1 / 14

Insertion Policy Selection Using Decision Tree Analysis

Insertion Policy Selection Using Decision Tree Analysis. Samira Khan, Daniel A. Jiménez University of Texas at San Antonio. Motivation. L1 and L2 filters the cache access Last Level Cache (LLC) does not have much temporal locality

jafari
Download Presentation

Insertion Policy Selection Using Decision Tree Analysis

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Insertion Policy Selection Using Decision Tree Analysis Samira Khan, Daniel A. Jiménez University of Texas at San Antonio

  2. Motivation • L1 and L2 filters the cache access • Last Level Cache (LLC) does not have much temporal locality • Large fraction of blocks brought to cache are never accessed again (zero reuse lines). • For SPEC CPU 2006 benchmarks, on average 60.18% lines are never accessed again while they are in the LLC

  3. Motivation • No cache bursts in LLC • Only small portion of hits occur near the MRU position

  4. Goal • Get rid of zero reuse lines as early as possible • Keep lines in cache for sufficient time to get the first hit • Minimal change to LRU policy • Use as little space as possible

  5. Insertion Position Selection • Find the optimal insertion position • Zero reuse lines will get evicted earlier • Most of the non zero reuse lines should be in cache before their first hit • This will get rid of zero reuse lines and make space for useful lines • Use Decision Tree Analysis via set dueling to find the position • This allows choosing among the insertion positions to set duel

  6. nearMRU pos nearLRU pos For 400.perlbench 66.67% lines brought to cache are never accessed again and 73.03% hits occur in between MRU and middle position LRU pos middle pos MRU pos Set dueling between middle and MRU pos MRU pos winner middle pos winner Set dueling between LRU and middle pos Set dueling between nearMRU and MRU pos nearMRU pos winner MRU pos winner LRU pos winner Middle pos winner Insert pos LRU Set dueling between nearLRU and middle pos Insert pos nearMRU Insert pos MRU nearLRU pos winner Middle pos winner Insert pos nearLRU Insert pos middle

  7. Adaptive Multi Set Dueling • Current multi set dueling • Have one leader set for each insertion policy • Partial follower sets duplicate the winner set policy • Each policy set duel in a tournament manner • Not scalable • Leader sets performing the looser policies hurt performance • Adaptive multi set dueling • Leader set adaptively chooses the policy • No need for partial follower set • Scalable

  8. Result

  9. Space Overhead Space overhead for a 1MB 16 way set associative LLC

  10. Conclusion • Insertion Position Selection using Decision Tree Analysis • Requires minimal change to LRU • Needs only 2069 bits extra space • Chooses the best insertion position adaptively • Gets rid of zero reuse lines without any storage hungry predictor • Makes multi set dueling scalable

  11. Questions

  12. Zero Reuse Lines in SPEC CPU 2006

  13. Adaptive Multi Set Dueling -1 pselab pa +1 pb -1 φab -1 pselcd pc psel1 +1 pd Leader sets in current multi set dueling scheme +1 φcd -1 pselef pe +1 pf -1 φef -1 pselgh pg psel2 +1 ph +1 φgh -1, if pa wins -1 psel1 pa +1 Leader sets in adaptive multi set dueling scheme pb psel2 +1 pα All sets in LLC +1, if pb wins

  14. Result MRU nearMRU middle nearLRU LRU psel2 psel1 s

More Related