与「Neocloud」相关的搜索结果

1/ ClusterMAX 3.0 发布。 这次全面评测了 77 家 Neocloud，并将行业观察范围扩大到 323 家，还访谈了 200 多位真实用户。 这可能是目前最完整的一份 Neocloud 行业地图。👇

ClusterMAX 3.0 is here! ClusterMAX 3.0 debuts with a comprehensive review of the neocloud industry, covering 77 providers. We increase our market view to cover 323 providers, up from 209 in ClusterMAX 2.0, 169 in ClusterMAX 1.0, and 124 in the original AI Neocloud Playbook and Anatomy article. We have now interviewed well over 200 end users of neoclouds as part of this research. We update our itemized list of criteria across 10 categories, and update our direct descriptions of our expectations for Slurm, Kubernetes, Standalone Machines, Monitoring Dashboards, and Health Checks. All of this content is live on our website. We encourage providers to use these lists when developing their offerings. We still consider these lists as an amalgamation of our experience interviewing end users, making them representative of the features that end users expect from their cloud providers. Nebius joins CoreWeave in the Platinum tier. While CoreWeave still sets the technical bar for others to follow, Nebius is now established as a provider that consistently commands a premium pricing over others. Strong business decisions by Nebius have put them in a position to serve an entire class of neolabs at seller’s prices. Google Cloud joins Oracle in the Gold tier. Azure moves to Silver, Fluidstack moves to Unavailable, and Crusoe drops to Bronze. Lambda, Firmus and TensorWave remain in Silver, while GMI moves up to Silver from Bronze. Many companies drop from Silver (or Gold) to Bronze or lower. We raise the bar this round as only 19 neoclouds globally achieve a Medallion rating. We establish a tier between Bronze and Underperforming: the Participation Ribbon tier. 15 providers join this rating, which more accurately describes our opinion that they do the bare minimum to get by.

Open router for chips is a $10B+ company waiting to be built Nobody moves custom inference workloads off Nvidia because porting to AMD, Cerebras, Etched, etc. is painful But with frontier models now you could build a pipeline for translating workloads from one chip to another When that starts happening reliably there there will be a rise of multi-chip neoclouds. Neoclouds won’t buy non-Nvidia chips today because developers don’t use them. That will change once you eliminate the code compatibility problem and performance is the driving factor If you build it, the play is either open source it from day 1 and become a strategic acquisition target for a big neocloud, or keep it private and build your own cloud, becoming the de facto multi-chip provider

过去一段时间，只要谈到AI投资，大家最先想到的往往还是英伟达。 但我越来越觉得，如果只盯着一家芯片公司，我们看到的可能只是AI浪潮中最显眼的一部分。训练和运行AI模型，不仅需要GPU，还需要数据中心、电力、云端算力、高速数据传输，以及最终将AI带入现实世界的机器人。 换句话说，AI不是一家公司或一类芯片的故事，而是一场覆盖整条基础设施供应链的长期建设。这也是我开始关注 @reserveprotocol 的原因。下面我想说说Reserve到做了什么？ Reserve围绕去中心化代币基金，也就是DTF来构建产品和基础设施。它通过Ondo Globa Markets的代币化股票，将AI供应链中的不同公司组合成五个链上主题篮子，让符合资格的用户能够分别了解和接触AI建设的不同环节。而且我认为这次产品设计最值得关注的地方，不是简单地把股票搬到链上，而是把原本比较分散的AI基础设施主题，拆解成了五条更容易理解的投资方向。 1. AI Infrastructure（BUILDOUT） BUILDOUT覆盖相对完整的AI硬件和基础设施技术栈。如果把AI产业比作一座正在扩建的城市，BUILDOUT关注的就是修建这座城市所需的各种底层设施，而不是只押注其中一家明星公司。 2. AI Power（POWER） AI运行需要大量电力。数据中心规模越大、模型训练越复杂，电力供应就越重要。我认为，能源是AI讨论中经常被低估的一环。算法可以快速升级，但发电设施、电网接入和长期供电合同并不能在一夜之间完成。POWER关注的正是为AI运行提供能源支持的企业。 3. AI Capacity & Neocloud（NEOCLOUD） NEOCLOUD聚焦将高性能计算资源作为服务提供给客户的企业。不是每家公司都有能力自行购买大量GPU、建设数据中心并维护整套计算基础设施。因此，算力租赁和新型云服务正在成为AI产业链中的重要组成部分。在我看来，这一方向值得关注的原因，是它直接连接了算力供给与企业的实际AI需求。 4. AI Photonics（PHOTON） AI不仅需要计算，还需要在芯片、服务器和数据中心之间快速传输大量数据。PHOTON关注的是光子技术和高速数据传输，也就是帮助数据高效流动的光学层。 很多人关注一块芯片能算多快，但我认为，芯片之间能否高效交换数据，同样可能影响整个AI系统的性能。 5. Robotics（ROBOTS） ROBOTS关注AI从数字世界进入物理世界的过程，包括机器人、自动化设备及相关产业链。如果说大语言模型让我们看到了AI理解和生成信息的能力，那么机器人代表的就是AI感知环境、作出判断并执行现实任务的可能性。这也许是五个方向中最直观、同时最具想象空间的一条主线。 了解了上面5个方向，接下来说下我怎么看这次机会。 我认为，Reserve AI DTF最值得讨论的地方，是它把AI交易从猜下一家明星公司，转向了观察哪些基础设施正在真实建设。市场情绪可以在一个月内发生很大的变化，但已经签署的数据中心建设计划、电力合同和芯片订单，这属于持续数年的商业承诺，而这些基础设施支出不一定会因为市场短期下跌而立即停止。因此，把基础设施建设看作观察AI产业一个重要指标，我认为是很有必要的，而不是只关注短期价格和社交媒体热度。当然这些并不意味着基础设施相关资产不会下跌，也不能说明AI DTF能够完全规避市场的波动风险。我的观点是：市场情绪和产业建设不总是同步的，而这种差异本身会有一些只得研究的地方。 现在想想，在过去，想在链上通过单一篮子覆盖这些主题并不容易。而Reserve让这个方向开始具备更清晰的产品结构，也更适合让人去找到合适自己的方向并且找到适合自己的产品标的。 最后如果让你从AI基础设施中选择一个未来几年最值得研究的方向，你会选哪一个？ 是AI芯片和硬件、电力、云端算力、光子技术，还是机器人？ 我自己会特别关注电力和算力服务。原因很简单：无论最终由哪一家模型公司胜出，AI系统都需要持续消耗算力和能源。但这只是我的观察，并不代表这些方向一定会获得更好的市场表现。 但是我肯定会持续关注 @reserveprotocol 带来的一些优质资产和设施，对于我来说确实降低了研究传统资产的门槛。

CoreWeave只用约25个季度，就做到AWS商业化第40个季度才达到的单季26亿美元收入。 如今AWS、Azure等Hyperscaler的绝对收入仍高出几个数量级，但按相同发展阶段比较，Neocloud的增长速度明显更快。 AI算力需求，正在把云计算早期二十年的增长曲线重新压缩一遍。

"The hyperscalers generate orders of magnitude more revenue per quarter than the neoclouds, but at the same time, it’s taken CoreWeave ~25 quarters to achieve the same $2.6B as AWS generated by quarter 40 since launch. Again, these are very fast growing businesses" (h/t @MosesSternstein @a16z)

底层看稀缺，顶层看客户！看到 @chamath 分享的这张 AI Stack，感觉这个架构划分和 Rewire Index 5 Layer 相当类似，分享一下我对每一层的理解： 1. 能源与基础设施层（最底层） 与 Rewired Index 的逻辑一致。电力会迎来爆发式增长，尤其是无需接入电网的独立供电商（IPP）；而土地受政策影响太大，弹性有限。在大众的舆论压力之下，太空基建应该会是未来几年的新机会，无需土地，无限电力⚡️ 2. 芯片层 做独立芯片，初创公司基本没有机会：性能要求极高、工艺极其复杂，最关键的是供应链已被完全锁死——这是头部玩家的战场。 真正的机会在融合与生态。从 Google TPU 的发展路径，到 Cerebras 等高速推理芯片的崛起，可以看出芯片会与云厂商、模型公司深度绑定。围绕芯片构建数据中心的整体供应与创新，机会很多；单做独立芯片，机会渺茫。 3. 云服务层 云可以分为 Hyperscaler 和 NeoCloud两类。模型商品化之后，几乎所有的负载都要靠云来承载，这会是非常赚钱的生意。但构建极其复杂，堪称 AI 时代的重资产业务——或者说，智能时代的房地产。 4. 模型层 模型公司面临的核心问题是正在被商品化： - 如果 Scaling Law 已到极限，模型百分之百会被商品化； - 即使 Scaling Law 还有很大空间，大家对「最好智能」的需求也在被分解——大量日常应用不需要最顶级的智能，中等水平模型和开源模型会逐渐接管这些需求，反而加速了商品化； - 最尖端的头部模型公司，更像是「先进制程」的芯片：能从中获取很高的价值，但并非所有任务都需要它。 5. 应用层：Harness vs Application 这张图最有趣的地方，是把应用层拆成了 Harness 和 Application 两层。 我的判断是：按目前模型的进化速度，Application 还没有任何机会，但 Harness 的机会已经大量出现，尤其在企业端。企业内化的知识只能通过 Context 和约束来落地，所以 Harness 就是新的企业应用层；它们会替代旧的 SaaS，或倒逼旧 SaaS 升级。 如果把 AI 扩展到大语言模型之外的更广义范畴，应用层更可能以垂直集成的形态出现，例如：自动驾驶 / RoboTaxi / 任何可端到端自动化工业流程和武器系统；生物研究 / Wet Labs；把执行能力直接部署进企业内部的模式（类似 Palantir 的服务方式）

Former $CRWV employee on why neoclouds are far more exposed to GPU generation cycles than hyperscalers ( $MSFT, $AMZN, $GOOGL ): - The expert describes GPU utilization tracking at hyperscale as a continuous and disciplined process built around two lenses. The first is infrastructure utilization, covering GPU occupancy, idle time, and memory utilization, noting that 95% booked usage can still mask inefficiency if jobs stall or batches have idle gaps. The second is outcome utilization, asking whether the compute is actually generating business value, measured by metrics such as tokens trained per dollar, time to reach target accuracy, and tokens per second per GPU. - The expert sees a meaningful difference between hyperscalers and neoclouds on GPU investment economics. Hyperscalers like $MSFT, $GOOGL, and $AMZN can tolerate a 3-5 year payback period given their ability to monetize the same infrastructure across multiple revenue streams. Neoclouds like $CRWV operate on a tighter 2-3 year window. - With higher financing costs and direct dependence on infrastructure cash yield, neoclouds are far more sensitive to utilization and GPU residual risk. The biggest risk is GPU generation cycles, where a slow payback means newer chips could erode pricing power before the asset has paid itself off. - The expert explains that for hyperscalers, roughly 60% of GPU capacity is allocated to external monetization, including GPU rentals, managed AI services, enterprise inference workloads, and startup model training. The remaining 40% is used internally, and of that internal portion, the majority is still indirectly monetized through products like M365 Copilot or GitHub. Around 40% is dedicated to pure R&D. - The GPU pricing mix has shifted meaningfully over the past few years. In 2023, around 70-80% of revenue was hourly as customers paid a premium just to get access to scarce GPUs. By 2025 that had moved to roughly 50% hourly and 30-50% committed, and the expert expects 2026 to tip further toward committed at around 65% for hyperscalers as AI matures and inference becomes more predictable. Neoclouds are moving in the same direction but more slowly. - By 2027-2030, the expert sees committed contracts settling at 55-65% as the norm, with hourly pricing remaining but losing its scarcity premium as more supply comes online.

The story of AI in the next few years is going to be compute: an essay on the future of AI. K3 in 2 days is already #10# on OpenRouter with ~140B tok/day, and it’s infra is crumbling. Throughput is down from 30tok/s to 13tok/s, E2E latency is up to 72s and time to first token is >20s! It would cost a minimum of $500k to buy the 8 B300s it would take to serve even quantized Kimi K3 and ~$4M for the more recommended GB300 NVL72 rack. I don’t think Moonshot has the compute available to scale to their demand! In fact, even the US based inference providers will likely not be able to scale capacity as much as they’d like even if they were to host it: a 2.8T model is no joke. GPU providers (neoclouds etc) are doing 3yr and I recently hear 5yr commits with an ungodly 30% down, and customers are chomping it up. Prices continue to go to the moon. The two big labs, hyperscaler clouds, Grok and Meta have compute deals locked in prior, and the rest are fighting for scraps. Tier 1 neoclouds (coreweave/nebius etc) are rumored to not even small “smaller” customers. Meta is the biggest wildcard here. With ~7GW of compute by eoy 2026 and no clear big model ties, they either get to frontier on their own or can host the most Kimi K3 capacity (unless they sell it to the labs). Even though the price of models has fallen over time, it’s worth noting that the price of frontier has not. 3yrs ago, GPT-4 released at $60/M, o1 at $60/M, Opus 4 at $75/M, GPT5 at $10/M, Fable at $50/M and now Sol at $30/M and K3 at $15/M. Even if you consider K3 frontier, that’s only a 4-5x flux in 3yrs. In that time, frontier demand has increased at least 3+ ooms and frontier intelligence performance has gone 32x at least by task time by METR. Essentially, so long as a) the demand for frontier intelligence continues to grow to near infinity, b) the frontier continues to grow in performance, even as c) if the price of frontier declines a little, the value accrued to frontier grows significantly! And there’s a tremendous bull case for those who have locked up compute if you’re bitter lesson pilled and believe larger models will always be smarter models.