Last-Touch Correlated Data Streaming

Abstract

Abstract

Temporal streams in commercial server applications

“…Furthermore, prior studies have considered only a single system organization, focusing either on uniprocessors [7,8,9,21] or multi-chip distributed-shared-memory systems [25]. As we will show, server applications' off-chip miss behavior changes drastically when all cores are located on a single chip, because coherence activity and contention are captured entirely within that chip.…”

“…Temporal streams extend the notion of address correlation to sequences rather than pairs of misses. Although the term "temporal stream" was introduced in [25], a wide variety of recent prefetchers rely on the same underlying phenomenon, including hot data stream prefetching [7], the global history buffer [19], the user-level memory thread [21], epoch-based correlation prefetching [8], and last-touch correlated data streaming [9].…”

“…Long streams amortize prefetch costs (e.g., off-chip lookup latency [8,9,25]). However, long streams may also imply the need to throttle stream retrieval-a long stream cannot be buffered entirely on chip without displacing other potentially-useful data.…”

“…To improve performance for these access patterns, over a decade of research has lead to the development of address-correlating prefetchers, which exploit correlation between consecutive memory accesses and are highly-effective for pointerbased structures [6,7,8,9,13,14,15,19,21,25]. Building on this line of research, the latest proposals prefetch extended sequences of memory accesses that recur over the course of program execution [7,9,19,21,25].…”

“…Building on this line of research, the latest proposals prefetch extended sequences of memory accesses that recur over the course of program execution [7,9,19,21,25]. We adopt the terminology of [25] and refer to these recurring memory access sequences as temporal streams.…”