CoreWeave 测验

当您参加测验时，就会出现答案和解释。

1. CoreWeave 主要提供什么？
   • 通过云访问大量用于 AI 的 Nvidia GPU
   • 消费类智能手机
   • 流媒体电影
   • 电动滑板车
2. 在专注于 GPU 云基础设施之前，CoreWeave 从事什么业务？
   • Food retail demand forecasting
   • 加密货币挖矿
   • Speech transcription services
   • A consumer social network
3. CoreWeave 将其 AI 云平台中的哪些服务列为组成部分？
   • Only a hosted language model with no infrastructure
   • Food inspection cameras and recipe tools
   • GPU compute together with networking, storage, and supporting software
   • A smartphone app for mining cryptocurrency
4. 为什么客户应在专用 GPU 云上对其自身的工作负载进行基准测试？
   • A specialized cloud guarantees a speedup for every model
   • The GPU name alone determines total cost
   • Cloud providers use identical networking and storage
   • Performance and cost depend on the workload, system configuration, and service requirements
5. CoreWeave 的文件指出了哪些针对客户和供应商的业务风险？
   • Concentration risks that can make the business sensitive to a small number of counterparties
   • Guaranteed demand from every customer
   • No dependency on hardware supply
   • A risk that only affects food safety inspections
6. CoreWeave 在 2025 年实现了哪些重要的企业里程碑？
   • 它停止使用 GPU
   • 它在股票市场上市
   • 它成为一个非营利组织
   • 它与一家咖啡连锁店合并
7. What is one infrastructure dependency a GPU cloud provider must manage for large AI clusters?
   • Only the model’s tokenizer vocabulary
   • The hotel’s room occupancy
   • Power, networking, cooling, and accelerator availability at data-center scale
   • The number of emotion labels in a dataset
8. Why should an enterprise compare end-to-end workload cost across GPU cloud providers?
   • The same nominal GPU always produces identical service cost
   • Model parameter count alone determines the bill
   • A public-company listing guarantees low prices
   • Hardware, interconnect, storage, utilization, and software affect total performance and cost