Your agent made it to production. Now the real work begins. Agents need a stack where the model, runtime, and infrastructure are designed together. That’s what Alibaba Cloud is building — the Qwen and Wan model family, model services, and an agent-native cloud layer — with the agent computing runtime running on AMD EPYC processors. Thanks to AMD for co-hosting Alibaba Cloud Agent Builder Day in Mountain View, and to HeyGen, Pika, OpenArt AI, Tinyfish, and our investor panel from Beta Fund, Hattrick Capital, Uphonest Capital, Llama Venture, and Translink Capital for sharing the stage. 100+ registered attendees. One clear signal: the agent-native stack is here. #AlibabaCloud# #AgentBuilderDay# #AgenticCloud# #AIAgents# #ShipTheNext#

The transition to Agentic Cloud has begun. Dr. Feifei Li, CTO and President of International Business at Alibaba Cloud, shared his perspective on what lies ahead: the next three years will be defined by the rise of the Agentic Cloud, where agents can leverage products and infrastructure to carry out sophisticated tasks.

FUTURUM'S FOUR LAYERS OF AGENTIC AI ECONOMY APPLICATIONS & AGENTS LAYER • $META scaling consumer agents through Muse & specialized models • $AXON putting AI on officers cameras with AI ARR now above $100M • $GOOGL bringing Gemini agents across Search, Android & Workspace • $MSFT distributing Copilot agents across world's largest enterprise customer base • $APP using next-gen AI models to price & target ads across gaming & e-commerce • $SHOP plugging merchants into every major shopping agent as AI-driven orders grow 3x • $PLTR giving AI agents operating context to act on real enterprise data through its ontology INFERENCE PLATFORMS & CLOUD LAYER • $IREN owning land, power & data centers under its AI cloud • $MSFT $AMZN $GOOGL $ORCL running largest AI clouds in world • $NBIS building vertically integrated AI cloud with zero Texas exposure • $CRWV providing specialized GPU cloud capacity for hyperscalers & labs • $NET running agents on isolates instead of containers as "the first agentic cloud" • $DOCN offering a simpler cloud for developers as AI agents start choosing infrastructure COMPUTE, NETWORKING, MEMORY & TIMING LAYER • $MU $SKHY supplying memory every AI system is short on • $CBRS delivering speed for inference with wafer-scale chips • $AVGO supplying Ethernet & custom AI chips for hyperscalers • $AEHR testing expensive chips before they go into production • $CRDO adding high-speed links as AI systems spread across more chips • $NVDA supplying accelerators, networking & software for AI infrastructure • $SITM keeping every chip, switch & optical link in sync with precision timing • $ARM collecting royalties on nearly every CPU challenging x86 in data center • $AMD powering agents with EPYC CPUs & Helios racks alongside Instinct GPUs • $ALAB connecting accelerators with content per chip roughly doubling each generation • $MRVL supplying custom AI chips & optical DSPs with its pending $LITE deal adding in-house lasers POWER, COOLING & PHYSICAL INFRASTRUCTURE LAYER • $GEV building gas turbines for self-powered campuses • $CEG $VST pairing data centers directly with nuclear power • $NVTS bringing GaN power chips into 800V AI data centers • $FPS supplying electrical gear that every power scenario needs • $BE generating on-site power for data centers that can't wait for grid • $VRT cooling & powering racks that get hotter & denser every generation • $ON converting power efficiently as data centers move to 800V architectures

Building what’s next for the agentic era. Join the Agentic Cloud Forum at #ApsaraConference2026# to explore how cloud infrastructure is evolving for AI and intelligent agents. Date & Time: October 29, 2026 | 14:00–16:05 CET Location: Beurs van Berlage, Amsterdam Register now: #ApsaraAmsterdam# #AgenticEra# #AIAgents# #AlibabaCloud#

🚀At Apsara2026, Alibaba Cloud introduced a suite of upgrates for its agentic cloud strategy, built around three pillars: Model, Harness, and Context. These upgrades drive AI innovation at scale, spanning AI Native Cloud, Agent Native Cloud, and Context Engine to empower enterprises in creating real-world value with cutting-edge AI technologies.

Agents should use tools that excel at what’s important for an AI model. Kitesurf is Cloudflare’s new stateless, highly scalable, and cost-effective web browser that runs entirely on top of Workers and was designed specifically for the Agentic Cloud. #AgentsWeek#

TRM Labs is hiring a Senior Cloud Security Engineer to own GCP IAM, lead TRM's FedRAMP 20x expansion, and drive the organization toward continuous, agentic cloud security posture management. If you have knowledge in Deep GCP, Terraform, and IAM experience, apply here 👉 #TRMCareers# #TRMLabs#

Shaping the future of intelligent AI. At #ApsaraConference2026#, Dr. Feifei Li, CTO of Alibaba Cloud, President of International Business, shared how Alibaba Cloud is advancing machine intelligence through world-class models, agentic cloud capabilities, and trusted AI infrastructure. Watch the full video here: Explore more: #ApsaraAmsterdam# #AgenticEra# #AIAgents# #AlibabaCloud#

Powering the next frontier of AI. At #ApsaraConference2026# in Amsterdam, Dr. Feifei Li, CTO of Alibaba Cloud, President of International Business, highlighted Alibaba Cloud's role in advancing open AI through scalable models and Agentic Cloud capabilities. Explore more: #ApsaraAmsterdam# #AgenticEra# #AIAgents# #AlibabaCloud#

What does it take to power the next generation of intelligent agents? At the Qwen Conference, Dr. Feifei Li, CTO and President of International Business at Alibaba Cloud, shared the four cornerstones driving agentic applications: Models, Agentic Cloud, Tools and Services, and Performance at Scale. Together, these pillars enable agents to reason, act autonomously, leverage tools like coding to accomplish complex tasks, and operate effectively at scale.