История изменений

Исправление HolyBoy, 02.02.13 15:45 (текущая версия) :

Хотя, сам спросил, сам ответил:

In attempting this project, the following information was learned:
- Memory copy overhead and latency.
CUDA recommends that memory buffers coming from the application be transfered to the GPU in a staged approach to maximize the parallel activity of the GPU. While this is a perfectly reasonable recommendation, it introduces additional latency in our real-time processing environment. It is important to remember that these latencies are a trade-off: if the algorithm has a marked improvement in execution time which offsets the introduced memory latency, then it is completely reasonable to go this route. This would be perfectly acceptable for video transcoding, because of the larger amount of data to process with possibility of more complex algorithms. This is not the case for audio transcoding.

Т.е. даже объём данных в аудиопотоке недостаточно велик для нормальной загрузки GPU отдельным куском этих данных и выигрыша нет. Ну что же, для транскодинга, значит, выгоднее использовать какие-нибудь блейды многопроцессорные и кучу VoIP-серверов.

Исходная версия HolyBoy, 02.02.13 15:43:

Хотя, сам спросил, сам ответил:

In attempting this project, the following information was learned:
- Memory copy overhead and latency.
CUDA recommends that memory buffers coming from the application be transfered to the GPU in a staged approach to maximize the parallel activity of the GPU. While this is a perfectly reasonable recommendation, it introduces additional latency in our real-time processing environment. It is important to remember that these latencies are a trade-off: if the algorithm has a marked improvement in execution time which offsets the introduced memory latency, then it is completely reasonable to go this route. This would be perfectly acceptable for video transcoding, because of the larger amount of data to process with possibility of more complex algorithms. This is not the case for audio transcoding.