摘要
以X264编码器作为研究对象,在指令集并行的基础上对其进行线程级并行优化,在Intel双核处理器平台上,针对非实时编码应用得到近2倍的加速比,针对实时编码应用得到1.5倍的加速比。
Taking X264 coder as researching object, the thread-level parallel optimization is realized based on instruetion level parallelism. On Intel dual-core platform, the speedup is almost 2 in non-real-time coding applications, and 1.5 in real-time coding applications.
出处
《电视技术》
北大核心
2008年第2期33-35,共3页
Video Engineering
关键词
H.264
指令级并行
线程级并行
多核
H.264
instruction level parallelism
thread level parallelism
muhi-core
