好,有意思,既然24楼拿出理论认为针对Read-After-Write (RAW)的“优化”对于桌面用户是“有通用性”的,那么我也专门看了一下这个论文,其中有这么一段:
We believe that the time interval from a write to a subsequent read is large due to large OS-level buffer caches and a smaller percentage of most overwritten blocks; as a result, the buffer cache can service most reads that oc- cur soon after a write, exposing only later reads that are issued after the block evictto the blockdevice .
论文中解释了为何Read-After-Write的发生几率如此之低,根据24楼举出的这个论文的数据,写入的数据在随后30秒内再度被读取的几率平均只有10%,这是因为Windows系统本身有一套缓存机制,会把刚存取过后的数据保留在内存当中备用,这就意味着多数RAW会被Windows缓存应付掉,硬盘对此进行优化显然是徒劳无功的。
不过24楼请不要忘记,840Evo释放缓存的延后时间并非30秒,根据三星自己的PDF,当840Evo的SLC区域填充满的情况下,可以在30秒内释放清空所有SLC区的数据。说白了就是即便我把SLC区写满,要清空他也只需要30秒。日常应用的写入量来说,要清空SLC区只怕用不了几秒钟吧?SLC区是随写随清空,再加上之前说到的Windows缓存机制,24楼你还坚持认为840Evo这种作弊能再实用中取得效果?
|