X

我们使用cookie,以确保适合你的经验

我们用饼干和其他跟踪技术来提高你的浏览体验在我们的网站上,显示个性化内容和有针对性的广告,分析网站流量,并理解我们的观众来自哪里。你也可以阅读我们的隐私政策,我们使用cookie来确保你在我们的网站上的最好的体验。

选择我接受,或继续在网站上,你同意我们使用饼干和条款和条件。

得到应用

\"\"
<\/span><\/figcaption><\/figure>By Jane Lanhee Lee<\/strong>

OAKLAND: Silicon Valley startup Cerebras<\/a> Systems, known in the industry for its dinner plate-sized chip<\/a> made for artificial intelligence work, on Monday unveiled its AI supercomputer<\/a> called Andromeda<\/a>, which is now available for commercial and academic research.

Andromeda is built by linking up 16 Cerebras CS-2 systems, the company's latest AI computer built around the over-sized chip called the Wafer-Scale Engine 2<\/a>.

Cerebras said Andromeda can perform 1 exaflop worth of AI computing - or at least one quintillion (10 to the power of 18) operations per second - based on a precision of 16 bit floating point format.

The fastest U.S. supercomputer<\/a> that is capable of doing nuclear weapons simulations called the Frontier at Oak Ridge National Laboratory breached the 1 exaflop performance based on 64 bit double precision format this year.

\"They're a bigger machine. We're not beating them. They cost $600 million to build. This is less than $35 million,\" said Andrew Feldman, founder and CEO of Cerebras when asked about the Frontier supercomputer.

He said that while complicated nuclear simulations and weather simulations historically ran in 64 bit double precision computers, this is a computationally expensive format, so researchers are looking into whether AI algorithms can eventually match such outcomes.

Feldman said Andromeda is owned by Cerebras and built at a high performance data center in Santa Clara, California called Colovore. Companies and researchers, including those from U.S. national labs can access it remotely, he said.



<\/body>","next_sibling":[{"msid":95514025,"title":"MTNL loss widens to Rs 737 crore in Q2","entity_type":"ARTICLE","link":"\/news\/mtnl-loss-widens-to-rs-737-crore-in-q2\/95514025","category_name":null,"category_name_seo":"telecomnews"}],"related_content":[],"msid":95514068,"entity_type":"ARTICLE","title":"Silicon Valley chip startup Cerebras unveils AI supercomputer","synopsis":"The fastest U.S. supercomputer that is capable of doing nuclear weapons simulations called the Frontier at Oak Ridge National Laboratory breached the 1 exaflop performance based on 64 bit double precision format this year.","titleseo":"telecomnews\/silicon-valley-chip-startup-cerebras-unveils-ai-supercomputer","status":"ACTIVE","authors":[],"Alttitle":{"minfo":""},"artag":"Reuters","artdate":"2022-11-14 20:21:12","lastupd":"2022-11-14 20:25:41","breadcrumbTags":["cerebras","Andromeda","AI supercomputer","chip","Wafer-Scale Engine 2","devices","supercomputer"],"secinfo":{"seolocation":"telecomnews\/silicon-valley-chip-startup-cerebras-unveils-ai-supercomputer"}}" data-authors="[" "]" data-category-name="" data-category_id="" data-date="2022-11-14" data-index="article_1">

硅谷芯片启动大脑揭示AI的超级计算机

美国最快的超级计算机,模拟核武器的能力被称为边界在橡树岭国家实验室违反了1 exaflop性能基于64位双精度格式。

  • 更新于2022年11月14日08:25点坚持

由简Lanhee李

奥克兰:硅谷创业公司大脑系统,在该行业的餐盘芯片为人工智能工作,周一公布了人工智能计算机被称为仙女座,目前用于商业和学术研究。

仙女座是由连接16大脑CS-2系统,该公司最新的人工智能计算机芯片称为围绕一词圆片规模引擎2。

大脑说仙女座可以执行1 exaflop人工智能计算的价值——或者至少一百亿亿(10的18)操作每秒——基于16位浮点格式的精度。

广告
最快的美国超级计算机能够做的核武器模拟称为边界在橡树岭国家实验室违反了1 exaflop性能基于64位双精度格式。

“他们更大的机器。我们不是击败他们。他们花费6亿美元建造。这是不到3500万美元,”创始人兼首席执行官安德鲁•费尔德曼表示,大脑当被问及前沿的超级计算机。

他说,虽然复杂的核模拟和天气模拟历史上跑64位双精度计算机,这是一个需要大量的计算格式,因此研究人员正在调查AI算法能否最终匹配结果。

Feldman说仙女座属于大脑和建造在圣克拉拉的高性能数据中心,称为Colovore加州。公司和研究人员,包括美国国家实验室可以远程访问它,他说。



  • 发布于2022年11月14日08:21点坚持
是第一个发表评论。
现在评论

评论

现在评论 阅读评论(1)

加入2 m +行业专业人士的社区

订阅我们的通讯最新见解与分析。乐动扑克

下载ETTelec乐动娱乐招聘om应用

  • 得到实时更新
  • 保存您最喜爱的文章
扫描下载应用程序

\"\"
<\/span><\/figcaption><\/figure>By Jane Lanhee Lee<\/strong>

OAKLAND: Silicon Valley startup Cerebras<\/a> Systems, known in the industry for its dinner plate-sized chip<\/a> made for artificial intelligence work, on Monday unveiled its AI supercomputer<\/a> called Andromeda<\/a>, which is now available for commercial and academic research.

Andromeda is built by linking up 16 Cerebras CS-2 systems, the company's latest AI computer built around the over-sized chip called the Wafer-Scale Engine 2<\/a>.

Cerebras said Andromeda can perform 1 exaflop worth of AI computing - or at least one quintillion (10 to the power of 18) operations per second - based on a precision of 16 bit floating point format.

The fastest U.S. supercomputer<\/a> that is capable of doing nuclear weapons simulations called the Frontier at Oak Ridge National Laboratory breached the 1 exaflop performance based on 64 bit double precision format this year.

\"They're a bigger machine. We're not beating them. They cost $600 million to build. This is less than $35 million,\" said Andrew Feldman, founder and CEO of Cerebras when asked about the Frontier supercomputer.

He said that while complicated nuclear simulations and weather simulations historically ran in 64 bit double precision computers, this is a computationally expensive format, so researchers are looking into whether AI algorithms can eventually match such outcomes.

Feldman said Andromeda is owned by Cerebras and built at a high performance data center in Santa Clara, California called Colovore. Companies and researchers, including those from U.S. national labs can access it remotely, he said.



<\/body>","next_sibling":[{"msid":95514025,"title":"MTNL loss widens to Rs 737 crore in Q2","entity_type":"ARTICLE","link":"\/news\/mtnl-loss-widens-to-rs-737-crore-in-q2\/95514025","category_name":null,"category_name_seo":"telecomnews"}],"related_content":[],"msid":95514068,"entity_type":"ARTICLE","title":"Silicon Valley chip startup Cerebras unveils AI supercomputer","synopsis":"The fastest U.S. supercomputer that is capable of doing nuclear weapons simulations called the Frontier at Oak Ridge National Laboratory breached the 1 exaflop performance based on 64 bit double precision format this year.","titleseo":"telecomnews\/silicon-valley-chip-startup-cerebras-unveils-ai-supercomputer","status":"ACTIVE","authors":[],"Alttitle":{"minfo":""},"artag":"Reuters","artdate":"2022-11-14 20:21:12","lastupd":"2022-11-14 20:25:41","breadcrumbTags":["cerebras","Andromeda","AI supercomputer","chip","Wafer-Scale Engine 2","devices","supercomputer"],"secinfo":{"seolocation":"telecomnews\/silicon-valley-chip-startup-cerebras-unveils-ai-supercomputer"}}" data-news_link="//www.iser-br.com/news/silicon-valley-chip-startup-cerebras-unveils-ai-supercomputer/95514068">