Which cache strategies/algorithms (especially for L2 Cache) are used in practice and don't exist solely in research/theory? There is a list on Wikipedia which does not state which algorithms are really used. Bélády's Algorithm for example would be as far as i understand it the best solution but is impossible to be used because of the unpredictability of cache behaviour.

Is there a difference when a virtualized environment is used or do hypervisors use the same strategies?

  • $\begingroup$ 1) "currently used" as in by Intel and AMD? --> offtopic. 2) Please restrict yourself to one question per post. $\endgroup$
    – Raphael
    Jun 3 '14 at 11:04
  • 2
    $\begingroup$ You might be interested in this Stack Overflow question about replacement policies that are actually used. In general, system software (OS/hypervisor) does not control replacement policy (minor exceptions include cache locking and PPC7500's choice of random algorithm), though write policy, cacheability, and other caching factors can often be controlled at page granularity. $\endgroup$ Jun 3 '14 at 11:25
  • $\begingroup$ I don't get the reason why this should be off topic. I have the following problem: several processes are writing a random amount of cache lines. I am researching which cache lines will be written. Therefore i need to know the replacement strategies or cache algorithms. Regarding the issue with "currently used" i will rephrase my question. $\endgroup$
    – asgaroth
    Jun 4 '14 at 9:19
  • 2
    $\begingroup$ Your problem is off-topic, since this site deals with conceptual question, not details of specific implementations (be it software or hardware). $\endgroup$
    – FrankW
    Jun 4 '14 at 9:54
  • 1
    $\begingroup$ I would suggest that you give more explicit references (such as the wikipedia list: what URL). Also caches appear in many places: chip level, operating system, web communication. Maybe being more explicit about your specific interest might help. But I do not make decisions. $\endgroup$
    – babou
    Jun 10 '14 at 12:28

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.