All articles| All Pictures| All Softwares| All Video| Go home page| Write articles| Upload pictures

Reading number is top 10 articles
英特尔清华等签署协议,将联合研发新型CPU - Intel,清华大学,英特尔
Sharp change, get Apple iPhone OLED big cake,
老年人手机并非老手机 - 老年人手机
票贩子玩新花样,2016春运来路不明的车票慎买 - 春运,2016春运,黄牛
马航MH370失事又现新解释:锂电池起火导致 - 马航,MH370
发改委原副主任:不能一味唱衰中国制造 - 中国制造
司机不认路,Uber做了一个手机游戏 - Uber
Tech giants are not supported: United Kingdom out of the EU influence on the IT industry,
东芝被踢出日经指数,面临9000亿日元违约金 - 东芝,日经指数
SOFTBANK to buy ARM, the latter’s share price rose almost 50%,
Reading number is top 10 pictures
9.3阅兵全景图1-抗战老兵和英雄连队梯队
Female model behind the bitterness, often being overcharged4
NeedWallpaper9
分手的感悟
真正的国产-非模拍 贵在是真实1
The money of more than 100 countries and regions2
Distribution of wealth in China survey status report
A beautiful girl to bud1
Exquisite decoration is not paying too much1
西方气质的东方美女2
Download software ranking
Unix video tutorial18
Tram sex maniac 2 (H) rar bag7
Sora aoi - one of more PK
Love the forty days
Unix video tutorial13
尖东毒玫瑰A
C#COM编程指南
dreamweaver8中文版
Unix video tutorial4
Professional killers2 for Android
归海一刀 published in(发表于) 2014/5/5 7:18:05 Edit(编辑)
The dreaded “brain“ speed ultra PC 9,000 times times

The dreaded “brain“ speed ultra PC 9,000 times times(恐怖的“大脑”运算速度超PC九千倍,)

The dreaded "brain" speed ultra PC 9,000 times times-brain, PC, operation speed, the computer-IT news The dreaded "brain" speed ultra PC 9,000 times times

According to foreign media reports, California, scientists at Stanford University have developed a new circuit board, it can help people understand the real power of their own brains. This neural network is based on the human brain as a model, it is able to simulate 1 million nerve cells and billions of synapses, far exceeding other mimic the brain's equipment.

Currently being conducted by researchers at Stanford University BRAIN project was to develop new tools to measure brain activity. Neural mesh size as the iPad , and with 16 special "neuro-core" chips, the chips more efficiently than ordinary computer, and speeds up to 9,000 times.

Associate Professor of bioengineering at Stanford University Kwabena Boahen says: "from the standpoint of pure energy, the brain is hard to match. "Bioengineering Division but Boahen and his colleagues hope that this project could pave the way for progress in the field of Robotics and computer, and find new ways of understanding how the brain works.

Only individuals can understand human brain neural network programming, but Boahen plans to create a "neuro-compiler", which enables people with no knowledge of the neurobiology scientists were also able to use this device. Researchers at Stanford University now also developing a chip similar to neural network control of prostheses, this chip can decode brain signals directly to campaigns.

Boahen says neural mesh is currently the most cost-effective tools developed so far, but each loop circuit continues to spend roughly $ 400, now researchers are looking for ways to reduce costs. In any case, research of simulating the human brain still has a long way to go. Boahen says: "with neurons in the human brain is a neural network of 80,000 times, energy consumption only it three times. Achieve this level of efficiency is an ultimate challenge. ”


(

恐怖的“大脑”运算速度超PC九千倍 - 大脑,PC,运算速度,计算机 - IT资讯
恐怖的“大脑”运算速度超PC九千倍

据国外媒体报道,加州斯坦福大学的科学家们发明了一种新的电路板,它能够帮助人们了解他们自己大脑的真正力量。这种神经网格是以人类大脑为模型的,它能够模拟一百万神经细胞和数十亿突触,远超其它模仿大脑的设备。

斯坦福大学的研究人员目前正在进行的BRAIN项目,就是为了研发出测量大脑活动的新工具。神经网格的大小与iPad相当,而且有着16个特制的“神经核心”芯片,这些芯片比普通电脑更高效,而且速度提升了9000倍。

斯坦福大学生物工程学副教授Kwabena Boahen称:“从纯能量角度来说,大脑非常难以匹配。”但是Boahen和他的生物工程师同事们希望,这个项目能够为机器人学和计算机领域的进步铺平道路,并且找到了解大脑如何工作的新方法。

目前只有了解人类大脑的个人才能够为神经网格编程,但是Boahen计划创造一种“神经编译器”,它能够使那些没有神经生物学知识的科学家们也能使用这种设备。斯坦福大学的研究人员们现在也在合作研发一种由类似于神经网格的芯片控制的假肢,这种芯片会将大脑信号直接解码为运动。

Boahen称,神经网格是目前为止研发出来的最合算工具,但是每个循环电路仍然耗费了大约400美元,现在研究人员正在寻找降低成本的方法。不管怎样,模拟人类大脑的研究仍然要有很长一段路要走。Boahen称:“人类大脑拥有的神经元数量是神经网格的8万倍,但消耗的能量只有它的三倍。达到这种能效水平是一项终极挑战。”


)


添加到del.icio.us 添加到新浪ViVi 添加到百度搜藏 添加到POCO网摘 添加到天天网摘365Key 添加到和讯网摘 添加到天极网摘 添加到黑米书签 添加到QQ书签 添加到雅虎收藏 添加到奇客发现 diigo it 添加到饭否 添加到飞豆订阅 添加到抓虾收藏 添加到鲜果订阅 digg it 貼到funP 添加到有道阅读 Live Favorites 添加到Newsvine 打印本页 用Email发送本页 在Facebook上分享


Disclaimer Privacy Policy About us Site Map

If you have any requirements, please contact webmaster。(如果有什么要求,请联系站长)
Copyright ©2011-
uuhomepage.com, Inc. All rights reserved.