IBM conducted a neurocomputer simulation comparable to that of a human brain.
At the Supercomputing 2012 conference, IBM presented the results of the Compass neuro-simulator on the Sequoia supercomputer (second place in the global supercomputer rating ). For the first time, a scale corresponding to the human brain was reached - 530 billion neurons and 137 trillion synapses. The simulation took place 1542 times slower than real time. All 1 572 864 cores and one and a half petabyte of memory were involved in it.
Compass is a simulation system for the new TrueNorth computer architecture, which at IBM is called the “big brother of Watson.” Unlike the familiar to us computers based on the von Neumann architecture, TrueNorth is built in the image and likeness of the brain, based on IBM-developed " neurosemantic cores " - chips containing silicon neurons and synapses. Each chip contains 256 neurons, 1024 axons and 256x1024 synapses. The area of ​​the chip produced by 45-nanometer technology is 4.5 mm 2 . In a record simulation, 2084 billion such nuclei were modeled.
This work was done as part of the DARPA SyNAPSE program , the ultimate goal of which is to create a neurocomputer comparable to the brain of higher mammals in terms of speed and cognitive abilities, as well as in terms of compactness and energy intensity. That is, in fact, we are talking about creating a brain for “terminators”. Here are some of its characteristics:
')
IBM has been working on creating a “cognitive computer” for DARPA for several years. So, in 2009, neurostimulation results were presented on a cat's brain scale. It is too early to talk about a realistic imitation of the human brain, because in addition to a sufficient number of neurons, it is also necessary to know the scheme of their connection and, most importantly, the program that controls the brain. There is progress in this direction too. One of the simulations in the framework of DARPA SyNAPSE was based on the diagram of the connections of neural complexes in the brain of a macaque obtained in the framework of the CoCoMac project.
Compass is a simulation system for the new TrueNorth computer architecture, which at IBM is called the “big brother of Watson.” Unlike the familiar to us computers based on the von Neumann architecture, TrueNorth is built in the image and likeness of the brain, based on IBM-developed " neurosemantic cores " - chips containing silicon neurons and synapses. Each chip contains 256 neurons, 1024 axons and 256x1024 synapses. The area of ​​the chip produced by 45-nanometer technology is 4.5 mm 2 . In a record simulation, 2084 billion such nuclei were modeled.This work was done as part of the DARPA SyNAPSE program , the ultimate goal of which is to create a neurocomputer comparable to the brain of higher mammals in terms of speed and cognitive abilities, as well as in terms of compactness and energy intensity. That is, in fact, we are talking about creating a brain for “terminators”. Here are some of its characteristics:
- 10 billion neurons
- 100 trillion synapses
- power consumption within 1 kW
- occupied volume - less than 2 dm 3
- interfaces for sensory perception and motor functions
')
IBM has been working on creating a “cognitive computer” for DARPA for several years. So, in 2009, neurostimulation results were presented on a cat's brain scale. It is too early to talk about a realistic imitation of the human brain, because in addition to a sufficient number of neurons, it is also necessary to know the scheme of their connection and, most importantly, the program that controls the brain. There is progress in this direction too. One of the simulations in the framework of DARPA SyNAPSE was based on the diagram of the connections of neural complexes in the brain of a macaque obtained in the framework of the CoCoMac project.
Source: https://habr.com/ru/post/159365/
All Articles