead investor at Investible.
"The team’s deep academic and materials development experience, together with their patented LEM technology, positions Syenta to become a world-leader in semiconductor advanced packaging."
)
