PulseAugur
实时 03:09:29
English(EN) With the recent surge in AI mega clusters, reaching hundreds of thousands to millions of AI accelerators, cloud providers encountered a new set of challenges th

SemiAnalysis:跨域扩展技术将推动网络市场在2028年达到140亿美元 · 追踪7个来源

SemiAnalysis报告称,由NVIDIA推广的“跨域扩展”(scale-across)技术对于将多个数据中心互联成单个集群至关重要。这种方法解决了云服务提供商在AI大型集群激增时面临的挑战,这些集群需要数百万个加速器。虽然Microsoft和Meta目前对其使用跨域扩展技术最为关注,其中Meta使用了Prometheus集群,Microsoft采用了AI-Backbone基础设施,但SemiAnalysis预计该技术将在行业内得到广泛应用。预计该技术将显著推动网络设备市场,特别是ZR/ZR+相干光模块和光线路系统市场,市场价值预计将从2026年的30亿美元增长到2028年的140亿美元。 AI

影响 这项技术对于扩展AI基础设施至关重要,通过互联多个数据中心,能够实现更大、更强大的AI集群。

排序理由 一家研究公司对技术趋势的分析,而非来自主要科技公司的直接产品发布或公告。

在 X — SemiAnalysis 阅读 →

AI 生成摘要 · Google Gemini · 来自 7 个来源。 我们如何撰写摘要 →

SemiAnalysis:跨域扩展技术将推动网络市场在2028年达到140亿美元 · 追踪7个来源

报道来源 [7]

  1. X — SemiAnalysis TIER_1 English(EN) · SemiAnalysis_ ·

    欲了解更多信息,请查看我们的网络模型,我们最近在该模型上发布了一篇关于这一技术转变的详尽说明。(7/7) https://t.co/kqt0IiSiFi

    To find out more, check our networking model, where we recently published an extensive note around this technological shift. (7/7) https://t.co/kqt0IiSiFi

  2. X — SemiAnalysis TIER_1 English(EN) · SemiAnalysis_ ·

    总的来说,跨域技术应能显著推动网络设备上的内容,特别是 ZR/ZR+ 相干可插拔设备和光模块上的内容。

    Overall, the scale-across technology should meaningfully drive content on networking equipment, and more specially on ZR/ZR+ coherent pluggables and on Optical Line Systems. We currently estimate pure-scale across content to represent about $3B in 2026, and growing to about $14B

  3. X — SemiAnalysis TIER_1 English(EN) · SemiAnalysis_ ·

    虽然微软和Meta是唯一公开谈论跨规模的公司,但我们相信几乎所有提供商都会转向这项技术

    While Microsoft and Meta have been the only ones talking this openly around scale-across, we believe that almost all providers will shift towards this technology as is becomes harder to build datacenters large enough to host millions of chips, and to find a provider that can

  4. X — SemiAnalysis TIER_1 English(EN) · SemiAnalysis_ ·

    截至目前,只有Meta和Microsoft公开宣布了他们使用scale-across的巨型集群,包括Meta的Prometheus集群以及他们的新的

    As of now, only Meta and Microsoft have publicly communicated around their mega-clusters using scale-across, with the Prometheus cluster for Meta and their new AI-Backbone (AIBB) infrastructure, and Fairwater for Microsoft, who intends to run real time AI training over Atlanta ht…

  5. X — SemiAnalysis TIER_1 English(EN) · SemiAnalysis_ ·

    主要混淆之处在于,规模扩展可以使用与传统数据中心互连相同的网络设备。事实上,作为互连

    The main confusion comes from the fact that scale across can use the same network equipment as for traditional datacenter interconnect. Indeed, as interconnections often reach a few kilometers minimum, operators can use ZR/ZR+ coherent pluggables and Optical Line Systems, which h…

  6. X — SemiAnalysis TIER_1 English(EN) · SemiAnalysis_ ·

    Nvidia去年广泛推广了Scale-across,但现在常被不精确地用作泛指,几乎谈论任何数据中心互连

    Scale-across has been widely popularized by Nvidia last year, but is now often used imprecisely as a generic term to talk almost about any Datacenter interconnect network. To make it short, scale-across refers to backend datacenter interconnect networks that are used to connect h…

  7. X — SemiAnalysis TIER_1 English(EN) · SemiAnalysis_ ·

    随着近期AI巨型集群的激增,AI加速器数量达到数十万至数百万,云服务提供商面临着一套新的挑战

    With the recent surge in AI mega clusters, reaching hundreds of thousands to millions of AI accelerators, cloud providers encountered a new set of challenges that forced them to run chip interconnect at a new scale: interconnecting multiple datacenters together. This is called ht…