Abstract:Many high end and next generation computing systems to incorporated alternative memory technologies to meet performance goals. Since these technologies present distinct advantages and tradeoffs compared to conventional DDR* SDRAM, such as higher bandwidth with lower capacity or vice versa, they are typically packaged alongside conventional SDRAM in a heterogeneous memory architecture. To utilize the different types of memory efficiently, new data management strategies are needed to match application usage to the best available memory technology. However, current proposals for managing heterogeneous memories are limited because they either: 1) do not consider high-level application behavior when assigning data to different types of memory, or 2) require separate program execution (with a representative input) to collect information about how the application uses memory resources. This work presents a toolset for addressing the limitations of existing approaches for managing complex memories. It extends the application runtime layer with automated monitoring and management routines that assign application data to the best tier of memory based on previous usage, without any need for source code modification or a separate profiling run. It evaluates this approach on a state-of-the-art server platform with both conventional DDR4 SDRAM and non-volatile Intel Optane DC memory, using both memory-intensive high performance computing (HPC) applications as well as standard benchmarks. Overall, the results show that this approach improves program performance significantly compared to a standard unguided approach across a variety of workloads and system configurations. Additionally, we show that this approach achieves similar performance as a comparable offline profiling-based approach after a short startup period, without requiring separate program execution or offline analysis steps.

Who Ate My Memory? Towards Attribution in Memory Management

Asking and Answering Questions During Memory Profiling

Graph Neural Networks Based Memory Inefficiency Detection Using Selective Sampling

A Lightweight Hybrid Hardware/software Approach for Object-Relative Memory Profiling

What Do You Mean by Memory? When Engineers Are Lost in the Maze of Complexity

Understanding object-level memory access patterns across the spectrum

Software Visualizations to Analyze Memory Consumption: A Literature Review

Quantifying Daily Evolution of Mobile Software Based on Memory Allocator Churn

Losing My Memory

Examem: Low-Overhead Memory Instrumentation for Intelligent Memory Systems

Pinpointing the Memory Behaviors of DNN Training

Memory visualization tool for training neural network

A Memory Hierarchical Layer Assigning and Prefetching Technique to Overcome the Memory Performance/Energy Bottleneck

Lossless instruction-to-object memory tracing in the Linux kernel

Spindle: Informed Memory Access Monitoring

BlueJay: A Platform to Quantifying the Impact of Memory Latency on Datacenter Application Performance

DJXPerf: Identifying Memory Inefficiencies via Object-centric Profiling for Java

Understanding Power Consumption Metric on Heterogeneous Memory Systems

A Tool for Estimating Memory Usage

Online Application Guidance for Heterogeneous Memory Systems

Best Memory Architecture Exploration under Parameters Variations accelerated with Machine Learning