Tag
Following a $750M Series D, Modal CEO Erik Bernhardsson discusses how each shift in software changes both the AI applications that can be created and the infrastructure needed to build them.
Anthropic 将 Claude Cowork 与聊天功能统一后的体验正式向 Claude Enterprise 客户开放，支持跨网页、桌面和移动端的云端代理工作，并新增 Efficient mode 等管理员成本控制功能。
Anthropic announces Claude Duet, a new standalone app where Claude acts as a proactive agent powered by Opus 5.5, working across 6,000 connectors/plugins, briefing users each morning, completing handed-off tasks autonomously, and seeking approval before sensitive actions. It is opening as a research preview on a waitlist for Claude Pro and Max subscribers in the US.
OpenAI and Google Cloud are partnering to bring OpenAI's frontier models and Codex agentic workflows into Google Cloud's Model Garden, available via Google Cloud Marketplace. General availability is planned in the coming weeks, while Microsoft remains OpenAI's primary cloud partner.
Morgan Stanley estimates that hyperscale cloud providers and neoclouds will spend nearly $3.4 trillion on infrastructure across 2027 and 2028, including operating leases, potentially marking the largest wave of infrastructure spending in human history.
Anthropic's leaked S-1 filing reveals it is struggling to secure cloud capacity as AWS, Google, and Microsoft increasingly view it as a competitor, while planning over $750 billion in cloud spending amid big losses ahead of a reported November IPO.
Shadeform announced Flex VMs and Shadeform Spot, aiming to put idle high-demand GPUs to use and give builders easier access to affordable AI infrastructure and agent compute for experiments and post-training.
A Google Cloud outage disrupted banking services and numerous other operations across Europe, highlighting reliance on cloud infrastructure for critical financial systems.
Neoclouds are moving beyond pure GPU rental and building moats by acquiring software companies, because renting compute alone is no longer scarce and inevitably leads to price wars. Nscale, Nebius, IREN, CoreWeave, Crusoe and others are all making acquisitions to boost customer stickiness and product differentiation.
Tech Fund projects that annualized AI revenue in markets outside China will reach $414 billion by the end of this year, with Anthropic and OpenAI combined accounting for $187 billion. It also estimates an overall IRR of roughly 12% on the $931 billion in AI capex spent over the past 3.5 years.
Amazon Web Services 发布了 TorchNeuron 开发工具，让 PyTorch 应用仅需一行配置即可迁移到其 Trainium AI 芯片上，直接挑战 Nvidia 在 AI 软件生态上的主导地位，并将相关改动贡献给 PyTorch 上游项目。
FT reports that AI compute has become strategic infrastructure, with adversaries targeting data centres like power plants. Iranian strikes on AWS Bahrain zones and Ukrainian drone attacks on Yandex's AI supercomputer sites exposed the fragility of cloud and data-localisation resilience.
Gartner forecasts AI providers will spend nearly $1T on infrastructure in 2026 and $6T more from 2027 to 2030, warning that over $3.5T of that could face write-downs as revenue fails to cover costs. Asset-light AI-native providers and exposed neoclouds are highlighted as key risks, with AI-optimized IaaS growth deemed too slow to recoup investments before current chips are replaced.
A commenter analyzes the release of DeepSeek-V4.1-Flash-1017 as a positive for the AI industry's profitability: improved token cost-efficiency will drive up AI penetration and demand for cloud providers' infrastructure. Unlike the earlier DeepSeek shock, market conditions have since changed.
AWS 意外取消了 Cloudflare 在 re:Invent 的赞助展台，Cloudflare CEO Matthew Prince 称此举显示 AWS 将其视为竞争对手，并表示将自行举办场外活动。业界评论指出，大厂年度活动中取消赞助商展台极为罕见，凸显云计算市场竞争加剧。
Jaya Gupta and Ashu Garg argue the AI economy is shifting from model training and inference to agent execution — where CPU-based work like scheduled checks, storage, and transaction recovery dominates cost — and that application builders, not model or GPU providers, will control who gets paid.
Modal 宣布年化运行率突破 10 亿美元，并完成由 Accel 领投的 7.5 亿美元 D 轮融资，估值达 157.5 亿美元，自 C 轮以来使用量增长三倍、同比增长超 15 倍。
AWS announces Lambda Web Functions, a new form factor that lets developers deploy standard port-listening HTTP servers (Express, Hono, Fastify, Next.js) directly on Lambda with multi-concurrency, active CPU billing, and no adapters, alongside Event Functions and MicroVMs.
乌克兰无人机袭击摧毁了俄罗斯Yandex的多个数据中心，导致支付、银行、电商等数字服务大规模中断，暴露出俄罗斯超集中的数字生态系统在战时的脆弱性。
微软发布第三代自研 AI 芯片 Maia 300，采用 TSMC N2P 芯粒架构，能效较领先 GPU 系统提升 40%，并计划 2027 年起在 Azure 大规模部署，以降低对 Nvidia 和 AMD 的依赖。