Efficient Load Value Prediction Using Multiple Predictors and Filters

Sheikh, Rami; Hower, Derek R.

doi:10.1109/hpca.2019.00057

Cited by 11 publications

(17 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We model criticality-unaware approaches by removing register-data-dependence edges sourcing from loads (0-100% randomly, in 10% increments) and re-calculating critical path to get CPI estimate (Figure 17). This represents predictors that have no regard for criticality-e.g., a criticality-unaware composite predictor [42], which achieves 40% coverage and 6.6% speed-up, and lands straight on our stochastic line in Figure 17.…”

Section: Criticality Insights Applied To Value Predictionmentioning

confidence: 60%

Calipers: A Criticality-aware Framework for Modeling Processor Performance

Golestani¹,

Sen²,

Young³

et al. 2022

Preprint

View full text Add to dashboard Cite

Computer architecture design space is vast and complex. Tools are needed to explore new ideas and gain insights quickly, with low efforts and at a desired accuracy. We propose Calipers, a criticality-based framework to model key abstractions of complex architectures and a program's execution using dynamic event-dependence graphs. By applying graph algorithms, Calipers can track instruction and event dependencies, compute critical paths, and analyze architecture bottlenecks. By manipulating the graph, Calipers enables architects to investigate a wide range of Instruction Set Architecture (ISA) and microarchitecture design choices/"what-if" scenarios during both early-and late-stage design space exploration without recompiling and rerunning the program. Calipers can model in-order and out-of-order microarchitectures, structural hazards, and different types of ISAs, and can evaluate multiple ideas in a single run. Modeling algorithms are described in detail.We apply Calipers to explore and gain insights in complex microarchitectural and ISA ideas for RISC and EDGE processors, at lower effort than cycle-accurate simulators and with comparable accuracy. For example, among a variety of investigations presented in the paper, experiments show that targeting only a fraction of critical loads can help realize most benefits of value prediction.

show abstract

Section: Criticality Insights Applied To Value Predictionmentioning

confidence: 60%

Calipers: A Criticality-aware Framework for Modeling Processor Performance

Golestani¹,

Sen²,

Young³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…In all the works above, values are predicted for all register-producing instructions (i.e., regardless of the type of their operation). There are also other studies that focus only on load-value prediction [1,13,14,30,31,34,35]. In particular, in Ref.…”

Section: Related Workmentioning

confidence: 99%

“…On the other hand, in Refs [14], [30], and [31], load-value prediction is implemented indirectly by using a different approach; rather than traditionally predicting the values of loads, these methods are based on predicting the memory addresses from where load-values are fetched. These predicted addresses are eventually used to retrieve the load-values earlier in a speculative manner.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Leveraging Value Equality Prediction for Value Speculation

Kalaitzidis

Seznec

2020

ACM Trans. Archit. Code Optim.

View full text Add to dashboard Cite

Value Prediction (VP) has recently been gaining interest in the research community, since prior work has established practical solutions for its implementation that provide meaningful performance gains. A constant challenge of contemporary context-based value predictors is to sufficiently capture value redundancy and exploit the predictable execution paths. To do so, modern context-based VP techniques tightly associate recurring values with instructions and contexts by building confidence upon them after a plethora of repetitions. However, when execution monotony exists in the form of intervals, the potential prediction coverage is limited, since prediction confidence is reset at the beginning of each new interval. In this study, we address this challenge by introducing the notion of Equality Prediction (EP), which represents the binary facet of VP. Following a twofold decision scheme (similar to branch prediction), at fetch time, EP makes use of control-flow history to predict equality between the last committed result for this instruction and the result of the currently fetched occurrence. When equality is predicted with high confidence, the last committed value is used. Our simulation results show that this technique obtains the same level of performance as previously proposed state-of-the-art context-based value predictors. However, by virtue of exploiting equality patterns that are not captured by previous VP schemes, our design can improve the speedup of standard VP by 19% on average, when combined with contemporary prediction models.

show abstract

“…al considered only load instructions by introducing DLVP [5], which forms a different approach of VP through memory address prediction. Finally, Sheikh and Hower presented a composite predictor [6] that combines different predictor models in order to increase coverage in the cost of some additional complexity imposed on the prediction critical path. Table. ETAGE is a context-based equality predictor that essentially copies the TAGE branch predictor structure.…”

Section: Related Workmentioning

confidence: 99%

Value Speculation through Equality Prediction

Kalaitzidis

Seznec

2019

2019 IEEE 37th International Conference on Computer Design (ICCD)

View full text Add to dashboard Cite

Modern context-based value predictors tightly associate recurring values with instructions and contexts by building confidence upon them. However, when execution monotony exists in the form of intervals, the potential prediction coverage is limited, since prediction confidence is reset at the beginning of each new interval. In this paper, we address this challenge by introducing the notion of Equality Prediction (EP), which represents the binary facet of value prediction. Following a twofold decision scheme (similar to branch prediction), EP makes use of control-flow history to determine equality between the last committed result read at fetch time, and the result of the fetched occurrence. When equality is predicted with high confidence, the read value is used. Our experiments show that this technique obtains the same level of performance as previously proposed state-of-the-art context-based predictors. However, by virtue of better exploiting patterns of interval equality, our design complements the established way that value prediction is performed, and when combined with contemporary prediction models, improves the delivered speedup by 19% on average.

show abstract

Efficient Load Value Prediction Using Multiple Predictors and Filters

Cited by 11 publications

References 15 publications

Calipers: A Criticality-aware Framework for Modeling Processor Performance

Calipers: A Criticality-aware Framework for Modeling Processor Performance

Leveraging Value Equality Prediction for Value Speculation

Value Speculation through Equality Prediction

Contact Info

Product

Resources

About