|
无与伦比的灵活性和强大功能
Pallas AI Flex 系列为 AI 计算提供无与伦比的灵活性和性能，采用多节点、英伟达支持的配置，可支持广泛的 AI 工作负载。它提供卓越的可扩展性和可定制性，并通过浦卓的集中式管理系统实现最佳服务器利用率。该系列适用于气隙环境，以确保数据完整性和安全性，使其成为需要强大且安全计算解决方案的政府、军事和大型企业应用程序的理想选择。
|
Advanced Processing Power |
Powered by the latest Intel Xeon processors, delivering exceptional performance for demanding applications |
|
Memory capacity |
Supports up to 40TB of DDR5 memory, accommodating memory-intensive AI tasks |
|
AI-Ready configuration |
Configurable with up to 120 Nvidia A100 GPUs, optimized for high-performance AI workloads |
|
NVIDIA AI Framework compatibility |
Fully compatible with Nvidia AI frameworks, providing a comprehensive platform for AI development and deployment |
|
NVIDIA approved hardware |
Certified for AI applications and supported by Nvidia, ensuring reliable and efficient AI operations |
| Form factor | 19” rack equipment |
| Processor |
Intel Xeon 5th generation scalable |
| Memory |
Up to 40TB DDR5 |
| GPU |
Up to 120 Nvidia H100 GPUs |
| Storage | Up to 5.3PB per cluster of Direct Attached Storage or Storage Array Network depending on the requirements of the application |
| Networking |
Infiniband HDR 200Gb/s fabric |
Ideal for extensive AI training and inference workloads
High scalability and reliability for critical enterprise applications
Enhanced performance and efficiency for decentralized data processing and AI applications
Robust performance for demanding computational tasks