Stock Markets February 2, 2026

SoftBank unit and Intel to jointly develop 'Z-Angle' memory technology

Saimemory and Intel aim for high-capacity, high-bandwidth, low-power chips for AI data centers with prototypes targeted by early 2028

By Nina Shah
SoftBank unit and Intel to jointly develop 'Z-Angle' memory technology

SoftBank Corp's unit Saimemory and Intel Corporation have agreed to collaborate on commercializing a new memory architecture called Z-Angle operation memory technology. The partners said the work targets memory chips that combine high capacity, high bandwidth and reduced power consumption, intended for use in artificial intelligence data centers. The effort will use Intel's next-generation DRAM bonding initiative, with prototypes expected by at least early 2028. SoftBank Corp is a listed unit of SoftBank Group Corp., which had agreed in mid-2025 to invest $2 billion in Intel.

Key Points

  • Saimemory and Intel will develop Z-Angle operation memory technology emphasizing high capacity, high bandwidth and low power consumption.
  • The technology is being targeted for AI data centers, where high-speed memory is essential for large generative AI models.
  • The partnership plans to leverage Intel's next-generation DRAM bonding initiative and aims to produce prototypes by at least early-2028.

SoftBank Corp's unit Saimemory and Intel Corporation announced a partnership to develop a novel class of memory chip technology they have named Z-Angle operation memory technology.

According to the companies, the collaboration focuses on bringing to market memory devices that seek to combine three attributes: high capacity, high bandwidth and low power consumption. The joint goal is to create chips that meet those technical targets and that can be integrated into environments demanding heavy computational throughput.

The statement from SoftBank notes the technology is being developed with artificial intelligence data centers specifically in mind. High-speed memory is cited as a critical component for AI data centers because running large generative AI models requires substantial memory bandwidth and capacity to sustain compute operations.

Under the terms described, Saimemory will work with Intel's next-generation DRAM bonding initiative as part of the development process. The companies said they aim to produce prototypes by at least early-2028, providing a concrete near-term timeline for initial hardware samples.

SoftBank Corp is identified as a publicly listed unit of SoftBank Group Corp. and serves as the conglomerate's primary operating company. The statement also notes that SoftBank Group had in mid-2025 agreed to make a $2 billion investment in Intel.

Intel framed the collaboration in the context of strengthening its memory and chip portfolio as it looks to expand its presence supplying technology for the AI sector. The development work on Z-Angle operation memory technology and the use of DRAM bonding techniques are positioned as steps to enhance Intel's competitive offering in this market.


Summary

Saimemory, a unit of SoftBank Corp (TYO:9434), has entered a development partnership with Intel Corporation (NASDAQ:INTC) to commercialize Z-Angle operation memory technology. The design goals are higher capacity, greater bandwidth and lower power usage, aimed at AI data center workloads. The project will leverage Intel's DRAM bonding initiative and targets prototype completion by at least early-2028. SoftBank Group (TYO:9984) had previously agreed to invest $2 billion in Intel in mid-2025.

Key points

  • Saimemory and Intel will co-develop Z-Angle operation memory technology focused on capacity, bandwidth and energy efficiency.
  • The technology is intended for artificial intelligence data centers, where high-speed memory is critical to running large models.
  • The partnership will use Intel's next-generation DRAM bonding initiative and aims to yield prototypes by at least early-2028.

Risks and uncertainties

  • Timing risk - the target to create prototypes by at least early-2028 represents a development timeline that may be subject to change.
  • Execution risk - commercializing a new memory architecture that meets the stated goals for capacity, bandwidth and power presents technical and integration challenges.
  • Market competitiveness - Intel's effort to shore up its chip offerings for the AI sector implies competitive pressure from other suppliers, which could affect adoption and market share.

Risks

  • Timing risk: the prototype target of early-2028 may be delayed, affecting deployment schedules - impacts semiconductor and AI infrastructure sectors.
  • Execution risk: technical and integration challenges in commercializing a new memory architecture could hinder performance targets - impacts memory and semiconductor markets.
  • Market competitiveness: Intel's push to expand its AI-related chip offerings indicates competitive pressure that could influence adoption and commercial success - impacts semiconductor suppliers and data center customers.

More from Stock Markets

Tesla Debuts New All-Wheel Drive Model Y Trim in U.S.; Premium Option Also Launched Feb 2, 2026 Eastroc Beverage Shares Start Trading in Hong Kong at Offer Price After $1.3 Billion IPO Feb 2, 2026 M EVO GLOBAL ACQUISITION CORP II Raises $300 Million in IPO Aimed at Critical Minerals Deals Feb 2, 2026 NRW Holdings Shares Rise After Securing A$175m Rio Tinto Earthworks Contract Feb 2, 2026 Federal Judge Blocks Attempt to End Temporary Protected Status for Haitians Feb 2, 2026