MRAM-Info: the MRAM experts

Researchers from Samsung's Advanced Institute of Technology (SAIT), have demonstrated what they say is the world’s first in-memory computing based on MRAM, targeting next-generation AI chips.

The researchers explain that In-Memory computing is a new paradigm that seeks to perform both data storage and data computing in a memory network. In such a computing system, a large amount of data, stored in the memory network, can be executed in a highly parallel manner. Power consumption in such systems is substantially reduced.

Researchers from the University of Arizona developed a new pMTJ structure that exhibits high magnetoresistance, strong retention and is likely to achieve fast switching times as well.

The new structure uses a multi-interface free layer (MIFL) which incorporate multiple materials with different properties. The researchers used a ferromagnetic CoFeB layer with nonmagnetic Mo or MgO layers. The magnetoresistance can be controlled by changing the thickness of the CoFeB layer. The researchers managed to demonstrate a magnetoresistance of over 200%.

2021 is soon over - and it was another year dominated by the global pandemic - but also with increased MRAM production, adoption and developments.

Here are the top 10 stories posted on MRAM-Info in 2021, ranked by popularity (i.e. how many people read the story):

1. TSMC shows its eMRAM technology roadmap (May 6)
2. Samsung improves its MRAM performance, will expand its target applications (Feb 5)
3. New SOT-MRAM device structure can be scaled up and is highly efficient (Aug 9)
4. Avalanche announces new 64Mb industrial x16-interface STT-MRAM memory devices (Sep 29)
5. Orthogonal Spin Transfer MRAM developer Spin Memory liquidates (Jul 15)
6. Everspin reports its financial results for Q1 2021 (May 7)
7. Everspin reported its preliminary Q4 2020 and full-year 2020 financial results (Mar 1)
8. Researchers developed a promising antiferromagnetic MRAM device structure (Jun 27)
9. Lucid Motors adopted Everspin's MRAM in its Lucid Air all-electric sedan (Jun 16)
10. Researchers use stochastic MRAM elements to create highly efficient AI devices (Jul 7)

Renesas Electronics announced that it has developed two new technologies that reduce the energy and voltage application time for the write operation of STT-MRAM chips.

On a 20-Mbit test chip with embedded MRAM memory cell array in a 16 nm FinFET logic process, a 72% reduction in write energy and a 50% reduction in the voltage application time were confirmed.

Researchers from Tohoku University managed to fabricate the world's smallest STT-MRAM MTJ, at 2 nm. In addition, the researchers demonstrated fast switching (3.5 ns) in sub-five-nm STT-MRAM MTJs.

The new MTJ were developed using a new multilayered ferromagnetic structure that can engineer characteristic relaxation time, which governs the magnetization dynamics in the ns regime.

Hprobe announced a new magnetic test head for MRAM Wafer Sort. The new module, the H3DM-XL, is at the heart of the latest addition to Hprobe’s IBEX line, the IBEX-WS.

The IBEX-WS test equipment integrates 3D magnetic field capabilities, while increasing field area and uniformity for wafer probing large MRAM arrays. It also features a unique patented robotized 3D Field Calibration Unit (FCU) for high-speed field mapping and monitoring.

The MRAM market is set for another record year in 2022, as MRAM chips and memories are increasingly being used in industrial, space, IT, automotive and other applications. If you are looking to promote your company's products and services for the MRAM market, increase industry awareness of your technology, and generate industry leads - MRAM-Info is the perfect choice!

MRAM-Info offers the easiest, most direct path to industry exposure, with specifically tailored marketing packages that will suit your needs. MRAM-Info is a well-established brand and a publication accessed by thousands of memory professionals and consumers each month, via our website, newsletter and social media channels.