Arsitektur ARM vs PowerPC

Oleh: Ihsan Hariadi

Dari sebuah diskusi di mailing list:
Pertanyaan: “btw ARM kan advance RISC Machine tapi kenapa beda sama power PC? ada yang tahu?”

ihsan:
———
Konon perbedaan yang penting dari ARM vs PowerPC ada di arsitektur bagian Cache nya, selebihnya saya hanya bisa salin-dan-tempel di bawah ini

——( ihsan )———————————-

( ditulis oleh orang dari PYTEC GmbH .de yg pernah   kontak dgn salah seorang rekan di Indonesia  )

http://www.linuxfordevices.com/files/article081/Bauer.pdf

An important difference between the two processors is the cache architecture.

The PXA270 is based on an ARM core that uses a virtual cache architecture. In this architecture, the cache is located between the CPU and the MMU. This design has the advantage that on a cache hit the address does not need to be converted to a physical
address. But the disadvantage of this architecture is that when a thread switch occurs the cache needs to be flushed which imposes an additionally latency.

In contrast, the MPC5200 is equipped with an e300 core, where the cache is placed between the MMU and the memory (physical cache). In this design, the address translation has to take place before the cache access. This leads to a slower access time
on the cache because of the address translation but the cache contents is consistent in different threads, in such a way that the cache does not need to be flushed.

In addition, the PXA270 is equipped with 32k data and instruction cache while the MPC5200 only has 16k caches. In general, a bigger cache is an advantage because it increases the probability of cache hits. But when it comes to task switching, a bigger cache is a drawback because flushing the cache takes more time. Disclaimer : Isi tanggung jawab pembaca.

This entry was posted in Uncategorized. Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *