Array Tracking Prefetcher for Indirect Accesses

Cavus, Mustafa; Sendag, Resit; Yi, Joshua J.

doi:10.1109/iccd.2018.00028

Cited by 2 publications

(10 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The four NT variants: prefetcher-only (ntpf), prefetcher with BOS (ntpf+bos), pre-execution unit (ntpx), and pre-execution unit with BOS (ntpx+bos) evaluated in this article achieve 4.1×, 11.9×, 14.9×, and 18.8× speedups, respectively, on average, over a no-prefetching baseline. In contrast, AMAC [1] and ATP [11] provided 2.4× and 1.2× speedups, respectively. Other recent prefetchers (BO [16], ISB [18], and SPP [17]) do not offer significant performance gains for the workloads in this article.…”

Section: Introductionmentioning

confidence: 90%

“…This allows multiple LDS traversal-misses from different lookups to overlap, increasing MLP. Similarly to recent prefetchers [3,11] and on-chip accelerators [4,12], NT reduces design cost and complexity through tight integration with the core, eliminating the need for dedicated TLB and cache units.…”

Section: Nt Architecture Overviewmentioning

confidence: 99%

“…KNAP is a programmable prefetcher that fetches the target key (i.e., the key to be searched) and computes the address of the node where the lookup should start. KNAP is similar to the recent sophisticated programmable prefetchers ATP [11] and ETP [5]. Both ATP and ETP can prefetch irregular patterns-they were initially designed to target indirect access patterns.…”

Section: Key and Node-address Providermentioning

confidence: 99%

“…KNAP's implementation was made similar to ATP because of ATP's simplicity and good performance. KNAP employs configurable registers and tables (see ATP [11]) that are configured by special instructions provided by compiler/programmer. It effectively creates a dependency graph between arrays using an array table (one entry for each array) and a relation table (keeps an entry for each relationship between two arrays).…”

Section: Key and Node-address Providermentioning

confidence: 99%

“…KNAP is programmed using three instructions (same as in ATP [11,64]): ATCL, ATAR, and ATRL (for more complex hashing behavior, a fourth instruction ATOP is utilized as described in References [11,64]). ATCL clears all KNAP tables.…”

Section: Key and Node-address Providermentioning

confidence: 99%

See 4 more Smart Citations

Fast Key-Value Lookups with Node Tracker

Cavus

Shatnawi

Sendag

et al. 2021

ACM Trans. Archit. Code Optim.

Self Cite

View full text Add to dashboard Cite

Lookup operations for in-memory databases are heavily memory bound, because they often rely on pointer-chasing linked data structure traversals. They also have many branches that are hard-to-predict due to random key lookups. In this study, we show that although cache misses are the primary bottleneck for these applications, without a method for eliminating the branch mispredictions only a small fraction of the performance benefit is achieved through prefetching alone. We propose the Node Tracker (NT), a novel programmable prefetcher/pre-execution unit that is highly effective in exploiting inter key-lookup parallelism to improve single-thread performance. We extend NT with branch outcome streaming (BOS) to reduce branch mispredictions and show that this achieves an extra 3× speedup. Finally, we evaluate the NT as a pre-execution unit and demonstrate that we can further improve the performance in both single- and multi-threaded execution modes. Our results show that, on average, NT improves single-thread performance by 4.1× when used as a prefetcher; 11.9× as a prefetcher with BOS; 14.9× as a pre-execution unit and 18.8× as a pre-execution unit with BOS. Finally, with 24 cores of the latter version, we achieve a speedup of 203× and 11× over the single-core and 24-core baselines, respectively.

show abstract

Section: Introductionmentioning

confidence: 90%

Section: Nt Architecture Overviewmentioning

confidence: 99%

Section: Key and Node-address Providermentioning

confidence: 99%

Section: Key and Node-address Providermentioning

confidence: 99%

Section: Key and Node-address Providermentioning

confidence: 99%

See 3 more Smart Citations

Fast Key-Value Lookups with Node Tracker

Cavus

Shatnawi

Sendag

et al. 2021

ACM Trans. Archit. Code Optim.

Self Cite

View full text Add to dashboard Cite

show abstract

A Cross-Prefetcher Schedule Optimization Methodology

2022

View full text Add to dashboard Cite

Prefetching offers the potential to significantly improve performance by speculatively loading application data so that it is available before it is needed. By their very nature, prefetching techniques are application behavior dependant. This implies that no universal prefetching solution exists. A combination of prefetching strategies need to be used to target a diverse set of applications. In this work, we develop the first comprehensive mathematical framework that allows a designer to better understand the prefetching opportunities of an application. We first use dynamic analysis to study the memory access behavior of an application and measure a series of metrics to both identify the optimized schedule, and estimate its achievable performance. To validate our model, we implement and evaluate three different prefetching strategies: helper threads, software prefetching and FPGA prefetching. We show that, for each individual scenario, our framework correctly generates the optimized schedule of prefetches and predicts the performance improvement with an accuracy of more than 95%. Using our framework, developers can choose the best prefetching strategy and parameters for their specific workload and use case.

show abstract

Array Tracking Prefetcher for Indirect Accesses

Cited by 2 publications

References 18 publications

Fast Key-Value Lookups with Node Tracker

Fast Key-Value Lookups with Node Tracker

A Cross-Prefetcher Schedule Optimization Methodology

Contact Info

Product

Resources

About