引用:
原帖由 qcmadness 於 2012-6-2 22:15 發表
下代又改架構
我估都係 native 128-bit support, improved branch predictor & prefetcher, clock improvements, deeper instruction windows, FMA3, AVX & XOP support 再加 revamped cache hierarchy (跟進 BD L1-L2 inclusiveness...) with L2 ECC protected (for embedded systems) 掛。
唔知 L2 仲會唔會 half-clocked 呢。
[
本帖最後由 Puff 於 2012-6-3 01:30 編輯 ]