TSMC

Researchers from TSMC, Sandford University, ITRI and National Yang Ming Chiao Tung University have fabricate a 64-kb SOT-MRAM based β-phase Tungsten that offers a spin–orbit torque switching of 1 ns, data retention of more than 10 years and a tunnelling magnetoresistance of 146%.

The researchers say that Tungsten is a promising heavy metal for such applications and can generate large spin–orbit torques when stabilized in its β-phase. However, the α-phase, which has a lower spin-Hall angle, is more thermodynamically stable. It is thus challenging to integrate metastable β-tungsten into complementary metal–oxide–semiconductor processes while maintaining phase stability under the back-end-of-line thermal constraints (400 °C for extended durations). 

Renesas is introducing a new microcontroller (MCU) group, called the RA8P1, that is designed for artificial intelligence (AI), machine learning (ML), and real-time analytics applications. The devices, produced at TSMC's 22 nm ulta-low leakage process, target edge and endpoint AI applications, supporting convolutional and recurrent neural networks with up to 256 MACs per cycle

The RA8P1 combines a 1GHz Arm Cortex-M85 core, a 250MHz Cortex-M33 core, and an Arm Ethos-U55 Neural Processing Unit (NPU). The MCU use a 1 Mb on-chip embedded MRAM memory.

TSMC is launching its first design center in Europe, with a focus on memory technologies for automotive applications. The company announced that it will start developing 5nm MRAM technologies, to complement its 22-nm (in production), 16-nm (ready for customers verification) and 12-nm (already in development) technologies. 

Scaling MRAM technology to 5nm will be a significant step, required, according to TSMC, to implement AI automotive chips and technologies. The company will also develop 6 nm RRAM memory in the new design center.

In 2022, Taiwan's Industrial Technology Research Institute (ITRI) announced an agreement with Taiwan Semiconductor Manufacturing Company (TSMC) to collaborate on SOT-MRAM R&D. ITRI and TSMC now announced that they have developed SOT-MRAM array chips that boasts a power consumption of merely one percent of a comparable STT-MRAM device. 

ITRI and TSMC published a new research paper that was presented at the 2023 IEE International Electron Devices Meeting (IEDM 2023). ITRI explains that the new unit cell achieves simultaneous low power consumption and high-speed operation, reaching speeds as rapid as 10 nanoseconds. And its overall computing performance can be further enhanced when integrated with computing in memory circuit design. 

NXP Semiconductors announced a collaboration with TSMC to offer automotive embedded MRAM, in TSMC's 16 nm FinFET technology. NXP says that as automakers transition to software-defined vehicles, they need to support multiple generations of software upgrades on a single hardware platform. Bringing together NXP’s high-performance S32 automotive processors with fast and highly reliable next-generation MRAM in 16 nm FinFET technology provides the ideal hardware platform for this transition.

The new MRAM memory will be able to udpate 20MB of code in ~3 seconds compared to flash memories that take about 1 minute, which will minimize the downtime associated with software updates. MRAM also provides a highly reliable technology for automotive mission profiles by offering up to one million update cycles, a level of endurance 10x greater than flash and other emerging memory technologies.

Taiwan's Industrial Technology Research Institute (ITRI) announced two new MRAM collaborations. The first one is with Taiwan's TSMC, for the development of SOT-MRAM array chips. The second collaboration is with National Yang Ming Chiao Tung University (NYCU) to develop magnetic memory technology that can perform across a wide operating temperature range of nearly 400 degrees Celsius.

Together with TSMC, ITRI is developing low-voltage and current SOT-MRAM, that features high write efficiecny and low write voltage. ITRI says that its SOT-MRAM achieves a writing speed of 0.4 nanoseconds and a high endurance of 7 trillion reads and writes. The memory also offers a data storage lifespan of over 10 years.

During a presentation during Persistent Memory Summit, a new slide from TSMC was shown that describes the company's eMRAM roadmap:

As we already know, TSMC is offering 22nm eMRAM option as an eFlash alternative. The company is also looking to develop 14/12 nm eMRAM option to replace SRAM memory (this is interesting as before it was revealed that the company is developing 16nm MRAM). Finally eMRAM is touted as a possible replacement for configuration memory (eFUSE / OTP / Flash).

TSMC announced, during the company's virtual European technology symposium, that it is developing MRAM technologies for its 16nm FinFET platform. The company expects to offer flash-like configuration risk production starting in 4Q21 and RAM-like risk production scheduled for 4Q22.

The company also expects to "explore smaller geometries" for its MRAM (and RRAM) solutions. TSMC started offering 22nm MRAM in 2018.

In June 2017 it was reported that Taiwan Semiconductor Manufacturing Company (TSMC) will start producing embedded MRAM in 2018 using a 22 nm process. In what may bet he first adoption of TSMC's eMRAM technology, AI accelerator startup Gyrfalcon Technology announced the commercial availability of its LightSpeeur 2802M, AI ASIC that include TSMC's eMRAM.

The 2802M ASIC has 40MB of eMRAM memory, which can support large AI models or multiple AI models within a single chip. Applications include image classification, voice identification, voice commands, facial recognition, pattern recognition and more.

According to reports, Taiwan Semiconductor Manufacturing Company (TSMC) is aiming to start producing embedded MRAM chips in 2018 using a 22 nm process. This will be initial "risk production" to gauge market reception.

TSMC also aims to start embedded RRAM chip production in 2019.