{"url":"https://newsforagent.com/news/2026-11-16/semifive-pfn-mn-core-l2000-chips","markdown_url":"https://newsforagent.com/news/2026-11-16/semifive-pfn-mn-core-l2000-chips.md","title":"SEMIFIVE and PFN to Build MN-Core L2000 AI Inference Chips","headline":"SEMIFIVE and Preferred Networks Team Up on MN-Core L2000 AI Inference Chips","description":"SEMIFIVE and Preferred Networks will develop the MN-Core L2000 AI inference chip series, targeting over twice the memory bandwidth of the L1000.","section":"Partnerships","published_at":"2026-11-16T22:24:09+00:00","modified_at":"2026-11-16T22:24:09+00:00","author":"News Desk","key_takeaways":["Preferred Networks and SEMIFIVE announced a collaboration to develop the MN-Core L2000 Series of AI inference chips.","The new series is planned to deliver more than twice the memory bandwidth of the MN-Core L1000 Series, with tape-out scheduled for the first quarter of 2027.","The logic chip will be manufactured using Samsung Foundry's platform technology, and the companies will jointly exhibit the series at SC26 in Chicago from November 17 to 19, 2026."],"companies":[{"name":"SEMIFIVE","is_primary":true,"profile_url":"https://akta.pro/company-profiles/0000k7v/semifive","website":"https://www.semifive.com"},{"name":"Preferred Networks","is_primary":false,"profile_url":"https://akta.pro/company-profiles/0000co3/preferred-networks","website":"https://www.preferred.jp"}],"source":{"publisher":"PR Newswire","title":"PFN and SEMIFIVE Collaborate to Develop the MN-Core L2000 Series of Next-Generation AI Inference Chips","url":"https://www.prnewswire.com/news-releases/pfn-and-semifive-collaborate-to-develop-the-mn-core-l2000-series-of-next-generation-ai-inference-chips-302940559.html"},"other_sources":[],"tags":["Partnerships","SEMIFIVE","Preferred Networks"],"body_markdown":"Preferred Networks and SEMIFIVE announced a collaboration on November 16, 2026, to develop the MN-Core L2000 Series of AI inference chips, planned to deliver more than twice the memory bandwidth of the MN-Core L1000 Series, with tape-out scheduled for the first quarter of 2027. The companies announced the project in a [PR Newswire](https://www.prnewswire.com/news-releases/pfn-and-semifive-collaborate-to-develop-the-mn-core-l2000-series-of-next-generation-ai-inference-chips-302940559.html) release.\n\n## How the MN-Core L2000 Series Is Built\n\nThe MN-Core L2000 Series is the successor to the MN-Core L1000 Series, which is scheduled for release in 2027. The L Series stacks logic chips built on the MN-Core architecture vertically with DRAM in a 3D structure, which the companies say achieves substantially higher memory bandwidth than GPUs and other AI processors. Preferred Networks handles processor architecture and logical design, while [SEMIFIVE](https://akta.pro/company-profiles/0000k7v/semifive) provides design implementation and supply chain management through its 3D-IC design solution. The logic chip will be manufactured using Samsung Foundry's platform technology, with Samsung also supplying 3D-IC design infrastructure for die connectivity and power structure optimization.\n\n\"Developing AI chips with a 3D-stacked structure presents more complex technical challenges than conventional chip development, particularly in areas such as power consumption, heat dissipation, and manufacturing yield,\" said Daisuke Okanohara, Co-Founder and CEO of [Preferred Networks](https://akta.pro/company-profiles/0000co3/preferred-networks). \"With SEMIFIVE's proven track record in turnkey solutions for custom semiconductors, we aim to tackle these technical hurdles.\"\n\n\"The MN-Core L2000 project with PFN represents a monumental milestone as SEMIFIVE's first full-scale 3D-IC initiative,\" said Brandon Cho, CEO and Co-Founder of SEMIFIVE.\n\n## SEMIFIVE's Recent 3D-IC and Turnkey Work\n\nAccording to akta.pro [company records](https://akta.pro/company-data), SEMIFIVE showcased 3D-IC and Big Die design solutions at the Samsung SAFE Forum 2026 in San Jose in May 2026, including the jointly developed HyperAccel Bertha LPU-based AI inference accelerator on Samsung's 4nm SF4X process. In March 2026, the company signed a KRW 18 billion turnkey design contract for a next-generation AI NPU on Samsung 4nm targeting on-premise real-time AI processing. In February 2026, it secured a KRW 10 billion design-win contract with U.S.-based Niobium to develop a fully homomorphic encryption accelerator on Samsung 8nm.\n\n## Earlier SEMIFIVE Orders and Production\n\nSEMIFIVE reported KRW 178.4 billion in third-quarter 2026 new orders, up 437% year-over-year, pushing its order backlog to a record KRW 245.5 billion, according to PR Newswire. In September 2026, SEMIFIVE began mass production of AI inference accelerators for data centers from HyperExcel using Samsung Foundry's 4nm process, Nate reported. In August 2026, SEMIFIVE reported a 97% year-over-year revenue surge to KRW 94.7 billion in the first half of 2026, PR Newswire reported.\n\n## SC26 Exhibition Dates\n\nPFN and SEMIFIVE will jointly showcase the MN-Core L2000 Series at PFN's Booth 4606 during SC26 in Chicago, Illinois, from November 17 to 19, 2026.\n","publisher":{"name":"News for Agent","url":"https://newsforagent.com"}}