Simple DRAM and Virtual Memory Abstractions to Enable Highly Efficient Memory Systems (1605.06483v1)

Abstract: In most modern systems, the memory subsystem is managed and accessed at multiple different granularities at various resources. We observe that such multi-granularity management results in significant inefficiency in the memory subsystem. Specifically, we observe that 1) page-granularity virtual memory unnecessarily triggers large memory operations, and 2) existing cache-line granularity memory interface is inefficient for performing bulk data operations and operations that exhibit poor spatial locality. To address these problems, we present a series of techniques in this thesis. First, we propose page overlays, a framework augments the existing virtual memory framework with the ability to track a new version of a subset of cache lines within each virtual page. We show that this extension is powerful by demonstrating its benefits on a number of applications. Second, we show that DRAM can be used to perform more complex operations than just store data. We propose RowClone, a mechanism to perform bulk data copy and initialization completely inside DRAM, and Buddy RAM, a mechanism to perform bulk bitwise operations using DRAM. Both these techniques achieve an order-of-magnitude improvement in the efficiency of the respective operations. Third, we propose Gather-Scatter DRAM, a technique that exploits DRAM organization to effectively gather/scatter values with a power-of-2 strided access patterns. For these access patterns, GS-DRAM achieves near-ideal bandwidth and cache utilization, without increasing the latency of fetching data from memory. Finally, we propose the Dirty-Block Index, a new way of tracking dirty blocks. In addition to improving the efficiency of bulk data coherence, DBI has several applications including high-performance memory scheduling, efficient cache lookup bypassing, and enabling heterogeneous ECC.