英偉達(dá)(Nvidia Corp.)的黃仁勛與亞洲首富穆克什·安巴尼達(dá)成了一項(xiàng)合作協(xié)議,將在全球人口最多的國家印度建設(shè)人工智能基礎(chǔ)設(shè)施和推廣人工智能技術(shù)。
上周四,這兩位高管在英偉達(dá)于孟買召開的人工智能峰會(huì)上同臺(tái)亮相,他們表示,安巴尼的信實(shí)工業(yè)(Reliance Industries Ltd.)在印度新建的大型數(shù)據(jù)中心,將使用美國公司英偉達(dá)的最新Blackwell芯片。英偉達(dá)還與印孚瑟斯(Infosys Ltd.)和塔塔咨詢服務(wù)公司(Tata Consultancy Services Ltd.)等多家印度綜合性企業(yè)建立了合作關(guān)系。
作為一個(gè)擁有14億人口的大國,印度已經(jīng)成為一個(gè)重要的人工智能市場(chǎng),該國的農(nóng)業(yè)、教育和制造業(yè)等行業(yè)正在采用人工智能技術(shù)提高效率。雖然人工智能技術(shù)僅貢獻(xiàn)了一小部分營收,但從英偉達(dá)到微軟(Microsoft Corp.)和Meta Platforms公司,全球科技公司都在押注印度這個(gè)快速增長的經(jīng)濟(jì)體能夠成為一個(gè)替代中國的增長型市場(chǎng)。
黃仁勛表示:“印度生產(chǎn)和出口軟件。未來,印度會(huì)出口人工智能?!?/p>
在印度古吉拉特邦,信實(shí)工業(yè)一座1千兆瓦規(guī)模的新數(shù)據(jù)中心正在建設(shè)中,該數(shù)據(jù)中心將使用英偉達(dá)的Blackwell芯片,使其成為最早采用這種強(qiáng)大的新型芯片的數(shù)據(jù)中心之一。亞馬遜云科技(Amazon Web Services,AWS)等英偉達(dá)的客戶也開始使用這款產(chǎn)品,AWS預(yù)計(jì)使用新芯片的數(shù)據(jù)中心將在明年上線。戴爾科技公司(Dell Technologies Inc.)也表示,基于Blackwell的服務(wù)器將在2025年初大規(guī)模上市。
對(duì)于使用芯片開發(fā)人工智能軟件和服務(wù)的數(shù)據(jù)中心運(yùn)營商而言,英偉達(dá)的產(chǎn)品已經(jīng)變成了昂貴的商品。位于加州圣克拉拉的英偉達(dá)在8月承認(rèn),Blackwell芯片的生產(chǎn)難度超出預(yù)期。英偉達(dá)表示,公司正在進(jìn)行一些改進(jìn)以提高產(chǎn)量,即工廠生產(chǎn)的功能正常的芯片數(shù)量。
英偉達(dá)還表示,將幫助印度的馬恒達(dá)科技公司(Tech Mahindra Ltd.)開發(fā)一個(gè)印地語大語言模型,并與電商公司Flipkart合作開發(fā)對(duì)話式客戶服務(wù)系統(tǒng)。英偉達(dá)將與印度的醫(yī)療保健公司合作,幫助他們提高患者保健與研究的效率。
這家美國公司向微軟和谷歌(Google)等科技巨頭供應(yīng)芯片,用于開發(fā)人工智能,站在全球人工智能熱潮的前沿。今年,黃仁勛巡回世界各地,推動(dòng)各國和企業(yè)采用被他稱為“新工業(yè)革命”的人工智能技術(shù)。
二十年前,英偉達(dá)開始在印度南部的班加羅爾開展業(yè)務(wù),并在印度其他三個(gè)城市設(shè)有研發(fā)中心,總計(jì)有約4,000名工程師,員工規(guī)模僅次于美國。
約一年前,英偉達(dá)與安巴尼的信實(shí)工業(yè)和塔塔集團(tuán)(Tata Group.)等印度綜合性企業(yè),達(dá)成了建設(shè)人工智能數(shù)據(jù)中心的初步協(xié)議。今年8月,安巴尼在信實(shí)工業(yè)的網(wǎng)絡(luò)直播股東大會(huì)上表示,公司正在開發(fā)名為JioBrain的一系列人工智能工具和應(yīng)用,他在會(huì)上至少80次提到了人工智能。
隨著中美緊張局勢(shì)升級(jí),印度對(duì)于全球科技公司的重要性提高。包括英偉達(dá)在內(nèi)的許多公司,因?yàn)槊绹南拗茰p少了與中國的業(yè)務(wù)往來。上個(gè)月,黃仁勛在美國與印度總理納倫德拉·莫迪會(huì)晤后,稱這是“印度的時(shí)刻”。
盡管印度的數(shù)字經(jīng)濟(jì)蓬勃發(fā)展,但其人工智能基礎(chǔ)設(shè)施仍在發(fā)展中。印度政府在“印度人工智能計(jì)劃“(IndiaAI Mission)下?lián)芸?2億美元,用于建設(shè)對(duì)構(gòu)建人工智能系統(tǒng)和技術(shù)商業(yè)化至關(guān)重要的數(shù)據(jù)中心。(財(cái)富中文網(wǎng))
譯者:劉進(jìn)龍
審校:汪皓
英偉達(dá)(Nvidia Corp.)的黃仁勛與亞洲首富穆克什·安巴尼達(dá)成了一項(xiàng)合作協(xié)議,將在全球人口最多的國家印度建設(shè)人工智能基礎(chǔ)設(shè)施和推廣人工智能技術(shù)。
上周四,這兩位高管在英偉達(dá)于孟買召開的人工智能峰會(huì)上同臺(tái)亮相,他們表示,安巴尼的信實(shí)工業(yè)(Reliance Industries Ltd.)在印度新建的大型數(shù)據(jù)中心,將使用美國公司英偉達(dá)的最新Blackwell芯片。英偉達(dá)還與印孚瑟斯(Infosys Ltd.)和塔塔咨詢服務(wù)公司(Tata Consultancy Services Ltd.)等多家印度綜合性企業(yè)建立了合作關(guān)系。
作為一個(gè)擁有14億人口的大國,印度已經(jīng)成為一個(gè)重要的人工智能市場(chǎng),該國的農(nóng)業(yè)、教育和制造業(yè)等行業(yè)正在采用人工智能技術(shù)提高效率。雖然人工智能技術(shù)僅貢獻(xiàn)了一小部分營收,但從英偉達(dá)到微軟(Microsoft Corp.)和Meta Platforms公司,全球科技公司都在押注印度這個(gè)快速增長的經(jīng)濟(jì)體能夠成為一個(gè)替代中國的增長型市場(chǎng)。
黃仁勛表示:“印度生產(chǎn)和出口軟件。未來,印度會(huì)出口人工智能?!?/p>
在印度古吉拉特邦,信實(shí)工業(yè)一座1千兆瓦規(guī)模的新數(shù)據(jù)中心正在建設(shè)中,該數(shù)據(jù)中心將使用英偉達(dá)的Blackwell芯片,使其成為最早采用這種強(qiáng)大的新型芯片的數(shù)據(jù)中心之一。亞馬遜云科技(Amazon Web Services,AWS)等英偉達(dá)的客戶也開始使用這款產(chǎn)品,AWS預(yù)計(jì)使用新芯片的數(shù)據(jù)中心將在明年上線。戴爾科技公司(Dell Technologies Inc.)也表示,基于Blackwell的服務(wù)器將在2025年初大規(guī)模上市。
對(duì)于使用芯片開發(fā)人工智能軟件和服務(wù)的數(shù)據(jù)中心運(yùn)營商而言,英偉達(dá)的產(chǎn)品已經(jīng)變成了昂貴的商品。位于加州圣克拉拉的英偉達(dá)在8月承認(rèn),Blackwell芯片的生產(chǎn)難度超出預(yù)期。英偉達(dá)表示,公司正在進(jìn)行一些改進(jìn)以提高產(chǎn)量,即工廠生產(chǎn)的功能正常的芯片數(shù)量。
英偉達(dá)還表示,將幫助印度的馬恒達(dá)科技公司(Tech Mahindra Ltd.)開發(fā)一個(gè)印地語大語言模型,并與電商公司Flipkart合作開發(fā)對(duì)話式客戶服務(wù)系統(tǒng)。英偉達(dá)將與印度的醫(yī)療保健公司合作,幫助他們提高患者保健與研究的效率。
這家美國公司向微軟和谷歌(Google)等科技巨頭供應(yīng)芯片,用于開發(fā)人工智能,站在全球人工智能熱潮的前沿。今年,黃仁勛巡回世界各地,推動(dòng)各國和企業(yè)采用被他稱為“新工業(yè)革命”的人工智能技術(shù)。
二十年前,英偉達(dá)開始在印度南部的班加羅爾開展業(yè)務(wù),并在印度其他三個(gè)城市設(shè)有研發(fā)中心,總計(jì)有約4,000名工程師,員工規(guī)模僅次于美國。
約一年前,英偉達(dá)與安巴尼的信實(shí)工業(yè)和塔塔集團(tuán)(Tata Group.)等印度綜合性企業(yè),達(dá)成了建設(shè)人工智能數(shù)據(jù)中心的初步協(xié)議。今年8月,安巴尼在信實(shí)工業(yè)的網(wǎng)絡(luò)直播股東大會(huì)上表示,公司正在開發(fā)名為JioBrain的一系列人工智能工具和應(yīng)用,他在會(huì)上至少80次提到了人工智能。
隨著中美緊張局勢(shì)升級(jí),印度對(duì)于全球科技公司的重要性提高。包括英偉達(dá)在內(nèi)的許多公司,因?yàn)槊绹南拗茰p少了與中國的業(yè)務(wù)往來。上個(gè)月,黃仁勛在美國與印度總理納倫德拉·莫迪會(huì)晤后,稱這是“印度的時(shí)刻”。
盡管印度的數(shù)字經(jīng)濟(jì)蓬勃發(fā)展,但其人工智能基礎(chǔ)設(shè)施仍在發(fā)展中。印度政府在“印度人工智能計(jì)劃“(IndiaAI Mission)下?lián)芸?2億美元,用于建設(shè)對(duì)構(gòu)建人工智能系統(tǒng)和技術(shù)商業(yè)化至關(guān)重要的數(shù)據(jù)中心。(財(cái)富中文網(wǎng))
譯者:劉進(jìn)龍
審校:汪皓
Nvidia Corp.’s Jensen Huang struck a partnership with Asia’s richest man, Mukesh Ambani, to build out artificial intelligence infrastructure and spur the technology’s adoption in the world’s most populous country.
The two executives shared the stage at Nvidia’s AI summit in Mumbai on Thursday and said a new major data center by Ambani’s Reliance Industries Ltd. in India is set to use the latest Blackwell chips from the US company. Nvidia also forged partnerships with Indian conglomerates including Infosys Ltd. and Tata Consultancy Services Ltd.
India has emerged as a potentially major AI arena, with the country of 1.4 billion adopting the technology in industries including agriculture, education and manufacturing to boost efficiency. While still a small part of their revenue, global tech companies from Nvidia to Microsoft Corp. and Meta Platforms Inc. are betting on the rapidly-growing economy as a growth market that can emerge as an alternative to China.
“India produced and exported software,” Huang said. “In the future, India will export AI.”
A 1 gigawatt Reliance data center under construction in Gujarat state will use Nvidia’s Blackwell, making it among the first to deploy the powerful new chips. Nvidia’s customers such as Amazon Web Services are also in the process of starting to use the product, with AWS expecting them to be online next year. Dell Technologies Inc. also has said that Blackwell-based servers will be generally available at the beginning of 2025.
Nvidia products have become a prized commodity among data center operators, which use the chips to develop AI software and services. The Santa Clara, California-based company acknowledged in August that Blackwell proved more difficult to produce than anticipated. The company said that it was making changes to improve its manufacturing yield — the number of functioning chips that come out of factories.
Nvidia also said it’ll help India’s Tech Mahindra Ltd. to build a Hindi large language model, and work with e-commerce company Flipkart on its conversational customer-service systems. It’ll collaborate with India’s health-care companies to help them improve productivity in patient care and research.
The US company has emerged at the forefront of a global AI boom, supplying the chips tech leaders like Microsoft and Google use to develop artificial intelligence. Huang has toured the globe this year, pushing countries and enterprises to adopt AI technologies he’s dubbed a “new industrial revolution.”
Nvidia began its operations in Bangalore, southern India, two decades ago and also has development centers in three other cities in the country, with a total of about 4,000 engineers, its largest employee base after its home country.
About a year ago, it struck initial pacts to build AI data centers with local conglomerates including Ambani’s Reliance and the Tata Group. Reliance Industries is building a range of AI tools and applications called JioBrain, Ambani said at the company’s live-streamed shareholders meeting in August, during which he mentioned the term AI at least 80 times.
India has risen in prominence for global tech companies as the US’s tensions with China have escalated. Nvidia is among companies whose business with China has been curtailed by Washington’s restrictions. Huang described “India’s moment” after meeting Prime Minister Narendra Modi in the US last month.
While India boasts a burgeoning digital economy, its AI infrastructure is still developing. The government has set aside $1.2 billion under the IndiaAI Mission to build data centers vital to building AI systems and commercializing technologies.