SoftBank’s subsidiary, Saimemory, has fashioned an alliance with Intel to advance the commercialisation of Z-Angle Reminiscence (ZAM), a next-generation reminiscence expertise.
This collaboration goals to deal with the rising calls for for enhanced computational efficiency in AI and high-performance computing sectors by providing improved reminiscence capability and bandwidth whereas decreasing energy consumption.
Entry deeper trade intelligence
Expertise unmatched readability with a single platform that mixes distinctive knowledge, AI, and human experience.
The timeline set by each events plans for operational graduation in early 2026, prototype growth by fiscal 12 months 2027, and full commercialisation by fiscal 12 months 2029.
Saimemory, established in December 2024 and based mostly in Tokyo, focuses on the analysis and growth of next-generation reminiscence options. This initiative is a part of SoftBank’s broader technique to reinforce Japan’s world competitiveness within the semiconductor trade.
SoftBank mentioned that Saimemory’s efforts are integral to the group’s key initiatives supporting future social infrastructure developments. By collaborating with Intel and different worldwide analysis establishments, SoftBank seeks to considerably contribute to the event of homegrown semiconductor applied sciences.
On this collaboration, Intel will function a expertise, innovation, and requirements collaborator, whereas Saimemory will present expertise innovation and lead the commercialisation of ZAM.
The ZAM expertise leverages foundational work beneath the Superior Reminiscence Expertise (AMT) analysis programme. Managed by the US Division of Vitality and Nationwide Nuclear Safety Administration, the AMT initiative entails contributions from Sandia Nationwide Laboratory, Lawrence Livermore Nationwide Laboratory, and Los Alamos Nationwide Laboratory.
Intel’s participation within the AMT programme is claimed to have contributed vital proof factors and efficiency validation for the stacked dynamic random-access reminiscence (DRAM) idea, which is central to ZAM’s structure.
Moreover, Intel’s developments by way of its Subsequent Technology DRAM Bonding (NGDB) initiative showcased greater DRAM density and bandwidth together with lowered latency and power consumption.
The shift from AMT to ZAM goals to strengthen U.S. and Japan expertise partnerships and speed up the deployment of analysis carried out in nationwide laboratories globally.
In keeping with Sanam Masroor, Intel’s Director of World Strategic Partnerships, the collaboration between Saimemory and Intel goals to deal with current bottlenecks in AI system scaling. By leveraging insights from the AMT and NGDB initiatives, the partnership seeks to reinforce packaging capabilities important for future expertise developments.
Intel Fellow and Intel authorities applied sciences chief expertise officer Joshua Fryman mentioned: “Intel’s Subsequent Technology DRAM Bonding (NGDB) initiative has demonstrated a novel reminiscence structure and revolutionary meeting methodology that considerably will increase DRAM efficiency, reduces energy consumption, and optimises reminiscence prices.
“Commonplace reminiscence architectures aren’t assembly AI wants, so NGDB outlined a complete new strategy to speed up us by way of the following decade.”
In August 2025, SoftBank dedicated to a $2bn funding in Intel, aligning with their technique to reinforce investments in superior expertise and semiconductor innovation within the US. At the moment, SoftBank mentioned that the funding displays its long-term imaginative and prescient of facilitating the AI revolution by rushing up the provision of superior applied sciences that underpin digital transformation, cloud computing, and next-generation infrastructure.
In late 2025, SoftBank Group closed the beforehand introduced acquisition of Ampere Computing, a US-based semiconductor design firm, for a worth of round $6.5bn. Ampere Computing is engaged in designing semiconductors for high-performance, power environment friendly AI computing utilizing the ARM compute platform.
Join our each day information round-up!
Give your enterprise an edge with our main trade insights.
