著名芯片設計師吉姆·凱勒加入了專門研發(fā)AI處理器的多倫多初創(chuàng)公司——Tenstorrent。
凱勒曾經(jīng)幫助AMD、蘋果(Apple)和特斯拉公司(Tesla)在芯片領域?qū)崿F(xiàn)過重大突破。2020年6月,在加入英特爾(Intel)僅僅兩年后,他就因為“個人原因”辭去了高級副總裁一職。
凱勒曾經(jīng)入職過多家頂尖的科技公司,雖然都只是短暫停留,但也都留下了輝煌的印記。近期,英特爾正面臨著產(chǎn)品延期發(fā)布的窘境,并且,在移動計算領域錯失良機之后,英特爾仍然在努力試圖奪回行業(yè)主導地位。因此,近幾年內(nèi),凱勒對英特爾的影響可能都不會有明顯的體現(xiàn)。
達到禪的境界
凱勒將成為Tenstorrent的總裁兼首席技術官。這位經(jīng)驗豐富的芯片大神將帶領這家已經(jīng)成立近五年的公司實現(xiàn)其目標:開發(fā)出超越當前水平的機器學習芯片。Tenstorrent的策略是,利用軟件去更好地衡量單個芯片的計算能力。
Tenstorrent的聯(lián)合創(chuàng)始人及首席執(zhí)行官留比薩·巴季奇曾經(jīng)在AMD的芯片設計團隊工作過十多年,凱勒與他的任職期剛好重疊——2012年至2015年期間,凱勒幫助AMD研發(fā)出了如今表現(xiàn)優(yōu)異的Zen架構。巴季奇還曾經(jīng)在英偉達(Nvidia)和VLSI工作過。
凱勒潛在影響力可能會為這家初創(chuàng)公司吸引來更多的投資和行業(yè)內(nèi)的合作。在一份聲明中,凱勒表示:“Tenstorrent已經(jīng)取得了顯著的進展,憑借著前景絕佳的架構,我們很可能會成為新一代的計算行業(yè)巨頭。”
AI芯片之爭
AI芯片領域的競爭相當激烈。從老牌芯片行業(yè)巨頭AMD、英特爾和英偉達,到包括Graphcore和Run:AI在內(nèi)的一眾初創(chuàng)公司,都在積極地研發(fā)更優(yōu)質(zhì)的AI芯片。
各個公司選擇的突破點也各不相同。英偉達安培架構的AI芯片嘗試著利用540億個晶體管(半導體電路的微觀組成部件)同時運行多項任務。位于加州的初創(chuàng)公司Cerebras則在研發(fā)集成了1.2萬億個晶體管巨型芯片。晶體管數(shù)量的增加可能會提高單個芯片的計算能力,但也會消耗更多電能,并提高其他方面的設計難度。
Tenstorrent公司表示,2020年推出的AI處理器Grayskull每秒可執(zhí)行368萬億次操作。雖然在實際應用中,芯片的性能會受到許多因素的影響,但這一數(shù)字,就已經(jīng)把宣稱每秒可執(zhí)行130萬億次到260萬億次操作的英偉達Tesla T4處理器甩在了后面。(財富中文網(wǎng))
譯者:殷圓圓
著名芯片設計師吉姆·凱勒加入了專門研發(fā)AI處理器的多倫多初創(chuàng)公司——Tenstorrent。
凱勒曾經(jīng)幫助AMD、蘋果(Apple)和特斯拉公司(Tesla)在芯片領域?qū)崿F(xiàn)過重大突破。2020年6月,在加入英特爾(Intel)僅僅兩年后,他就因為“個人原因”辭去了高級副總裁一職。
凱勒曾經(jīng)入職過多家頂尖的科技公司,雖然都只是短暫停留,但也都留下了輝煌的印記。近期,英特爾正面臨著產(chǎn)品延期發(fā)布的窘境,并且,在移動計算領域錯失良機之后,英特爾仍然在努力試圖奪回行業(yè)主導地位。因此,近幾年內(nèi),凱勒對英特爾的影響可能都不會有明顯的體現(xiàn)。
達到禪的境界
凱勒將成為Tenstorrent的總裁兼首席技術官。這位經(jīng)驗豐富的芯片大神將帶領這家已經(jīng)成立近五年的公司實現(xiàn)其目標:開發(fā)出超越當前水平的機器學習芯片。Tenstorrent的策略是,利用軟件去更好地衡量單個芯片的計算能力。
Tenstorrent的聯(lián)合創(chuàng)始人及首席執(zhí)行官留比薩·巴季奇曾經(jīng)在AMD的芯片設計團隊工作過十多年,凱勒與他的任職期剛好重疊——2012年至2015年期間,凱勒幫助AMD研發(fā)出了如今表現(xiàn)優(yōu)異的Zen架構。巴季奇還曾經(jīng)在英偉達(Nvidia)和VLSI工作過。
凱勒潛在影響力可能會為這家初創(chuàng)公司吸引來更多的投資和行業(yè)內(nèi)的合作。在一份聲明中,凱勒表示:“Tenstorrent已經(jīng)取得了顯著的進展,憑借著前景絕佳的架構,我們很可能會成為新一代的計算行業(yè)巨頭。”
AI芯片之爭
AI芯片領域的競爭相當激烈。從老牌芯片行業(yè)巨頭AMD、英特爾和英偉達,到包括Graphcore和Run:AI在內(nèi)的一眾初創(chuàng)公司,都在積極地研發(fā)更優(yōu)質(zhì)的AI芯片。
各個公司選擇的突破點也各不相同。英偉達安培架構的AI芯片嘗試著利用540億個晶體管(半導體電路的微觀組成部件)同時運行多項任務。位于加州的初創(chuàng)公司Cerebras則在研發(fā)集成了1.2萬億個晶體管巨型芯片。晶體管數(shù)量的增加可能會提高單個芯片的計算能力,但也會消耗更多電能,并提高其他方面的設計難度。
Tenstorrent公司表示,2020年推出的AI處理器Grayskull每秒可執(zhí)行368萬億次操作。雖然在實際應用中,芯片的性能會受到許多因素的影響,但這一數(shù)字,就已經(jīng)把宣稱每秒可執(zhí)行130萬億次到260萬億次操作的英偉達Tesla T4處理器甩在了后面。(財富中文網(wǎng))
譯者:殷圓圓
Star chip designer Jim Keller has joined Tenstorrent, a Toronto startup that makes specialized computer processors for artificial intelligence.
Keller previously designed breakthrough chips at AMD, Apple, and Tesla. He left his most recent role as senior vice president at Intel in June 2020 after just two years citing “personal reasons.”
Keller’s career is filled with short but significant stints at major technology companies. His imprint on Intel, which has faced recent product delays and struggled to reclaim its industry dominance after missing the boat on mobile computing, may not be obvious for a few more years.
Achieving a Zen-like state
At Tenstorrent, Keller will assume the titles of president and chief technology officer. The veteran chip whiz will be tasked with guiding the almost five-year-old company toward achieving its goal: creating chips for training machine learning programs that leapfrog current designs. Tenstorrent’s strategy is using software to better allocate the use of computing power on each chip.
Ljubisa Bajic, Tenstorrent’s cofounder and CEO, spent more than a decade designing chips at AMD. His tenure overlapped with Keller’s when Keller helped design AMD’s now successful Zen architecture from 2012 to 2015. Bajic has also worked at Nvidia and VLSI.
Keller’s implicit endorsement could help the startup attract further investment and industry partners. “Tenstorrent has made impressive progress, and with the most promising architecture out there, we are poised to become a next gen computing giant,” Keller said in a statement.
A. I.-yai-yai
The race to create better chips for A.I. applications is crowded. Competitors range from established chip giants AMD, Intel, and Nvidia to a host of startups including Graphcore and Run:AI.
Each company has a different approach. Nvidia’s Ampere A.I. chip seeks to take on multiple tasks with 54 billion transistors, the microscopic building blocks of semiconductor circuitry. California-based startup Cerebras is, on the other hand, making chips that are physically huge, with 1.2 trillion transistors each. Putting more transistors on a chip may increase the computing power but also draws more power and complicates the design in other ways.
Tenstorrent says its Grayskull processor introduced last year is capable of 368 trillion operations per second on A.I. tasks. That’s more than the 130 trillion to 260 trillion operations claimed by Nvidia’s Tesla T4 system, although many variables go into a chip’s performance on real-world applications.