Samsung demos world’s first in-memory computing based on MRAM

Samsung claims the MRAM in-memory computing chip achieved an accuracy of 98% in classification of hand-written digits and a 93% accuracy in detecting faces from scenes.

Samsung has demonstrated the world's first in-memory computing based on Magnetoresistive Random Access Memory (MRAM), a type of non-volatile memory, the South Korean firm announced on Thursday.

In-memory computing, a new computing paradigm that seeks to perform both data storage and data computing in a memory network, has emerged as one of the promising technologies to realize next-generation low-power AI semiconductor chips.

Samsung says the new scheme can process a large amount of data stored within the memory network itself without having to move the data, and also because the data processing in the memory network is executed in a highly parallel manner, power consumption is substantially reduced.

While non-volatile memories such as RRAM (Resistive Random Access Memory) and PRAM (Phase-change Random Access Memory), have been actively used for demonstrating in-memory computing, it's difficult to use MRAM for in-memory computing. Given the low resistance of MRAM, it cannot enjoy the power reduction advantage when used in the standard in-memory computing architecture, Samsung explains.

To resolve this issue, Samsung researchers successfully developed an MRAM array chip that demonstrates in-memory computing, by replacing the standard, 'current-sum' in-memory computing architecture with a new, 'resistance sum' in-memory computing architecture, which addresses the problem of small resistances of individual MRAM devices.

Samsung claims the MRAM in-memory computing chip achieved an accuracy of 98% in classification of hand-written digits and a 93% accuracy in detecting faces from scenes.

The researchers have also suggested that not only can this new MRAM chip be used for in-memory computing, but it also can serve as a platform to download biological neuronal networks.

The research was led by Samsung Advanced Institute of Technology (SAIT) in close collaboration with Samsung Electronics Foundry Business and Semiconductor R&D Center. The paper on this innovation, titled A crossbar array of magnetoresistive memory devices for in-memory computing, has been published online by Nature and will also be published in the upcoming print edition of the journal.