Продолжение истории одинокого листочка. Еще раз напоминаю – особо чувствительным лучше ...
Продолжение истории одинокого листочка. Еще раз напоминаю – особо чувствительным лучше воздержаться от чтения. С частью 1 можно ознакомиться тут. История одинокого листка. Часть 2 Я – листок блокнота Moleskine, я ищу своего брата, мы с ним были двойными листочками. 25.01.2013 мы спокойно лежали в сложенном карманном блокноте, изредка общаясь с белыми нитками, удерживающими нас от падения [...]
В течение пребывания в поезде №160 “Сапсан” сообщением “Москва – Санкт- ...
В течение пребывания в поезде №160 “Сапсан” сообщением “Москва – Санкт-Петербург” мое воображение посетила одна мысль, которая впоследствии стала отправной точкой формирования небольшого рассказа. И этот рассказ я не могу не опубликовать. Прошу людей, страдающих вследствие душераздирающих переживаний, читать очень осторожно. Я предупредил. Предыстория. Я ехал в поезде, писал различные мысли в карманном блокноте Moleskine. Внезапно [...]
... descended from. Who
all things that ...
... old (format 10,
by Subversion 1 ...
... would allow to
artificial reason already ... software allows to
the virtual supercomputer ...
Problem of creation of machine intelligence mankind заинтересоввалось rather recently, approximately from the middle of the last century. Currently already there are theoretical practices and reasons in the field. It would allow to create artificial reason already today but while everything rests against high-speed performance of the modern means, and often and in their cost.
Recently came across one very interesting development which allows to create a supercomputer to a quantity of participants on the basis of the P2P networks. Development is absolutely free (Open Source) and is available in repositories of top distribution kits of Linux, wears the name CPUShare. The developer of this project is Andrea Arcangeli. The software allows to create the virtual supercomputer within several minutes. This program obsepecheniye works at the majority of widespread equipment rooms and all system (OS) platforms. Project site: http://www.cpushare.com. Be curious, won’t regret spent time.
Use of P2P of networks for implementation of the distributed computation gives conclusive priority – the number of participants can be very big, respectively and computational capabilities can be on orders above existing capacities of supercomputers in the modern data-centers. Besides, in view of decentralization of P2P of a network, this virtual supercomputer will be the most fault-tolerant, i.e. the functionality to remain regardless of if the part of a network fails for any reasons. For example, natural disasters, hostilities, diversions etc. This software as it was mentioned above, can be unrolled on any operating systems (Linux, Unix, Windows), and it will allow to cut down for orders expenses on creation of powerful computing systems, or to make their computational capabilities still big and fault-tolerant. And in case of universal mandatory implementation of a client part of such system, for example, in a number of branches, the departments, the separate enterprises, will allow to finish computational capability to inaccessible before heights.
It also is of interest for normal users, for example, by means of it it is possible to receive necessary additional high-speed performance on unproductive hardware platforms which for some reasons while there is no possibility to replace, for example, computer games. Options there can be a set. And commercial use isn’t excluded, it is possible to lease excessive computational capabilities for a certain board.
Nikitushkin Andrey
On July 07, 2012.
https://nikitushkinandrey.wordpress.com/2012/07/17/supercomputer-on-the-basis-of-p2p-of-networks-one-more-step-on-a-way-to-machine-intelligence-creation/
