# ATLANT 3D launches NANOFABRICATOR PRO to build AI-designed materials

> Source: <https://runtimewire.com/article/atlant-3d-nanofabricator-pro-ai-materials-fabrication>
> Published: 2026-08-17 10:56:11+00:00

[ATLANT 3D](https://atlant3d.com/?ref=runtimewire) founder [Maksym Plakhotnyuk](https://dk.linkedin.com/in/maksym-plakhotnyuk?ref=runtimewire) launched NANOFABRICATOR PRO on August 17, betting that the next valuable layer in AI materials discovery will be the machinery that turns a model's predictions into physical samples and prototype devices.

The Copenhagen-based manufacturer said in [its launch announcement](https://www.prnewswire.com/news-releases/atlant-3d-launches-nanofabricator-pro-the-worlds-first-and-only-physical-platform-for-ai-driven-materials-discovery-302851741.html?ref=runtimewire) that the system combines programmable atomic-scale fabrication, experimental validation and device prototyping. ATLANT 3D calls it the world's first and only physical platform for AI-driven materials discovery, a distinction unsupported by public comparative benchmarks.

Plakhotnyuk has spent eight years working toward this product. During his nanotechnology doctorate at the Technical University of Denmark, he found conventional semiconductor equipment slow, rigid and expensive for rapid experimentation. He founded ATLANT 3D in 2018 with Ivan Kundrata and thin-film materials professor Julien Bachmann to commercialize Direct Atomic Layer Processing, or DALP, a method designed to deposit materials selectively without masks or conventional lithography constraints.

"The next frontier is turning those discoveries into real-world innovations," Plakhotnyuk said in the announcement. That sentence captures the commercial bet behind PRO: AI can generate candidate materials faster than many laboratories can fabricate and test them, creating a new bottleneck in the physical workflow.

### From a doctoral frustration to industrial hardware

Plakhotnyuk arrived at ATLANT 3D through research in nanoelectronics, quantum devices and semiconductor fabrication. A [Fulbright program record](https://fulbright.org.ua/wp-content/uploads/2020/04/yearbook_2010_11.pdf?ref=runtimewire) identifies him as a researcher from Vinnytsia National Technical University who worked at the University of Illinois Chicago, while an [industry program biography](https://efds.org/wp-content/uploads/2020/12/EFDS_WSALD_02-03-12-2020_VL8.pdf?ref=runtimewire) records his later doctorate and postdoctoral work at DTU.

In a [2025 founder post](https://www.linkedin.com/posts/maksym-plakhotnyuk_in-2018-i-founded-atlant-3d-in-copenhagen-activity-7340278030252060675-kbNZ?ref=runtimewire), Plakhotnyuk said NASA became ATLANT 3D's first customer, Sony its first investor and ATLANT 3D had raised over $32 million. ATLANT 3D's [investor page](https://atlant3d.com/investor-relations/?ref=runtimewire) also names West Hill Capital, Sony, the European Commission and Innovation Fund Denmark as backers. Funding totals reported elsewhere vary, so Plakhotnyuk's figure remains the clearest attributed total.

ATLANT 3D delivered a NANOFABRICATOR 0G system to NASA in 2022 and introduced NANOFABRICATOR LITE commercially in 2023, according to ATLANT 3D's investor timeline. PRO moves the product line toward industrial research and small-batch prototyping, with specifications intended to support larger substrates and more complex material stacks.

### What NANOFABRICATOR PRO adds

ATLANT 3D's [NANOFABRICATOR PRO product overview](https://atlant3d.com/nanofabricator-pro/?ref=runtimewire) lists support for as many as six materials, samples up to 200 millimeters, line widths of 100 micrometers and deposition speeds up to 500 millimeters per second. Those figures are ATLANT 3D specifications rather than independently tested performance results.

The distinction from NANOFABRICATOR LITE is capacity. ATLANT 3D lists LITE as supporting two materials, samples up to 100 millimeters and speeds of up to 200 millimeters per second. LITE targets academic and research laboratories, while PRO is positioned for semiconductor, advanced packaging, quantum and industrial R&D work.

Direct Atomic Layer Processing uses localized microreactors to deposit materials along programmed paths. ATLANT 3D says its DALP stack can work with over 450 materials and reduce chemical use and waste, though that material count and the claimed efficiency gains have not been independently benchmarked. The strategic value lies in making deposition programmable: a materials model can propose a structure, the fabrication system can create it, and measurement tools can feed the result into the next computational cycle.

ATLANT 3D is working with Automated Industrial Robotics Silicon Valley to manufacture PRO in the United States. AIR Silicon Valley, the Fremont, California operation previously known as Owens Design, [specializes in customized manufacturing equipment](https://industrialrobotics.com/wp-content/uploads/2025/10/Enterprise-Sales-Manager-Job-Posting.pdf?ref=runtimewire). ATLANT 3D says the resulting system is SEMI-compliant, giving Plakhotnyuk a U.S. production base and a route into semiconductor customers that expect equipment built around established industry requirements.

The U.S. partnership also addresses a practical problem for a European hardware developer. Selling capital equipment requires installation, service, replacement parts and manufacturing repeatability near major customers. AIR gives ATLANT 3D an industrialization partner while Plakhotnyuk's group concentrates on deposition processes, materials workflows and software.

### The physical layer of AI materials discovery

AI materials research is drawing well-capitalized competitors that approach the problem from different layers. [CuspAI](https://cusp.ai/?ref=runtimewire) is assembling an AI Materials Foundry across model developers, data providers, industrial groups and laboratories, and lists ATLANT 3D among its laboratory partners. [Lila Sciences](https://www.lila.ai/tech?ref=runtimewire) is building AI Science Factories where models propose, run and learn from automated experiments across materials, chemistry and life sciences.

ATLANT 3D's opening is narrower and hardware-heavy. Plakhotnyuk is building the fabrication and validation machinery that model companies and autonomous labs will need if their predicted materials are to become working devices. NANOFABRICATOR PRO is therefore both a product and a bid to become infrastructure for other AI materials platforms.

There is an early sign of demand, although ATLANT 3D has kept the customer anonymous. On July 16, ATLANT 3D [said a global AI hyperscaler had ordered a NANOFABRICATOR LITE](https://www.prnewswire.com/news-releases/a-leading-global-ai-hyperscaler-selects-atlant-3ds-nanofabricator-platform-for-setting-ai-driven-materials-discovery-lab-302826293.html?ref=runtimewire) for a materials discovery laboratory. That order involved the smaller LITE system; it supports Plakhotnyuk's thesis that AI developers need physical experimentation capacity alongside compute and models.

PRO's next test will happen inside customer laboratories. ATLANT 3D has yet to publish independent results demonstrating the complete loop from an AI-generated material prediction through fabrication, measurement and a functional device at production quality. Price, delivery timing and an initial PRO customer will also shape whether the platform becomes shared infrastructure or remains specialized research equipment.

Plakhotnyuk's bet is grounded in a constraint he encountered before the current AI materials boom: digital design can move only as quickly as researchers can build and test its output. NANOFABRICATOR PRO gives ATLANT 3D a larger machine for that bottleneck and a U.S. partner capable of manufacturing it. The commercial case will depend on whether customers can turn those capabilities into repeatable experimental cycles, then carry the winning materials into production.
