AI云算力AI cloud compute
租用远程AI计算资源
每条都摘自下方编号的来源原文;只写事实,不写预测或建议。
技术简介
- 远程可视化虚拟机提供远程GPU算力,使用户无需强大的本地硬件即可处理复杂图形任务。6
Remote Visualization: For applications requiring high-end visualization capabilities, such as CAD and visual effects, NC-series VMs can provide the necessary GPU power remotely, allowing users to work on complex graphical tasks without needing powerful local hardware.
- AI训练和推理需要高密度计算、先进网络、优化存储和管理复杂分布式系统的软件。7
Both training and inference require high-density compute, advanced networking, optimized storage, and software capable of managing complex distributed systems.
已有应用
- Azure NCads H100 v5虚拟机使用英伟达H100 NVL GPU和第四代AMD EPYC Genoa处理器。6
The NCads H100 v5 series virtual machines are powered by NVIDIA H100 NVL GPU and 4th-generation AMD EPYC™ Genoa processors.
- IREN称Horizon 1已为微软投入服务。1
“This quarter Horizon 1 went into service for Microsoft with NVIDIA Exemplar status, AI Cloud revenue nearly tripled to $200 million, and operating ARR doubled to a billion,” said Daniel Roberts, Co-Founder and Co-CEO of IREN.
- Cognition在CoreWeave云上运行Devin的训练、强化学习和生产推理。3
Cognition, the applied AI lab behind Devin, runs training, reinforcement learning and production inference for Devin on CoreWeave.
为何此时
- 微软在2026财年第四季度电话会上表示Azure客户需求继续超过供给。8
In Azure, we expect revenue growth of approximately 45% in constant currency and we remain focused on delivering efficiencies that help us bridge the gaps we see as customer demand continues to exceed supply.
- 微软财务负责人在2026年10月28日表示,需求仍超过供应。2
"We still have demand that exceeds supply," she said.
主要参与者
- CoreWeave部署了采用英伟达GB200和GB300 NVL72的戴尔PowerRack系统,并已部署Vera Rubin。3
CoreWeave, in collaboration with Dell Technologies, was among the first cloud providers to deploy Dell PowerRack systems featuring NVIDIA GB200 and GB300 NVL72, and is one of the first to deploy NVIDIA Vera Rubin.
- AMD的Helios将Instinct MI455X GPU、EPYC Venice CPU、Pensando网络及ROCm软件集成为大规模AI训练和推理机架平台。4
AMD Helios combines AMD Instinct™ MI455X GPUs, AMD EPYC™ “Venice” CPUs, Pensando™ networking and ROCm™ software in an open, integrated rackscale platform built for large-scale AI training and inference.
- IREN提供用于AI训练和推理的大规模GPU集群。5
IREN is a leading AI Cloud Service Provider, delivering large-scale GPU clusters for AI training and inference.
未定事项
- IREN将超过40亿美元合同年化收入对应产能的投运目标从年底推至1月,并称可能延至2月底,原因是客户要求改设计及供应商时点。1
“The more than $4 billion of contracted ARR is funded and now targeted for January, after a customer-requested design change and supplier timing moved it from year-end; it may extend to the end of February.
- 微软财务负责人在2026年10月28日表示,原本预期第一季度交付的部分第三方算力延期。2
"Some third-party capacity deliveries we expected in Q1 were delayed," the company's finance chief, Amy Hood, said on a conference call with analysts.
来源
- iren.gcs-web.com · IREN Reports Q1 FY27 Results | IREN · 2026-11-05
- CNBC · Microsoft stock jumps 4% as cloud growth drives earnings and revenue beat · 2026-10-28
- CoreWeave · CoreWeave Delivers NVIDIA Vera Rubin NVL72 Performance at Production Scale, Starting With Cognition · 2026-09-30
- newsroom.amd.com · Microsoft to Deploy Next-Gen AMD Instinct and AMD EPYC… · 2026-07-20
- iren.gcs-web.com · IREN Secures $9.7bn AI Cloud Contract with Microsoft | IREN · 2025-11-03
- learn.microsoft.com · NC family VM size series - Azure Virtual Machines
- SEC · crwv-20251231 · 2026-03-02
- microsoft.com · Microsoft Fiscal Year 2026 Fourth Quarter Earnings Conference Call