12.07.2015 Views

NoC design and optimization for Multi-core media processors

NoC design and optimization for Multi-core media processors

NoC design and optimization for Multi-core media processors

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

CHAPTER 2. RELATED WORK 25Figure 2.1: Floorplan used in estimating wire lengths. Wire lengths estimated from thesefloorplans are used as input to Intacte to arrive at a power optimal configuration <strong>and</strong>latencyinclockcycles. HorizontalR-R:Linkbetweenneighboringroutersinthehorizontaldirection, Vertical R-R: Link between neighbouring routers in the vertical direction.Cache Based Microarchitectural TechniquesStridedprefetching[106]hasbeencomparedwithblockmigration<strong>and</strong>onchiptransmissionlines to manage on-chip wire delay in CMP caches to improve per<strong>for</strong>mance. Work in [106]combines block migration strategies, transmission line caches <strong>and</strong> stride-based hardwareto reduce cache based communication latencies in CMPs. Dynamic Non-Uni<strong>for</strong>m CacheArchitecture (DNUCA)[107] uses block migration techniques to reduce cache latencies.Block migration involves moving frequently accessed blocks in the cache into banks withlower latencies. On chip transmission lines are used to provide low latency to all banks inTransmission Line Caches[108]. Around 40%-60% of L2 cache hits are satisfied in centralbanks in CMPs with shared L2 rendering block migration ineffective. TLCs suffer fromb<strong>and</strong>width contention reducing its advantage.Communication energy <strong>and</strong> delay can be minimized by migrating frequently accessedcachelinestobanksclosesttotheaccessing<strong>processors</strong>[109][110][111]. Non-Uni<strong>for</strong>mCacheArchitectures (NUCAs) with policies allow important cache lines to migrate toward theprocessor within the same level of the cache have been proposed in [109][110]. The cachearchitecture proposed in [109][110] shows that a dynamic NUCA structure achieves higherIPC than a traditional Uni<strong>for</strong>m Cache Architecture (UCA) when maintaining the samesize <strong>and</strong> manufacturing technology. D-NUCA mapping policy attempts to provide fastestbankaccessto all banksetsby sharingtheclosest banksamong multiplebanksets. In this

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!