¶ Overall Architecture of the File Storage Systems on MONTAGE
MONTAGE runs on a modern parallel storage system for end-to-end collaborative common enterprise data platforms, analytics, and AI workflows - IBM Storage Scale (a.k.a. Spectrum Scale or GPFS).
¶ GPUDirect Storage (GDS)
IBM Spectrum Scale's support for NVIDIA's GPUDirect Storage (GDS) enables a direct path between GPU memory and storage. This solution addresses the need for higher throughput and lower latencies. File system storage is directly connected to the GPU buffers to reduce latency and load on CPU. For IBM Spectrum Scale, this means that data can be read directly from an NSD server's pagepool and it is sent to the GPU buffer of the IBM Spectrum Scale clients by using RDMA. IBM Spectrum Scale with GDS requires an InfiniBand fabric.
GDS is useful where significant I/O is involved and the CPU stands as a bottleneck to overall system performance. This happens when the CPU cycles are heavily used for managing data transfers into and out of CPU memory and CPU DRAM bandwidth is used.