MRAM applications

High-performance STT-MRAM developer Numem announced that it expects to start samplings its MRAM chipsets, to enable non-volatile, high-speed and ultra-low power solutions to chip designers.

Numem says that the company's chiplets deliver up to 4TB/sper 8-die memory stack, exceeding existing AI memoryHBM solutions.  It supports 4GB per stack package, and is optimized for AI applications across OEMs and hyperscalers.

Vertןcal Compute, a new startup that was spun-off from the imec institute to commercialize its MRAM technology for AI applications, has raised 20 million Euro. The seed round was led by imec.xpand and supported by Eurazeo, XAnge, Vector Gestion, and imec.

The company says that it will develop an MRAM in-memory computing chiplet technology. The chiplet, according to the company, can reduce power consumption by 80% and speed up the execution of large language models for AI by 100X. It is believed that the company's MRAM is based on SOT-MRAM technology.

Everspin Technologies announced that Lucid Motors has chosen the company's PERSYST MRAM technology to be used in its latest all-electric Gravity SUV. 

Lucid's Gravity uses Everspin's MR25H256A, a 256Kb serial MRAM. Everspin says it was selected because it meets the AEC Q100 Grade 1 specification of -40C to 125C temperature operation. The MRAM is used in the Gravity to handle data logging and parameter storage to assist in the efficient operation of the all-electric powertrain.

US-based Carnegie Mellon University spinoff Efficient has signed a strategic partnership with GlobalFoundries to launch its low-power Fabric chip on GlobalFoundries' 22FDX low power silicon on insulator process. Efficient will adopt embedded MRAM in its chip, and expects to launch it (first samples) by the summer of 2025.

Efficient is an early stage company that raised over $16 million to develop the power-efficient Fabric processor, targeting edge applications such as machine learning-enabled extreme-edge machine vision, continuous audio intelligence and versatile sensory and signals intelligence.

GlobalFoundries has been offering eMRAM 22nm process solutions since 2017, based on Everspin's pMTJ technology and IP.

Researchers from the University of Minnesota Twin Cities developed a new AI hardware accelerator, based on MRAM memory, that can accelerate AI computation by a factor of 1,000. 

The so-calledd Computational Random-Access Memory device, or CRAM, is a machine learning inference accelerator, that is not just 1,000 faster, it also exhibits energy saves of 1,700 to 2,500 times compared to traditional methods. 

Alif Semiconductor announced a new microcontroller in its Ensemble family, the E1C, that offers 46 giga-operations per second (GOPS) of minimum-precision compute for on-device machine learning and artificial intelligence workloads.

The Alif E1C offers several memory options, depending on the model, ranging from 256Kb of SRAM in the entry-level option, up to 1.9Mb of MRAM and 2Mb of SRAM in the high-end option.

High-performance STT-MRAM developer Numem announced that in collaboration with ASIC/SoC designer IC’ALPS it has developed an integrated circuit with RISC-V processors, 2MBytes of NuRAM and a DSP/AI Custom Datapath Accelerator.

The two companies co-developed the SoC in an advanced technology node, taking advance of the high-performance and low power consumption of Numem's memory technology. The company also details that the physical implementation of this integrated circuit was made in a secure space (isolated location, network, and servers, and encrypted exchanges) to meet with the stringent protection of sensitive data required by this program. Numem and IC’Alps intend to extend their partnership to serve new SoC projects for customers.

Everspin announced that IBM has chosen the PERSYST EMD4E001G 1Gb STT-MRAM for use in their FlashCore Module 4.

Everspin says that its EMD4E001G high-performance persistent memory ensures critical data integrity even during power loss. With a DDR4 interface, it delivers 2.7 gigabytes/second of both read and write bandwidth, coupled with instant non-volatility.

pMTJ STT-MRAM developer Avalanche Technology announced that it is providing its Space Grade Persistent SRAM (P-SRAM) STT-MRAM products for satellite-based storage systems developed by RAiTEK. 

RAiTEK has designed in a 16Mb device from Avalanche's Gen 2 Space Grade family to monitor data stream status, and a 1Gb Gen 3 Space Grade to support large configuration images for the onboard FPGA used in the drives, as part of their growing portfolio of data storage and processing solutions for the space industry.