Tokyo, September 18, 2026 – Preferred Networks, Inc. (PFN) and Preferred Computing Infrastructure, Inc. (PFCI) today announced plans to offer computing resources powered by PFN’s proprietary AI chip, MN-Core™ 2, within 2026 for selected calculation tasks on Matlantis™, a universal atomistic simulator for materials discovery marketed by a PFN group company Matlantis Corporation. 
Under the plan, MN-Core 2’s computing resources will be provided via PFCI’s AI cloud service, Preferred Computing Platform™ (PFCP™), to Matlantis. This will enable all Matlantis users to select MN-Core 2 as their computing resource for eligible high-speed simulations—such as crystal structure calculations—achieving up to a fourfold speedup over GPUs, all without changes to their existing operating environments.

Matlantis is an AI-accelerated cloud service that simulates the physical properties of arbitrary combinations of 96 elements up to tens of millions of times faster than conventional methods. Matlantis relies on Matlantis PFP, a foundation model (machine learning interatomic potential) developed by PFN. The service is currently used by more than 150 companies and academic institutions worldwide to discover novel materials for next-generation batteries, catalysts, semiconductors and more.

While Matlantis PFP is currently powered by GPUs, testing has confirmed that MN-Core 2 can simulate structures such as battery material crystals up to four times faster than GPUs while maintaining accuracy. MN-Core 2 features a higher local memory capacity per compute unit and greater flexibility in software control compared to GPUs, enabling significant speed improvements for simulation workloads for specific structures. PFN is currently developing the capability for release within 2026, following trial evaluations with select Matlantis users to verify practical efficacy.

MN-Core 2, a proprietary AI chip developed by PFN

By developing the entire AI technology stack in-house—from AI chips and computing infrastructure to foundation models and products/solutions—PFN has consolidated expertise across every layer. While PFN flexibly combines proprietary and third-party technologies to deliver optimal solutions for customer needs, this combination of in-house PFN group technologies was seen as ideal for Matlantis users, prompting the integration of MN-Core 2’s computing power into the platform.

About Preferred Networks
Guided by its mission to “Make the real world computable and create the future together,” Preferred Networks, Inc. (PFN) develops advanced software and hardware technologies through a vertically integrated approach spanning the entire AI value chain—from AI chips and computing infrastructure to generative AI foundation models and solutions across a wide range of industries. Founded in 2014 in Tokyo, PFN currently develops and delivers the MN-Core™ series of AI processors, the PFCP™ cloud platform for AI computing, and the Japan-developed generative AI foundation model PLaMo™. 

About Preferred Computing Infrastructure
Preferred Computing Infrastructure, Inc. (PFCI) builds high-efficiency computing infrastructure powered primarily by Preferred Networks (PFN)’s MN-Core™ series AI accelerators and provides it as the AI cloud service Preferred Computing Platform™ (PFCP™). PFCI was established in 2025 as a PFN Group company.