算力紧缩下 Fal、Together、Baseten 转向自购 Nvidia 芯片
Fal、Together、Baseten 这些靠租 GPU 做推理服务的公司，现在开始自己买 Nvidia 芯片了。算力越来越难租到，说明推理这条赛道的成本和供应格局在变。做部署的朋友可以看看别人怎么重新算这笔账。
据 The Information 报道，AI 推理初创公司正因算力获取收紧而被迫调整商业模式。Fal、Together 和 Baseten 开始自购 Nvidia 芯片，不再完全依赖云服务商的 GPU 供给。
这三家公司的核心业务是为开发者提供模型推理服务，GPU 供应的稳定性直接决定其服务质量与成本结构。此前它们主要通过租用云端算力来承接流量，如今算力紧张与成本压力促使其转向自建硬件。
这一动向反映了推理层基础设施玩家的普遍处境：当云厂商优先保障自家模型业务时，第三方推理平台需要通过持有自有芯片来掌控产能。对下游依赖这些平台部署模型的企业来说，供应链变化可能影响推理服务的价格与可用性。
AI inference startups are being forced to rethink their business models as access to computing power tightens. Fal, Together and Baseten are moving to buy their own Nvidia chips rather than rely entirely on cloud providers.
Read more: https://t.co/TncGpO541i
This report is categorized under AITOP's "Industry" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows. Key entities and technical topics under continuous tracking include Fal, Together AI, Baseten, Nvidia.
Source attribution: curated from The Information (@theinformation).
