Search for: All records

Current state-of-the-art resource management systems leverage Machine Learning (ML) methods to enable the efficient use of heterogeneous memory hardware, deployed across emerging computing platforms. While machine intelligence can be effectively used to learn and predict complex data access patterns of modern analytics, the use of ML over the exploded data sizes and memory footprints is prohibitive for its practical system-level integration. For this reason, recent solutions use existing lightweight historical information to predict the access behavior of majority of the application pages, and train ML models over a small page subset. To maximize application performance improvements, the pages selected for machine learning-based management are identified with elaborate page selection methods. These methods involve the calculation of detailed performance estimates depending on the configuration of the hybrid memory platform. This paper aims to reduce such vast operational overheads, that further exacerbate the existing high overheads of using machine intelligence, in return for high performance and efficiency. To this end, we build Cronus, an image-based pipeline for selecting pages for ML-based management. We visualize memory access patterns and reveal spatial and temporal correlations among the selected pages, that current methods fail to leverage. We then use the created images to detect patterns and select page groups for machine learning model deployment. Cronus drastically reduces the operational costs, while preserving the effectiveness of the page selection and achieved performance of machine intelligent hybrid memory management. This work makes a case that visualization and computer vision methods can unlock new insights and reduce the operational complexity of emerging systems solutions. 

Emerging workloads benefit from massive memory capacities provided by hybrid memory platforms. Recent system-level hybrid memory management solutions integrate machine learning methods to better predict complex data access behaviors. Given the substantial associated learning overheads, such solutions train parallel recurrent neural networks to learn the access patterns at the granularity of a page for a carefully selected page subset. Our observation reveals that the size of this subset varies immensely across workload classes, sizes and patterns. Increasing the granularity at the level of a page group will help reduce the aggregate learning overheads. Yet, unsupervised machine learning clustering methods are not practical to use in this context. Instead, this paper builds Coeus - a page grouping mechanism for machine learning-based hybrid memory management. Coeus is simple, robust and efficient. Coeus leverages data reuse insights to fine-tune the granularity at which patterns are interpreted by the system. As a result, Coeus creates large clusters of pages that share the same access behavior, in a practical way. Coeus reduces by almost 3x the associated learning overheads. In addition, Coeus achieves 3x higher application performance, by the combined effects of applying machine learning to more pages and by performing management operations at better granularity, compared to configurations of existing hybrid memory managers. 

Current state-of-the-art systems for hybrid memory management are enriched with machine intelligence. To enable the practical use of Machine Learning (ML), system-level page schedulers focus the ML model training over a small subset of the applications’ memory footprint. At the same time, they use existing lightweight historical information to predict the access behavior of majority of the pages. To maximize application performance improvements, the pages selected for machine learning-based management are identified with elaborate page selection methods. These methods involve the calculation of detailed performance estimates depending on the configuration of the hybrid memory platform. This paper explores the opportunities to reduce such operational overheads of machine learning-based hybrid memory page schedulers via use of visualization techniques to depict memory access patterns, and reveal spatial and temporal correlations among the selected pages, that current methods fail to leverage. We propose an initial version of a visualization pipeline for prioritizing pages for machine learning, that is independent of the hybrid memory configuration. Our approach selects pages whose ML-based management delivers, on average, performance levels within 5% of current solutions, while reducing by 75 × the page selection time. We discuss future directions and make a case that visualization and computer vision methods can unlock new insights and reduce the operational complexity of emerging systems solutions. 

Emerging hybrid memory systems that comprise technologies such as Intel's Optane DC Persistent Memory, exhibit disparities in the access speeds and capacity ratios of their heterogeneous memory components. This breaks many assumptions and heuristics designed for traditional DRAM-only platforms. High application performance is feasible via dynamic data movement across memory units, which maximizes the capacity use of DRAM while ensuring efficient use of the aggregate system resources. Newly proposed solutions use performance models and machine intelligence to optimize which and how much data to move dynamically. However, the decision of when to move this data is based on empirical selection of time intervals, or left to the applications. Our experimental evaluation shows that failure to properly conFigure the data movement frequency can lead to 10%-100% performance degradation for a given data movement policy; yet, there is no established methodology on how to properly conFigure this value for a given workload, platform and policy. We propose Cori, a system-level tuning solution that identifies and extracts the necessary application-level data reuse information, and guides the selection of data movement frequency to deliver gains in application performance and system resource efficiency. Experimental evaluation shows that Cori configures data movement frequencies that provide application performance within 3% of the optimal one, and that it can achieve this up to 5 x more quickly than random or brute-force approaches. System-level validation of Cori on a platform with DRAM and Intel's Optane DC PMEM confirms its practicality and tuning efficiency. 

Application performance improvements in emerging systems with hybrid memory components, such as DRAM and Intel’s Optane DC persistent memory, are possible via periodic data movements, that maximize the DRAM use and system resource efficiency. Similarly, predominantly used NUMA DRAM-only systems benefit from data balancing solutions, such as AutoNUMA, which periodically remap an application and its data on the same NUMA node. Although there has been a significant body of research focused on the clever selection of the data to be moved periodically, there is little insight as to how to select the frequency of the data movements, i.e., the duration of the monitoring period. Our experimental analysis shows that fine-tuning the period frequency can boost application performance on average by 70% for systems with locally attached memory units and 5x when accessing remote memory via interconnection networks. Thus, there is potential for significant performance improvements just by cleverly selecting the frequency of the data movements apart from choosing the data itself. While existing solutions empirically set the duration of the period, our work provides insights into the application-level properties that influence the choice of the period. More specifically, we show that there is a correlation between the application-level data reuse distance and migration frequency. Future work aims to solidify this correlation and build a profiling solution that provides users with the data movement frequency which dynamic data management solutions can then use to enhance performance. 

For hosting data-serving and caching workloads based on key-value stores in clouds, the cost of memory represents a significant portion of the hosting expenses. The emergence of cheaper, but slower, types of memories, such as NVDIMMs, opens opportunities to reduce the hosting costs for such workloads. The question explored in this paper is how to determine adequate allocations of different memory types in future systems with heterogeneous memory components, so as to retain desired performance SLOs and maximize the cost efficiency of the memory resource. We develop Mnemo, a memory sizing and data tiering consultant, that permits quick exploration of the cost-benefit tradeoffs associated with different configurations of the hybrid memory components used by key-value store workloads. Using experimental evaluation with different workload patterns, Mnemo is able to afford applications such as Redis, Memcached and DynamoDB, with substantial reduction in their hosting costs, at negligible impact on application performance, thus improving the overall system memory cost efficiency. 

We evaluated Intel ® Optane™ DC Persistent Memory and found that Intel's persistent memory is highly sensitive to data locality, size, and access patterns, which becomes clearer by optimizing both virtual memory page size and data layout for locality. Using the Polybench high-performance computing benchmark suite and controlling for mapped page size, we evaluate persistent memory (PMEM) performance relative to DRAM. In particular, the Linux PMEM support maps preferentially maps persistent memory in large pages while always mapping DRAM to small pages. We observed using large pages for PMEM and small pages for DRAM can create a 5x difference in performance, dwarfing other effects discussed in the literature. We found PMEM performance comparable to DRAM performance for the majority of tests when controlled for page size and optimized for data locality. 

The increasing demand of big data analytics for more main memory capacity in datacenters and exascale computing environments is driving the integration of heterogeneous memory technologies. The new technologies exhibit vastly greater differences in access latencies, bandwidth and capacity compared to the traditional NUMA systems. Leveraging this heterogeneity while also delivering application performance enhancements requires intelligent data placement. We present Kleio, a page scheduler with machine intelligence for applications that execute across hybrid memory components. Kleio is a hybrid page scheduler that combines existing, lightweight, history-based data tiering methods for hybrid memory, with novel intelligent placement decisions based on deep neural networks. We contribute new understanding toward the scope of benefits that can be achieved by using intelligent page scheduling in comparison to existing history-based approaches, and towards the choice of the deep learning algorithms and their parameters that are effective for this problem space. Kleio incorporates a new method for prioritizing pages that leads to highest performance boost, while limiting the resulting system resource overheads. Our performance evaluation indicates that Kleio reduces on average 80% of the performance gap between the existing solutions and an oracle with knowledge of future access pattern. Kleio provides hybrid memory systems with fast and effective neural network training and prediction accuracy levels, which bring significant application performance improvements with limited resource overheads, so as to lay the grounds for its practical integration in future systems.