Share your vclHashcat-lite speeds

[epixoip]

sort of.

supercl reduces the amount of kernel executions that occur over the network. instead of making one call for each kernel execution, it allows you to make just one call for *all* your kernel executions. so in that regard, it eliminates the latency issues with gige.

however, it doesn't really reduce the amount of bandwidth that is required -- it just shifts where it's required from between the broker and compute nodes, to between the compute nodes and your network storage.