The goal of the data-sharing and analytics effort is promoting collaboration among chipmakers and materials suppliers.
A technology partnership between the semiconductor materials arm of German multinational Merck and Palantir Technologies, the data analytics specialist, will tackle chip shortages via a collaboration platform aimed at chipmakers and their suppliers.
Merck KGaA, Darmstadt, Germany, and Denver-based Palantir said this week the partnership would leverage data analytics and AI to advance semiconductor manufacturing via a platform dubbed Athinia. The collaboration tool would connect fab operators with chemical and materials suppliers, allowing them to securely share and analyze fab data to improve efficiencies, thereby addressing current supply chain disruptions.
Along with the joint venture with Palantir, the Merck unit also said it would invest $1 billion in its U.S. operations through 2025 at facilities in Arizona, California, Pennsylvania and Texas. EMD Electronics, Merck KGaA’s North American business unit, will oversee the initiative aimed at the semiconductor and display sectors.
“The chip shortage needs industry-wide cooperation to resolve the supply chain issues consumers are currently facing,” Kai Beckmann, CEO of the German firm’s electronics unit, said in a statement. “We are investing in the U.S. to expand our production capacity….”
Athinia is billed as a secure data analytics platform based on Palantir’s Foundry operating system aimed at large enterprises. The new platform would crunch data on IC manufacturing supplies ranging from gases and deposition materials to lithography and photoresist chemicals. The goal is to help rationalize supply chains as U.S. chipmakers seek to ramp IC production.
Chipmakers are tightfisted when it comes to sensitive data on chip yields, for example. Palantir’s Alec McShane, who will serve as executive vice president of Athinia, said there would be “no IP contamination” on its secure platform.
Rather, the data analytics platform would leverage encrypted, anonymized data on fab operations, combining it with supplier information to help U.S. chipmakers ramp leading-edge 7-nm process nodes. The partners are emphasizing collaboration around materials as a way to boost fab capacity utilization while bringing badly needed new capacity online, addedLaura Matz, Athinia’s CEO and chief science and technology officer at Merck KGaA.
Shyam Sankar, Palantir’s chief operating officer, will also lead the partnership. Merck previously formed a healthcare joint venture with Palantir.
Athinia would be used to integrate databases covering raw materials, batch data and manufacturing timelines, Matz said in an interview. Palantir’s data analytics technology is touted as handling disparate data sources ranging from structured, unstructured and time-series. Those data sets are analyzed to discern patterns that can be used to boost manufacturing efficiencies.
Fab data are anonymized and scaled to create what the partners assert is a secure flow of information between suppliers and fab operators.
Along with data encryption, Athinia’s cloud infrastructure complies with “military-grade” data security, access controls such as “granular permissions” along with infrastructure hardening in the form of certifications, encrypted network traffic and audit logs.
Along with big data analytics, the collaborative platform would leverage AI models to help chipmakers and materials suppliers share, aggregate and analyze data on materials and manufacturing. The goal is identifying efficiencies based on insights into the nexus between raw materials and fab processes.
For example, the collaboration platform could be used to analyze materials requirements along with “in-fab” quality performance data, the partners said. One result would be improved materials quality to boost chip yields and overall fab performance.
The data analytics initiative emphasizes “the sensitivity of materials quality and device manufacturing” as a way to reduce defects, Matz said.
The joint ventures with Merck are among the latest for high-flying Palantir. The company went public last year on the strength of government contracts and joint ventures covering areas such as insurance and fraud detection.
Palantir has also drawn scrutiny for its data mining technologies supplied to several U.S. military branches, including U.S. Special Operations Command. It also provides data integration tools to the U.S. Space Force, among a growing list of other U.S. government customers that also includes U.S. immigration services.
The partners declined to identify potential customers for the collaboration platform, saying only they are targeting “leading-edge chipmakers.”
This article was originally published on EE Times.
George Leopold has written about science and technology from Washington, D.C., since 1986. Besides EE Times, Leopold’s work has appeared in The New York Times, New Scientist, and other publications. He resides in Reston, Va.