Technical / Research - Page 11

Toshiba announced a new computing architecture that uses only STT-MRAM to perform both operations and storage. The idea support computing capability, register file, primary cache and secondary cache all on the same perpendicular STT-MRAM, and Toshiba says it could lead to CPUs that are drastically faster and more efficient.

Toshiba explains that in the new architecture, the results of operations (answers) that correspond to combinations of certain inputs are prepared in advance in the form of a table and stored in the memory. In response to an input, an answer is read out of the memory. This is equivalent to an operation carried out by a CPU. Because the computation answer is read once, it drastically improves processing speed and power consumption.

According to Nikkei, over 20 Japanese and US companies have teamed up to develop MRAM technologies, in particular a new mass production method. Participants in this ambitious project include Tokyo Electron (who's merging with Applied Materials), Renasas, Hitachi and Shin-Etsu Chemical from Japan and Micron Technology from the US.

Japan's Tohoku University, a leader in Spintronics and MRAM research, will also join the project. The companies will finance several dozens researchers at the University. They plan to start development in February 2014, and continue to seek more companies from the US and Europe to join. The aim is to complete materials and processes development by 2017 and start mass production by 2018.

Crocus Technology filed an Inter Partes Review petition to the US PTO, claiming that US patent #6,980,469 describes a technology already used in prior art, in particular in the patent portfolio of Crocus. The patent in question describes a high-speed low power magnetic devices based on current induced spin-moment transfer, and is owned by New York University (although crocus says in their PR that it is owned by Spin Transfer Technologies (STT).

Crocus currently holds 154 patents, describing their Magnetic Logic unit (MLU) design and manufacturing as well as generic technologies like STT (Spin Torque Transfer).

During Flash Memory Summit 2013, Sony delivered a lecture on ReRAM (resistive random-access memory) - a technology Sony has been developing together with Micron. Sony hopes to start producing 16Gb ReRAM chips by 2015 (20nm) for the storage-class memory (SCM) market (SCM fills the performance gap between DRAM and NAND).

Sony says they chose ReRAM as its SCM because its memory cell can be made using advanced process technologies and a cross-point cell array can be used for it, making it easy to increase memory density.

Yesterday we reported that Crocus Technology raised €34 million (about $45 million) in their Series D round of funding. Crocus said that part of that money will be used to ramp up its manufacturing at Tower Semiconductors (in Israel), and today they reported about a major breakthrough in that technology development, which they will bring to market by Q4 2013.

Crocus, together with TowerJazz now offers SRAM-performance with non-volatile MRAM. The MRAM solution has demonstrated best-in-class performance and more than two billion cycles of program/erase were demonstrated on a 4M bit NVM product. The first market that the two companies will address is Battery Backed SRAM (BB-SRAM). The MRAM technology offers fast access times, low power and unlimited write cycles - without the need for cost, recharging and long-term disposal of a Lithium battery.

One year ago, the French National Research Agency (ANR) launched an MRAM project called MARS (MRAM based Architecture For Reliable and low power Systems). The main focus of this project is the study of MRAM technology's contribution to embedded processors architectures. MARS also aims to build new MRAM architectures and design new software models.

The MARS project already reached several milestones:

• An open-source, generic STT and TAS compact Model (Spinlib library) for electrical simulation
• Developed new Non-Volatile Flip Flop (several patents were already requested) and a new non-volatile SRAM/MRAM memory (also patented)
• A full study on reliability on the STT-MRAM comprising noise and stochastic effects
• A study on the use of MRAM on the embedded processor hierarchy (MRAM cache memory based architecture)
• New MRAM/DRAM cells was designed and successfully tested on silicon

Samsung Electronics launched a new global research outreach program aimed towards STT-MRAM innovation. The Samsung Global MRAM Innovation (SGMI) is looking for colleges, universities and research labs from all over the world to explore breakthrough and innovative STT-MRAM research.

Samsung invites submissions for novel ideas on STT-MRAM research - and the selected proposals will receive financial support from Samsung. Samsung will give around $70,000 to $150,000 for one year research (which may be extended for up to three years). They may also offer larger funds for exceptional proposals. The deadline for submission is September 28, 2013.

Crocus Technology signed a license and joint-promotion agreement with TowerJazz, for the use of Crocus' Magnetic Logic Unit (MLU) process technology. TowerJazz customers will be able to use Crocus' technology in embedded System-On-Chip (SoC) applications. First products based on MLU are expected on the market before the end of 2013 as some customers are already in the process of adopting the technology.

Crocus's MLU blocks are used in magnetic sensors and secure embedded memories. The memory is fast (like DRAM) and non volatile. It also includes unique security options. Crocus actually plans to produce (at TowerJazz's Migdal Haemek manufacturing facility in Israel) their own magnetic sensor and non-volatile memory products.

GlobalFoundries announced it is joining Qualcomm and Imec (and other companies) in their joint development effort to advance STT-MRAM technology. GlobalFoundries is the first IC maker to join imec's R&D program on emerging memory technologies. Imec says that they now have the complete infrastructure necessary for STT-MRAM R&D.

Imec and the other members aim to explore the potential of STT-MRAM, including performance below 1ns and scalability beyond 10nm for embedded and standalone applications.

Researchers from the University of Delaware confirmed that electrons generate a magnetic field in a ferromagnetic material This new field does not radiate beyond the ferromagnetic material and so may lead towards high-density MRAM memory cells (in current technology it is difficult to shield the magnetic fields between memory cells).

In materials made from two layers of a heavy metal and a ferromagnetic material, the spin current diffuses into the ferromagnetic material. When this happens, a magnetic field is generated. This magnetic field does not radiate beyond the ferromagnetic material (unlike regular magnetic fields).