[an error occurred while processing the directive]
RSS
Логотип
Баннер в шапке 1
Баннер в шапке 2

Pohoiki Beach (neyromorfny system)

Product
The name of the base system (platform): Intel Loihi (neyromorfny processor)
Developers: Intel
Date of the premiere of the system: July, 2019
Branches: Information technologies
Technology: Processors

Content

2019: Announcement

On July 15, 2019 the Intel corporation announced the neyromorfny computing system on the basis of Loihi chips which purpose is modeling of work of a human brain. A system under the name Pohoiki Beach consists of 64 Loihi chips which include analogs of 130 thousand neurons and 130 million synapses. Thus a system in general have the computing power which is equivalent to 8 million neurons.

Intel provided the computing system which imitates a human brain

Intel claims that Loihi chips have the improbable speed of training thanks to capability to imitate work of a human brain. The solution can also use information which it already has to draw conclusions about new data, accelerating learning process in geometrical progression. The idea of neyromorfny calculations means that chips were developed on the basis of architecture of a human brain and modern neurological theories of its work.

According to the statement of Intel, Loihi chips can process information up to 1000 times faster and 10,000 times more effectively, than traditional central processors for customized applications. At the same time Loihi chips differ in extremely low power consumption. At the moment, according to Intel, the actual capacity of Pohoiki Beach consists in the solution of some of the most difficult computing tasks for Internet of Things and other self-contained units thanks to extremely low power consumption of Loihi. Intel is going to provide a new system to 60 partners in researches to study possibilities of new architecture.

The Pohoiki Beach system is only the base of researches in the field of neyromorfny calculations. In 2019 Intel is going to release more powerful system under the name Pohoiki Springs which performance will correspond to 100 million neurons.[1]

You See Also

Intel Pohoiki Springs (neyromorfny system)

Notes