As embedded intelligence is extending its way into more zones of our lives, domains spanning from autonomous driving to tailored medicine are producing huge data amounts. But just as the overflow of data is attaining enormous proportions, the potential of computer chips to route it into valuable data is stymied.

The existing computers consist of a chip for computing and one more for storing data. As increased bulks of data are analyzed, the restricted rate at which information can be relocated between the chips is generating a communication blockage. However, the researchers at MIT and the Stanford University have developed a new chip to surpass this obstacle.

The new prototype chip utilizes several nanotechnologies along with new 3D computer architecture to converse this trend. Professor at Stanford University, Subhasish Mitra, said “The new 3D computer design offers fine-grained and dense incorporation of data storage and computing, considerably surpassing the limitation of moving data between chips. Consequently, the chip is capableof storing enormous volumes of data and executing on-chip processing to convert a data flood into valuable data.”

The team incorporated over 1 Million RRAM (resistive random-access memory) cells, 2 Million carbon nanotube transistors for processing, and a new type of memory storage creating dense 3D computer design with interweaving layers of memory and logic.By introducing between these layers the ultra-dense wires, the 3D design assures to tackle the connection “bottleneck”.

Stanford’s Philip Wong mentioning about a conventional memory called as the dynamic random-access memory, said, “Logic composed of carbon nanotubes can be much more energy-efficient in comparison to the existing logic composed silicon and, likewise, RRAM can be faster, denser, and more energy-efficient.”A major benefit of this invention is that the new chip is well-suited with existing silicon structure, both in terms of design and fabrication.

The team is functioning to enhance the basic nanotechnologies while studying the new 3-D computer design. For the Assistant Professor at MIT, Max Shulaker, the following step is functioning with Analog Devices, which is aMassachusetts-based semiconductor firm, to create new versions of the tool that take benefit of its potential to perform data processing and sensing on the same chip.