Exploiting Program Semantics to Place Data in Hybrid Memory

Wei, Wei; Jiang, Dongxiang; McKee, Sally A.; Xiong, Jin; Chen, Mingyu

doi:10.1109/pact.2015.10

Cited by 51 publications

(48 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Different hybrid/heterogeneous main memory structures have been proposed, such as DRAM + NVM [12,13,6,14,3,15,4], on-chip memory + off-chip memory [16], and DRAMs with different feature values [17,8]. Though with different technologies, all of them share the common characteristic that different parts of the address space yield different access cost.…”

Section: Related Workmentioning

confidence: 99%

Efficient Management for Hybrid Memory in Managed Language Runtime

Wang

Cao

Zigman

et al. 2016

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Hybrid memory, which leverages the benefits of traditional DRAM and emerging memory technologies, is a promising alternative for future main memory design. However popular management policies through memory-access recording and page migration may invoke non-trivial overhead in execution time and hardware space. Nowadays, managed language applications are increasingly dominant in every kind of platform. Managed runtimes provide services for automatic memory management. So it is important to adapt them for the underlying hybrid memory. This paper explores two opportunities, heap partition placement and object promotion, inside managed runtimes for allocating hot data in a fast memory space (fast-space) without any access recording or data migration overhead. For heap partition placement, we quantitatively analyze LLC miss density and performance effect for each partition. Results show that LLC misses especially store misses mostly hit nursery partitions. Placing nursery in fast-space, which is 20% total memory footprint of tested benchmarks on average, causes only 10% performance difference from all memory footprint in fast-space. During object promotion, hot objects will be directly allocated to fast-space. We develop a tool to analyze the LLC miss density for each method of workloads, since we have found that the LLC misses are mostly triggered by a small percentage of the total set of methods. The objects visited by the top-ranked methods are recognized as hot. Results show that hot objects do have higher access density, more than 3 times of random distribution for SPECjbb and pmd, and placing them in fast-space further reduces their execution time by 6% and 13% respectively.

show abstract

Section: Related Workmentioning

confidence: 99%

Efficient Management for Hybrid Memory in Managed Language Runtime

Wang

Cao

Zigman

et al. 2016

Lecture Notes in Computer Science

View full text Add to dashboard Cite

show abstract

“…For individual KPA allocations, StreamBox-HBM further considers the critical path. StreamBox-HBM does not migrate existing KPAs, which are ephemeral, unlike other software systems for hybrid memory [61,63,72]. Figure 6 plots StreamBox-HBM's state space.…”

Section: Dynamically Managing Hybrid Memorymentioning

confidence: 99%

StreamBox-HBM

Miao¹,

Jeon²,

Pekhimenko³

et al. 2019

Proceedings of the Twenty-Fourth International Conference on Architectural Support for Programming Languages and Operating Syst

View full text Add to dashboard Cite

Stream analytics has an insatiable demand for memory and performance. Emerging hybrid memories combine commodity DDR4 DRAM with 3D-stacked High Bandwidth Memory (HBM) DRAM to meet such demands. However, achieving this promise is challenging because (1) HBM is capacitylimited and (2) HBM boosts performance best for sequential access and high parallelism workloads. At first glance, stream analytics appears a particularly poor match for HBM because they have high capacity demands and data grouping operations, their most demanding computations, use random access.This paper presents the design and implementation of StreamBox-HBM, a stream analytics engine that exploits hybrid memories to achieve scalable high performance.StreamBox-HBM performs data grouping with sequential access sorting algorithms in HBM, in contrast to random access hashing algorithms commonly used in DRAM. StreamBox-HBM solely uses HBM to store Key Pointer Array (KPA) data structures that contain only partial records (keys and pointers to full records) for grouping operations. It dynamically creates and manages prodigious data and pipeline parallelism, choosing when to allocate KPAs in HBM. It dynamically optimizes for both the high bandwidth and limited capacity of HBM, and the limited bandwidth and high capacity of standard DRAM.StreamBox-HBM achieves 110 million records per second and 238 GB/s memory bandwidth while effectively utilizing all 64 cores of Intel's Knights Landing, a commercial server with hybrid memory. It outperforms stream engines with sequential access algorithms without KPAs by 7× and stream engines with random access algorithms by an order Permission to make digital of magnitude in throughput. To the best of our knowledge, StreamBox-HBM is the first stream engine optimized for hybrid memories.

show abstract

“…Wei et al [56] explore read and write predictability for allocation sites in native C applications. C semantics limit the applicability of their approach, and their applications allocate a majority of the heap from a small number of (up to 4) sites.…”

Section: Related Workmentioning

confidence: 99%

Crystal Gazer

Akram

Sartor

McKinley

et al. 2019

Proc. ACM Meas. Anal. Comput. Syst.

View full text Add to dashboard Cite

Non-volatile memories (NVM) offer greater capacity than DRAM but suffer from high latency and low write endurance. Hybrid memories combine DRAM and NVM to form scalable memory systems with the promise of high capacity, low energy consumption, and high endurance. Automatically managing hybrid NVM-DRAM memories to achieve their promise without changing user applications or their programming models remains an open question. This paper uses garbage collection in managed languages to exploit NVM capacity while preventing NVM wear out in hybrid memories with no changes to the programming model.We introduce profile-driven write-rationing garbage collection. Allocation sites that produce frequently written objects are predicted based on previous program executions. Objects are initially allocated in a DRAM nursery space. The collector copies surviving nursery objects from highly written sites to a mature DRAM space and read-mostly objects to a mature NVM space. Write-intensity prediction for 15 Java benchmarks accurately places objects in the correct space, eliminating expensive object monitoring from prior write-rationing garbage collectors. Furthermore, our technique exposes a Pareto tradeoff between DRAM usage and NVM lifetime, unlike prior work. Experimental results on NUMA hardware that emulates hybrid NVM-DRAM memory demonstrates that profile-driven write-rationing garbage collection reduces the number of writes to NVM compared to prior work to extend its lifetime, maximizes the use of NVM for its capacity, and achieves good performance. S. Akram et al.worrisome as emerging applications have an insatiable desire for memory. Expecting further DRAM supply shortages, Facebook is already experimenting with hybrid DRAM and non-volatile memory (NVM) systems [23].Production NVM uses phase-change memory (PCM). PCM offers high capacity, low idle power, and non-volatility, while being byte addressable, unlike some NVM. It however suffers two major drawbacks over DRAM: long latency and limited write endurance [16,37,38]. New materials may eventually bridge the latency gap [30,42]. However, endurance will remain a problem because PCM writes change the material form [16]. Although prior hardware and software approaches improve PCM lifetime, an endurance gap remains [3,37,38,48].Memory system analysts and others report that Intel's Optane SSD, a commercially available NVM with 375 GB of capacity, can sustain a write rate of up to 10 TB/day, i.e., 140 MB/s without wearing out within the 3 year warranty period [23,26]. Our experimental results (using an optimistic PCM latency) show that 13 out of our 15 modern Java applications exceed this write rate when using a PCM-only main memory. In practice, memory systems will therefore need hybrid DRAM and PCM memories. Similar products with a smaller capacity (up to 32 GB) only support write rates of up to 1.5 MB/s [26]. All of our Java applications exceed this write rate even using state-of-the-art hybrid memory management.Including PCM into the main memory system requires solutions to t...

show abstract

Exploiting Program Semantics to Place Data in Hybrid Memory

Cited by 51 publications

References 23 publications

Efficient Management for Hybrid Memory in Managed Language Runtime

Efficient Management for Hybrid Memory in Managed Language Runtime

StreamBox-HBM

Crystal Gazer

Contact Info

Product

Resources

About