What AI infrastructure and model deployments did Nvidia and its partners announce around AI Day Singapore on September 22–23, including AolaIllustrative AI-generated image; it does not depict Aolani’s planned factories.
AI 提示詞
Create a landscape editorial hero image for this Studio Global article: What AI infrastructure and model deployments did Nvidia and its partners announce around AI Day Singapore on September 22–23, including Aola. Article summary: Around AI Day Singapore on September 22–23, 2026, Nvidia and its partners presented plans for more regional AI computing capacity and wider use of its models. The clearest quantified infrastructure announcement was Aolan. Topic tags: general, general web, user generated. Style: premium digital editorial illustration, source-backed research mood, clean composition, high detail, modern web publication hero. Use reference image context only for broad subject, composition, and topical grounding; do not copy the exact image. Avoid: logos, brand marks, copyrighted characters, real person likenesses, fake screenshots, UI text, readable text, watermarks, charts with fa
報導指出,Sea Limited是東南亞國協(ASEAN)首家採用NVIDIA Vera Rubin平台的企業。Sea計畫在包括Shopee、Monee及Garena的業務中,利用該平台擴展AI模型與智慧代理。不過,現有報導未列出各業務的確定工作負載或正式投入運作日期,不能據此認定三者已全面使用Vera Rubin。1012
Nemotron:從公共安全研究到語言模型微調
在新加坡,負責支援公共安全等工作的科技機構HTX(Home Team Science and Technology Agency)正研究將Nemotron 3 Super與Nemotron 3 Nano Omni用於公共安全。NVIDIA指出,前者可能支援複雜推理與代理式工作流程,後者則具備結合視覺、音訊及語言的能力。這仍是研究與潛在用途,並非已確認的正式部署。5
在越南,Viettel AI正微調Nemotron,以用於越南語應用;馬來西亞的YTL AI Labs則為企業及民眾服務微調模型。現有資料未能確認泰國有哪一家特定機構部署了Nemotron,因此不宜替泰國項目指定合作對象。11
比較各項宣布,先看清數字屬於誰
2.2萬顆GPU及2027年初的時間表,僅屬Aolani這次的擴建計畫,不能套用到其他區域項目。NVIDIA較早前曾表示,Firmus Technologies正擴大其在澳洲及東南亞的AI工廠布局,並已與ST Telemedia Global Data Centres合作,在新加坡部署AI基礎設施;但該資料沒有提供可與Aolani相比的9月GPU承諾數字。現有資料同樣不足以確認AWS有與這批宣布相連的特定區域GPU數量或時程。4023