WEST LAFAYETTE, Ind., March 02, 2026 -- Electronics manufacturers could benefit from patented and copyrighted Purdue University simulation engines that inexpensively and quickly model semiconductors scaled as small as 2 or 3 nanometers.

The engines, collectively called the Quantum Code Library, have been created and developed since 2010. These modeling tools are composed of more than 600,000 lines of code contributed by hundreds of researchers, including university faculty and their graduate students.

Tillmann Kubis, the Katherine Ngai Pesic and Silvaco Associate Professor of Electrical and Computer Engineering in Purdue’s Elmore Family School of Electrical and Computer Engineering, leads the development of the Quantum Code Library.

“The library currently powers several simulation engines including NEMO5, which is commercialized by Silvaco Group Inc. as Victory Atomistic,” he said. “We have collaborated with Silvaco on the library for several years.”

Purdue researchers have disclosed the Quantum Code Library to the Purdue Innovates Office of Technology Commercialization, which has received patents and registered copyrights to protect the intellectual property.

The Quantum Code Library is organized so software packages can be created for targeted customers and their applications, meaning new nonexclusive licenses can be issued. OTC also is exploring providing access to library modules on a software-as-a-service basis.

Industry-driven development sets Purdue library apart

Kubis said Purdue’s Quantum Code Library, like other high-performing simulation tools developed worldwide, focuses on atomistic simulations: modeling the charge, heat or spin transport and properties of nanodevices with a subatomic resolution.

“What sets the Purdue library apart is that much of our development and funding was a result of projects with industry-leading companies, including several in the Fortune 500,” he said. “Most of the time we were an extended workbench to those companies. This very close interaction guaranteed every physical effect, device geometry and material aspect we included answered industrial questions within less than a month from the release date.”

The problem of a single atom

Kubis said next-generation nanotechnology demands semiconductors that are ultrasmall yet high performing and resilient. Precision modeling, therefore, is a requisite. He said a 2-nanometer semiconductor has about 16 atoms of silicon.

“If the design has one more or fewer atom than the average, the entire system can change nonlinearly beyond 10%,” he said. “When you have a computer chip with trillions of transistors, you need them to behave the same. If some have one atom more or one atom fewer, the transistors behave differently and the chip won’t work.”

Atomistic simulations can address the problem of a single atom, but Kubis said a common perception is that the simulations are expensive. But the Quantum Code Library allows users to model nanoscale electronics quickly and inexpensively.

Kubis said atomistic nanodevice simulations conducted in the 2010s could easily cost millions of CPU hours, which typically required Purdue researchers to get project allocations on national supercomputers. Handling atomic-scale deviations from the ideal device structure with reasonable statistics of several hundred samples was simply not feasible.

“During the past few years, our team at Purdue has developed systematic and automatized spectral transformations that reduce the numerical costs by many orders of magnitude without sacrificing the atomic accuracy and prediction precision,” he said. “We can run the same simulations of the 2010s and far more complicated ones so fast, I did not have to apply for supercomputer allocations for many years.”

Continually enhancing the library

Kubis said the Quantum Code Library user experience is being further refined, including enhancing its code framework, called Rhino.

“The Rhino framework allows for the customization of libraries,” he said. “It allows users to streamline the flexibility and broad applications available into easy-to-use tools for specific customized application scenarios.”

Kubis said further development is underway to interface the Quantum Code Library with other tools engineers use.

“Semiconductor engineers are well versed in nonatomistic technology computer-aided design, or TCAD, tools. Rhino allows the Quantum Code Library to feed atomistic information into TCAD, effectively creating atomistic TCAD,” he said. “In a similar way, we interface material and physical chemistry tools such as density functional theory models with our open-boundary quantum simulators. This allows us to solve interface and chemical reaction problems with millions of atoms on a small computer cluster.”

Quantum research is a component of Purdue Computes, a comprehensive university initiative that emphasizes four key pillars of Purdue’s extensive technological and computational environment — computing departments, physical AI, quantum science and semiconductor innovation.