alt Hacker NewsFUSE generally has low overall performance because of an additional data transfer process between the kernel space and user space, which is less than ideal for AI training.
FUSE generally has low overall performance because of an additional data transfer process between the kernel space and user space, which is less than ideal for AI training.
The MLPerf performance of JuiceFS last year, FUSE client, and TCP/IP network. There're some analysis about performance and bottleneck. -> https://juicefs.com/en/blog/engineering/mlperf-storage-v2-ai...