黃仁勳越來越會玩了：不僅化身數字人，還要造“第二顆地球”

{"type":"doc","content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"blockquote","content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"在2021年度GTC大會上，英偉達創始人兼CEO黃仁勳公佈了一系列最新技術，從自動駕駛汽車到虛擬化身、再到機器人乃至地球氣候建模，英偉達正着力拓展真實與虛擬世界的模擬邊界。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"英偉達認爲，黃仁勳此次公佈的新技術，有望顛覆價值數萬億美元的龐大產業體系。"}]}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"heading","attrs":{"align":null,"level":2},"content":[{"type":"text","text":"英偉達的元宇宙暢想"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"黃仁勳首先展示了英偉達Omniverse（用於3D工作流程的虛擬世界模擬與協作平臺）如何將英偉達引以爲傲的技術結合起來。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"黃仁勳表示，“我們的發展主題始終不變 — Omniverse將在倉儲、車間與車房、物理及生物系統、5G邊緣場景、機器人、自動駕駛汽車乃至化身的數字孿生等層面提供模擬功能。”"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"關於Omniverse，黃仁勳強調，“我們現在已經擁有了能夠創建新的3D世界，或者對真實物理世界進行建模的技術。”"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"據悉，Omnivers 的新能力包括："}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"bulletedlist","content":[{"type":"listitem","attrs":{"listStyle":null},"content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"Omniverse Avatar連接了NVIDIA的語音AI、計算機視覺、自然語言理解、推薦引擎和模擬方面技術，是一個能夠用於生成交互式AI虛擬化身的技術平臺。"}]}]}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"bulletedlist","content":[{"type":"listitem","attrs":{"listStyle":null},"content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"Omniverse Replicator是能夠生成用於深度神經網絡的合成數據的強大引擎。"}]}]}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"Omniverse 內的新功能包括："}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"bulletedlist","content":[{"type":"listitem","attrs":{"listStyle":null},"content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"NVIDIA CloudXR 是一個企業級沉浸式串流框架，已經集成到 Omniverse Kit（用於構建原生 Omniverse 應用程序和微服務的工具包），允許用戶以交互方式將 Omniverse 體驗串流至他們的移動 AR 和 VR 設備。"}]}]},{"type":"listitem","attrs":{"listStyle":null},"content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"Omniverse VR 推出領先的完全圖像、實時光線追蹤 VR – 支持開發者在平臺上構建自己的 VR 工具，最終用戶可以直接享受 VR 功能。"}]}]},{"type":"listitem","attrs":{"listStyle":null},"content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"Omniverse Remote 提供 AR 功能和虛擬攝像頭，支持設計師查看他們通過 iOS 和 Android 設備全光線追蹤的資產。"}]}]},{"type":"listitem","attrs":{"listStyle":null},"content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"Omniverse Farm 允許團隊同時使用多個工作站或服務器，爲渲染、合成數據生成或文件轉換等任務提供支持。"}]}]},{"type":"listitem","attrs":{"listStyle":null},"content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"Omniverse Showroom 作爲 Omniverse 公測版中的應用程序提供，允許非技術用戶使用 Omniverse 技術演示，展示平臺的實時物理和渲染技術。"}]}]}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"NVIDIA 還宣佈 Omniverse Enterprise 已進入正式發佈階段。它允許跨多個軟件套件工作的全球 3D 設計團隊在共享虛擬空間中通過任意設備實時協作。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"自去年 12 月發佈 Omniverse 公測版以來，已有 70000 多個創作者下載。此外，還有 700 多家公司的專業人士使用。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"此次發佈盡顯英偉達對元宇宙的暢想和願景。Bernstein的分析師斯Stacy Rasgon認爲，黃仁勳的演講將進一步激發人們對於元宇宙的興奮感。"}]},{"type":"heading","attrs":{"align":null,"level":2},"content":[{"type":"text","marks":[{"type":"strong"}],"text":"Omniverse Avatar"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong"}],"text":"爲了幫助開發者們使用Omniverse創建出能夠觀察、言說、就各類主題開展討論並理解自然語言表達含義的交互式角色，英偉達宣佈推出Omniverse Avatar。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"Omniverse Avatar是一個生成交互式AI虛擬形象的技術平臺，英偉達表示，這些技術可以用於支持一系列虛擬代理。在該平臺上創建的虛擬形象是具有光線追蹤3D圖像效果的交互式角色，可以看到、說話、談論各種主題，以及合理地理解表達意圖。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"Omniverse Avatar 所使用的語音AI、計算機視覺、自然語言理解、推薦引擎、面部動畫和圖像元素，通過以下技術實現："}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"bulletedlist","content":[{"type":"listitem","attrs":{"listStyle":null},"content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"其語音識別基於NVIDIA Riva。這個軟件開發工具包可以識別多種語言的語音。Riva還可通過文本語音轉換功能生成接近真人的語音反饋。"}]}]},{"type":"listitem","attrs":{"listStyle":null},"content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"其自然語言理解基於Megatron 530B大型語言模型。該模型可以識別、理解和生成人類語言。Megatron 530B是一個預訓練模型，它可以在很少或沒有訓練的情況下完成句子、回答主題廣泛的問題、總結長而複雜的故事、翻譯成其他語言，並應對許多它未接受過專門訓練的情景。"}]}]},{"type":"listitem","attrs":{"listStyle":null},"content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"其推薦引擎由NVIDIA Merlin™提供。該框架使企業可以建立能夠處理大量數據的深度學習推薦系統，以提出更明智的建議。"}]}]},{"type":"listitem","attrs":{"listStyle":null},"content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"其感知功能來自用於視頻分析的計算機視覺框架——NVIDIA Metropolis。"}]}]},{"type":"listitem","attrs":{"listStyle":null},"content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"其虛擬形象動畫由AI賦能的2D和3D面部動畫和渲染技術——NVIDIA Video2Face和Audio2Face™驅動。"}]}]}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"這些技術被集成在一個應用，並使用NVIDIA統一計算框架進行實時處理。這些技術被打包成可擴展、自定義微服務後，能夠通過NVIDIA Fleet Command™在多個地點安全地部署、管理和協調。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"Omniverse Avatar可以根據各個行業的需求定製。AI助手可以幫助處理數十億次日常客戶服務互動，如餐廳訂單，銀行交易，個人預約和預訂等。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"黃仁勳展示了Project Maxine for Omniverse Avatar如何將計算機視覺、Riva語音AI以及化身動畫與圖像結合起來，構成一個能夠實時對話的AI機器人。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"image","attrs":{"src":"https:\/\/static001.infoq.cn\/resource\/image\/f2\/b9\/f24546e61a67b7897be653cf7894ccb9.png","alt":null,"title":"","style":[{"key":"width","value":"75%"},{"key":"bordertype","value":"none"}],"href":"","fromPaste":false,"pastePass":false}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":"center","origin":null},"content":[{"type":"text","marks":[{"type":"size","attrs":{"size":10}}],"text":"黃仁勳展示了Project Maxine for Omniverse Avatar如何將計算機視覺、Riva語音AI、化身動畫與圖像共同結合起來，構成一個能夠實時對話的AI機器人——圖中爲黃仁勳的Omniverse Avatar玩偶。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"他還帶來了Project Tokkio演示，這是餐廳銷售臺場景中的客戶服務化身，能夠看到客戶、與兩位客戶同時交談並理解對方的意圖。該角色使用面部跟蹤技術與顧客保持眼神交流，並對他們的面部表情做出迴應。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"不過，黃仁勳在演示中指出，虛擬化身的響應時間只有兩秒，比人類的響應時間還慢，如果顧客趕時間，勢必會引起不滿。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"黃仁勳還展示了Project Maxine爲虛擬協作及內容創建類應用提供的先進視頻與音頻功能。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"在一段演示中，一位女士在嘈雜的咖啡廳內使用英語進行視頻通話，但接聽方聽到的卻是消去背景噪音、清晰順暢的語音結果。另外，她說的話可以被實時轉錄並翻譯爲法語、德語和西班牙語。更有趣的是，在Omniverse的幫助下，所有這些翻譯後的語言內容都由虛擬化身以與她相同的聲音和語調錶達了出來。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"黃仁勳表示：“智能虛擬助手的前景一片光明。Omniverse Avatar將我們的基礎圖像、模擬和AI技術結合在一起，打造出一些有史以來最複雜的實時應用。協作機器人和虛擬助手的案例令人難以置信，影響深遠。”"}]},{"type":"heading","attrs":{"align":null,"level":2},"content":[{"type":"text","marks":[{"type":"strong"}],"text":"Omniverse Replicator"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong"}],"text":"爲了幫助開發人員獲取訓練AI模型所需要的大量數據，英偉達宣佈推出Omniverse Replicator——一套用於訓練深度神經網絡的合成數據生成引擎。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"英偉達共開發出兩種replicator：一種用於通用型機器人技術，即NVIDIA Isaac Sim™；另一種用於自動駕駛汽車，即NVIDIA DRIVE Sim™。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"真實世界數據是昂貴、費力、需要人工標記的，它容易出錯且不完整，而Omniverse Replicator擴展了這些數據，該引擎能夠創建大量、多樣化的精確物理數據以滿足自動駕駛汽車和機器人開發者的需求。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"這兩個Replicator應用使開發者能夠以人類無法做到的方式引導AI模型、填補現實世界的數據空白，並標記真值數據。在這些虛擬世界中產生的數據可以涵蓋各類不同場景，包括在現實世界中無法經常經歷或安全體驗的罕見和危險情況。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"使用這些數據構建的自動駕駛汽車和機器人可以在一系列虛擬環境中掌握技能，然後再應用到物理世界中。Omniverse Replicator已成爲NVIDIA DRIVE自動駕駛汽車團隊和Isaac機器人團隊的重要數據生成引擎。該引擎將於明年上線，幫助開發者構建特定領域數據生成引擎。"}]},{"type":"heading","attrs":{"align":null,"level":2},"content":[{"type":"text","marks":[{"type":"strong"}],"text":"AI模型與系統"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"AI大模型毫無疑問是今年的熱點。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"演講中，黃仁勳介紹了英偉達在預訓練大模型方面的進展。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"黃仁勳介紹了英偉達的訓練大型語言模型Nemo Megatron（威震天）。他表示這類模型“將成爲有史以來規模最大的主流高性能計算（HPC）應用”。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"新的NeMo威震天框架可以使企業克服訓練複雜的自然語言處理模型的挑戰，訓練具有數萬億參數的語言模型。結合英偉達DGX™系統，這些工具提供了一個生產就緒的企業級解決方案，以簡化大型語言模型的開發和部署。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"作爲現代數據科學中的一種重要數據結構，大家現在可以利用Deep Graph Library（簡稱DGL）這個全新Python包將圖數據投射至各類深度神經網絡框架當中。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"在高性能計算領域，英偉達發佈了一個可以構建並訓練出能夠學習並遵循物理定律的機器學習模型Modulus。它是爲構建能夠大規模執行物理模擬的人工智能而設計的。這些模型的數據既來自真實世界，也來自對數字孿生的模擬。模數的設計充分利用了英偉達計算架構在大型超級計算機上的優勢。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"能夠適應各類工作負載的推理服務器Triton也迎來升級，現在能夠推理森林模型併爲大型語言模型執行多GPU多節點推理。Triton目前有2.5萬多家客戶在使用。Triton Inference Server軟件的關鍵更新包括與Amazon SageMaker的新集成，對Arm cpu和多gpu多節點功能的支持。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"黃仁勳還介紹了三個新庫。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"bulletedlist","content":[{"type":"listitem","attrs":{"listStyle":null},"content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"ReOpt——用於價值10萬億美元的物流產業。"}]}]},{"type":"listitem","attrs":{"listStyle":null},"content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"cuQuantum – 用於加速量子計算研究。"}]}]},{"type":"listitem","attrs":{"listStyle":null},"content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"cuNumeric – 用於幫助Python社區的科學家、數據科學家、機器學習及AI研究人員們實現NumPy加速。"}]}]}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"爲了將英偉達AI技術服務進一步推向邊緣場景，黃仁勳又帶來了英偉達Launchpad。目前英偉達正與全球數據中心巨頭Equinix合作，計劃將英偉達AI技術預安裝並整合至全球多處數據中心。"}]},{"type":"heading","attrs":{"align":null,"level":2},"content":[{"type":"text","marks":[{"type":"strong"}],"text":"全棧、數據中心規模的開放平臺"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"黃仁勳表示，加速計算催生出現代AI，而由此掀起的浪潮正在席捲整個科學及工業世界。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"這個時代始於三種芯片——GPU、CPU與DPU，各類系統（DGX、HGX、EGX、RTX及AGX）也開始由雲端擴展至更多邊緣地帶。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"英偉達爲全球300萬開發者創建出150個加速庫，用途涵蓋圖形、AI、科學及機器人等多個領域。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"在本次GTC大會上，英偉達再次發佈65項全新及更新SDK，包括CuNumeric——一個爲Python社區的科學家、數據科學家和機器學習研究人員加速NumPy的新庫。它實現了NumPy API，可以自動擴展到多gpu和多節點系統，無需任何代碼更改。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"黃仁勳指出，“英偉達加速計算已經成爲一套全棧、數據中心規模的開放平臺。”"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"他還同時介紹了“有史以來最先進的網絡平臺”— 英偉達Quantum-2，將配合BlueField-3 DPU迎接雲原生超級計算新時代。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"Quantum-2由Nvidia Quantum-2交換機、ConnectX-7網絡適配器和BlueField-3數據處理單元和支持軟件組成。據稱，Quantum-2能夠提供雲計算服務商及超級計算中心所需要的極致性能、廣泛的可訪問性及強大的安全保障。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"黃仁勳說，隨着超級計算中心開始向更多研究人員開放訪問，雲計算供應商開始提供超級計算服務，英偉達的Quantum-2平臺，爲超級計算機架構提供了重要的改變。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"“Quantum-2是第一個提供超級計算機性能和雲計算共享能力的網絡平臺”， 黃仁勳說，“這在以前是不可能的。有了Quantum-2，你的超級計算機將是原生雲計算，並且得到更好的利用。”"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"如今，網絡安全已經成爲企業乃至全球各國面臨的首要問題，而黃仁勳則公佈了一套基於三大支柱的零信任應對框架。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"其中BlueField DPU負責將應用程序與基礎設施隔離開來，作爲最新BlueField SDK的英偉達DOCA 1.2用於支持下一代分佈式防火牆；而英偉達Morpheus則假設入侵者已經駐留在內部系統當中，並使用“加速計算與深度學習等超能力檢測對方活動”。"}]},{"type":"heading","attrs":{"align":null,"level":2},"content":[{"type":"text","marks":[{"type":"strong"}],"text":"機器人"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"英偉達的Isaac生態系統目前已經吸引到全球700多家企業及合作伙伴，而且參與者數量過去四年內已經快速增長了5倍。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"黃仁勳宣佈，英偉達Isaac機器人平臺現可輕鬆被集成至機器人操作系統（簡稱ROS）當中，與後者中得到廣泛應用的機器人應用軟件庫及工具順暢協同。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"以Omniverse爲基礎構建而成的Isaac Sim則成爲有史以來仿真度最高的機器人模擬器。黃仁勳強調，“我們的目標是讓機器人分不清自己身處模擬世界、還是真實場景。”"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"爲了達成這個目標，Isaac Sim Replicator能夠生成合成數據以訓練機器人。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"Replicator模擬傳感器會生成經過自動標記的數據，並使用域隨機化引擎創建出豐富多樣的訓練數據集。"}]},{"type":"heading","attrs":{"align":null,"level":2},"content":[{"type":"text","marks":[{"type":"strong"}],"text":"自動駕駛汽車"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"黃仁勳認爲，“到2024年，絕大多數新的電動汽車都將具備強大的自動駕駛能力。”而展望未來，一切車輛都將在很大程度上、徹底完全實現自主運作。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"英偉達DRIVE是英偉達面向自動駕駛汽車打造的全棧開放平臺，而Hyperion 8則是英偉達的最新完整硬件與軟件架構。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"DRIVE Hyperion 8是一款用於全自動駕駛系統的計算機架構和傳感器。最新一代的技術是爲最高水平的功能安全而設計的。其傳感器套件包含12個攝像頭、9個雷達、12個超聲波雷達與1個前置激光雷達，所有信息均由兩個英偉達Orin SoC負責處理。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"黃仁勳還詳細介紹了Hyperion中內置的幾項新技術，包括Omniverse Replicator for DRIVE Sim——一種基於Omniverse的自動駕駛汽車合成數據生成器。DRIVE Sim可以縮小虛擬世界和現實世界之間的差距，提高逼真度，幫助自動駕駛汽車更好地感知和導航環境，並對快速變化做出反應。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"英偉達目前還運行有一套基於Hyperion 8傳感器、4d感知以及深度學習技術的全新多傳感器融合、特徵跟蹤與規劃引擎。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"以這些技術成果爲技術，汽車的內部也將迎來革命性的變化。英偉達Maxine技術將重新設計我們與車輛之間的交互方式。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"黃仁勳總結道，“有了Maxine，汽車將變身爲貼心的禮賓服務”。 他宣佈，DRIVE Hyperion可以用於2024年的車型。"}]},{"type":"heading","attrs":{"align":null,"level":2},"content":[{"type":"text","marks":[{"type":"strong"}],"text":"Earth Two"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"最後，黃仁勳宣佈英偉達將打造E-2項目，即Earth Two地球二號，用以模擬並預測蔚藍星球上的氣候變化。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"“地球2號”是英偉達於6月推出的價值1億美元的超級計算機“劍橋1號”之後的又一個新產品。“劍橋1號”可供醫療保健研究人員使用。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"他甚至做出斷言，“截至目前，我們發明的一切技術都是讓Earth Two成爲可能的必要條件，我想不出比這更偉大、更重要的消息了。”"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"英偉達沒有詳細說明E2的架構，但黃仁勳表示，要對地球氣候進行足夠詳細的建模，以準確預測未來10年、20年或30年的氣候，是一個極其困難的問題，氣候模擬比天氣模擬困難得多。英偉達正在使用其新的模數框架開發物理機器學習模型來解決這個問題。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"儘管地球2號的暢想充滿想象力和富有遠見，不過有分析人士認爲，細節的缺乏讓人懷疑E2是否真的存在。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"size","attrs":{"size":10}}],"text":"參考鏈接："}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"link","attrs":{"href":"https:\/\/www.nvidia.cn\/gtc-global\/keynote\/","title":"","type":null},"content":[{"type":"text","text":"https:\/\/www.nvidia.cn\/gtc-global\/keynote\/"}],"marks":[{"type":"size","attrs":{"size":10}}]}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"link","attrs":{"href":"https:\/\/blogs.nvidia.com\/blog\/2021\/11\/09\/nvidia-ceo-accelerated-computing-ai-omniverse-avatars-robots-gtc\/","title":"","type":null},"content":[{"type":"text","text":"https:\/\/blogs.nvidia.com\/blog\/2021\/11\/09\/nvidia-ceo-accelerated-computing-ai-omniverse-avatars-robots-gtc\/"}],"marks":[{"type":"size","attrs":{"size":10}}]}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"link","attrs":{"href":"https:\/\/mp.weixin.qq.com\/s\/T2aejAz_3EdMRjjrfoMrGw","title":"","type":null},"content":[{"type":"text","text":"https:\/\/mp.weixin.qq.com\/s\/T2aejAz_3EdMRjjrfoMrGw"}],"marks":[{"type":"size","attrs":{"size":10}}]}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"link","attrs":{"href":"https:\/\/mp.weixin.qq.com\/s\/oh3jynSWPlHsuxJJvpocGg","title":"","type":null},"content":[{"type":"text","text":"https:\/\/mp.weixin.qq.com\/s\/oh3jynSWPlHsuxJJvpocGg"}],"marks":[{"type":"size","attrs":{"size":10}}]}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"link","attrs":{"href":"https:\/\/mp.weixin.qq.com\/s\/504S7mDzkDeJx_wxnHcoZg","title":"","type":null},"content":[{"type":"text","text":"https:\/\/mp.weixin.qq.com\/s\/504S7mDzkDeJx_wxnHcoZg"}],"marks":[{"type":"size","attrs":{"size":10}}]}]}]}