Problem

Atum Works is targeting a fundamental bottleneck in semiconductor manufacturing: conventional lithography etches flat, two-dimensional patterns onto silicon, while modern chips increasingly need greater density, wiring, and three-dimensional integration. That approach is approaching physical limits, making it harder to continue increasing computational power—especially for demanding AI hardware applications.

The economic promise is to replace or augment expensive, inflexible 2D fabrication with a 3D process that can make complex structures more cheaply and flexibly. The company claims its approach could make some components more than 10,000 times cheaper in low-volume production and roughly 10 times more affordable at scale. The clearest killer use case is printing three-dimensional semiconductor structures and interconnects directly inside chips rather than fabricating them through increasingly complex stacks of etched layers.

Product / Service

Atum Works is developing what it describes as the world’s first scalable 3D lithography foundry. Its process is intended to fabricate arbitrary metal structures with submicron resolution; the company and Y Combinator describe multi-material structures at 100-nanometer resolution across wafer-scale areas. In practical terms, the platform uses additive, lithographic manufacturing to place material in three dimensions instead of merely patterning a flat surface.

The likely delivery model is a specialized manufacturing and technology platform, with the foundry producing advanced microstructures and early products potentially incorporating the company’s printer technology. The benefit is a route to more three-dimensional chip architectures, greater wiring and computational density, and potentially lower costs and more design flexibility than conventional lithography. The exact commercial mix between foundry services, equipment, and shipped products has not been publicly detailed.

Market

Atum Works sits at the intersection of hard tech, semiconductor manufacturing, nanomanufacturing, and advanced 3D printing. Its principal incumbent alternative is the established 2D lithography ecosystem, including tools from ASML, while companies such as 3D Systems represent an adjacent additive-manufacturing competitor category rather than a clearly identified direct rival. Atum Works’ differentiation is the combination of wafer-scale processing, nanoscale resolution, multiple materials, and genuinely three-dimensional structures.

The company was founded in 2024, entered Y Combinator’s Spring 2025 batch, and is listed as active with a small team. A 2025 third-party estimate puts revenue at approximately $330,000, so the company should not be treated as pre-revenue, although that figure is not an audited company disclosure. A trade-publication report also says Atum Works had raised more than $10 million, built a working printer prototype, partnered with NVIDIA, and was preparing early products; however, another database reports zero funding, so financing and commercial traction remain publicly inconsistent and should be viewed as preliminary rather than fully verified.