Hybrid main memory systems combine both performance and capacity advantages from heterogeneous memory technologies. With larger capacities, higher associativities, and finer granularities, hybrid memory systems currently exhibit significant metadata storage and lookup overheads for flexibly remapping data blocks between the two memory tiers. To alleviate the inefficiencies of existing designs, we propose Trimma, the combination of a multi-level metadata structure and an efficient metadata cache design. Trimma uses a multi-level metadata table to only track truly necessary address remap entries. The saved memory space is effectively utilized as extra DRAM cache capacity to improve performance. Trimma also uses separate formats to store the entries with non-identity and identity mappings. This improves the overall remap cache hit rate, further boosting the performance. Trimma is transparent to software and compatible with various types of hybrid memory systems. When evaluated on a representative DDR4 + NVM hybrid memory system, Trimma achieves up to 2.4$\times$ and on average 58.1\% speedup benefits, compared with a state-of-the-art design that only leverages the unallocated fast memory space for caching. Trimma addresses metadata management overheads and targets future scalable large-scale hybrid memory architectures.

翻译：混合主存系统通过异构内存技术融合了性能与容量双重优势。随着容量增加、关联度提升及粒度细化，当前混合内存系统在灵活重映射两个内存层级间的数据块时，面临显著的元数据存储与查找开销。为解决现有设计的低效问题，我们提出Trimma，该方案结合了多层级元数据结构与高效元数据缓存设计。Trimma采用多级元数据表仅追踪必要地址重映射条目，将节省的内存空间有效用作额外DRAM缓存容量以提升性能。同时，Trimma使用独立格式分别存储恒等映射与非恒等映射条目，从而提升整体重映射缓存命中率，进一步优化性能。Trimma对软件透明且兼容多种类型混合内存系统。在代表性DDR4+NVM混合内存系统上的评估显示，与仅利用未分配快速内存空间进行缓存的现有最优设计相比，Trimma可实现最高2.4倍、平均58.1%的加速比增益。Trimma致力于解决元数据管理开销问题，面向未来可扩展的大规模混合内存架构。