フリー問題

問題文

A team reports that adding a second device did not speed up their pipeline at all. Which explanation fits if the code was written with the single-device library?

選択肢

1. The second device was assigned a lower ordinal than the first, so the library used it as the primary device and left the original one idle, which is why the total throughput stayed the same instead of doubling.
2. The single-device library uses one device; using several requires the distributed framework with an accelerator-aware cluster.
3. The pipeline is bound by the host's memory bandwidth, so a second device only adds cores that sit idle behind the same narrow channel.
4. The second device needs to be made visible with an environment variable, so the library then spreads the work across both of them.