Puncalc: task-based parallelism and speculative reevaluation in spreadsheets

Bock, Alexander Asp; Biermann, Florian

doi:10.1007/s11227-019-02823-8

Cited by 3 publications

(9 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…When there is only a single cell in a cell's support set, it would be wasteful to enqueue it in the global queue as the thread could just directly evaluate it instead and save concurrent pulls from the global queue in search of more work. In our previous work 15 on a task‐based spreadsheet interpreter, a variant was developed that used thread‐local queues for evaluating such sequential chains of supported cells. The algorithm presented in this article could be similarly extended where the single supported cell is enqueued in a recalculation thread's local queue.…”

Section: Discussionmentioning

confidence: 99%

“…In earlier work, 15 we presented a task‐based parallel spreadsheet interpreter that automatically discovers parallelism, finds cyclic references in parallel and also targets shared‐memory multiprocessors without requiring modification of the spreadsheetor user interaction. Tasks were spawned using the Task Parallel Library (TPL) 16 and the algorithm obtained roughly a maximum 16× speed‐up on the same set of benchmark spreadsheets used in this article.…”

Section: Related Workmentioning

confidence: 99%

“…Cell ownership was originally introduced in earlier work 15 for two purposes. First, as a way for threads to claim exclusive ownership of a cell.…”

Section: Parallel Funcalcmentioning

confidence: 99%

“…The groups were then run on multiple processors using the TPL 16 . The article presented three extensions, two of which incorporate work from the task‐based spreadsheet interpreter 15 and Biermann's cell array rewriting 11 . While the algorithm yielded overall positive speed‐ups, partitioning time was largely dominated by applying the cost semantic rules to estimate the evaluation time of each cell resulting in partitioning taking up toward one hour to finish.…”

Section: Related Workmentioning

confidence: 99%

“…A resolution strategy must be in place to break the cycle and avoid the five threads waiting for each other indefinitely. For example, our task‐based spreadsheet interpreter 15 used speculative reevaluation to detect and resolve cycles in parallel. Threads with lower identifiers were given precedence over threads with higher identifiers.…”

Section: Parallel Funcalcmentioning

confidence: 99%

See 4 more Smart Citations

Parallel spreadsheet evaluation and dynamic cycle detection

Bock

2021

Concurrency and Computation

Self Cite

View full text Add to dashboard Cite

Summary It was estimated there would be 72 million users using spreadsheets monthly in 2017 some of which build complex financial, scientific, and mathematical models. Most of these end‐users are not trained IT professionals but domain experts. In the age of multicore computing and ever‐increasing amounts of data, how can end‐users access this powerful, parallel hardware to accelerate spreadsheet computation? Some existing solutions are usually not fully automatic and require a level of interaction from end‐users to facilitate parallel execution. Ideally, an end‐user tool would transparently exploit the underlying hardware and automatically discover available parallelism in the spreadsheet without any required interaction. This article presents an algorithm for automatic parallel evaluation of the cells in a spreadsheet and dynamic, parallel cycle detection. It is implemented in the Funcalc research spreadsheet application which supports higher‐order functions in the spreadsheet paradigm. Altogether, this promises a powerful and expressive platform for end‐user development. Our results on a 48 logical core machine show a maximum 7–14× speed‐up on a set of benchmark spreadsheets and a maximum 21–32× speed‐up on a set of synthetic spreadsheets with predefined topologies.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

“…Cell ownership was originally introduced in earlier work 15 for two purposes. First, as a way for threads to claim exclusive ownership of a cell.…”

Section: Parallel Funcalcmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Parallel Funcalcmentioning

confidence: 99%

See 3 more Smart Citations

Parallel spreadsheet evaluation and dynamic cycle detection

Bock

2021

Concurrency and Computation

Self Cite

View full text Add to dashboard Cite

show abstract

The Power of Spreadsheet Computations

Tyszkiewicz

2020

Fields of Logic and Computation III

View full text Add to dashboard Cite

Building Custom Spreadsheet Functions with Python: End-User Software Engineering Approach

Elserwy,

Raslan,

Ali

et al. 2024

JSEA

View full text Add to dashboard Cite

End-user computing empowers non-developers to manage data and applications, enhancing collaboration and efficiency. Spreadsheets, a prime example of end-user programming environments widely used in business for data analysis. However, Excel functionalities have limits compared to dedicated programming languages. This paper addresses this gap by proposing a prototype for integrating Python's capabilities into Excel through on-premises desktop to build custom spreadsheet functions with Python. This approach overcomes potential latency issues associated with cloud-based solutions. This prototype utilizes Excel-DNA and IronPython. Excel-DNA allows creating custom Python functions that seamlessly integrate with Excel's calculation engine. IronPython enables the execution of these Python (CSFs) directly within Excel. C# and VSTO add-ins form the core components, facilitating communication between Python and Excel. This approach empowers users with a potentially open-ended set of Python (CSFs) for tasks like mathematical calculations, statistical analysis, and even predictive modeling, all within the familiar Excel interface. This prototype demonstrates smooth integration, allowing users to call Python (CSFs) just like standard Excel functions. This research contributes to enhancing spreadsheet capabilities for end-user programmers by leveraging Python's power within Excel. Future research could explore expanding data analysis capabilities by expanding the (CSFs) functions for complex calculations, statistical analysis, data manipulation, and even external library integration. The possibility of integrating machine learning models through the (CSFs) functions within the familiar Excel environment.

show abstract

Puncalc: task-based parallelism and speculative reevaluation in spreadsheets

Cited by 3 publications

References 15 publications

Parallel spreadsheet evaluation and dynamic cycle detection

Parallel spreadsheet evaluation and dynamic cycle detection

The Power of Spreadsheet Computations

Building Custom Spreadsheet Functions with Python: End-User Software Engineering Approach

Contact Info

Product

Resources

About