Schedule Synthesis for Halide Pipelines on GPUs

Sioutas, Savvas; Stuijk, Sander; Basten, Twan; Corporaal, Henk; Somers, Lou

doi:10.1145/3406117

Cited by 11 publications

(13 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We achieve a geomean speedup over Sioutas et al 's Halide GPU autoscheduler [Sioutas et al 2020] of 1.07× in the One Shot case, 1.30× in the Top 5 case, and 1.66× in the autotuning case. Our autotuned results are on par with our best known manual schedules (0.95×).…”

Section: Discussionmentioning

confidence: 89%

“…No benchmarking is involved in selecting the schedule. This approach is directly comparable to Sioutas et al [Sioutas et al 2020]'s autoscheduler, and can run in seconds.…”

Section: Resultsmentioning

confidence: 99%

“…In many cases these human schedules substantially improve on the ones found in the Halide repository, which were used as-is by prior work. We also compare to the best existing Halide GPU autoscheduler from Sioutas et al [Sioutas et al 2020]. We use our technique in 3 different modes of operation, which trade off compile time and the ability to benchmark for quality of results:…”

Section: Resultsmentioning

confidence: 99%

“…• A new automatic scheduling algorithm that scales orders of magnitude better than prior work, making it possible to efficiently explore a large, rich space of GPU schedules. It delivers state of the art performance on a suite of real world imaging and vision pipelines, with a geomean speedup of up to 1.7× over the prior state of the art GPU autoscheduler [Sioutas et al 2020], and competitive (0.95×) with what the best human experts were able to achieve in an active effort to beat our automatic results. • A set of schedule features that capture the architectural intricacies required to predict performance of GPU programs.…”

Section: Introductionmentioning

confidence: 88%

“…Throughput of our technique at three levels of compile time budget (from one shot with no benchmarkto autotuning 1600 samples) relative to the prior state-of-the-art GPU autoscheduler [Sioutas et al 2020] and highly tuned human expert schedules. The benchmarks are a super-set of those in prior work [Adams et al 2019;Sioutas et al 2020], and span a diverse set of imaging and learning programs from a few to around one hundred stages. For the One Shot and Top 5 modes, the results presented are the median time over 100 independent trials.…”

Section: States Evaluatedmentioning

confidence: 99%

See 4 more Smart Citations

Efficient automatic scheduling of imaging and vision pipelines for the GPU

Anderson

Adams

et al. 2021

Proc. ACM Program. Lang.

View full text Add to dashboard Cite

We present a new algorithm to quickly generate high-performance GPU implementations of complex imaging and vision pipelines, directly from high-level Halide algorithm code. It is fully automatic, requiring no schedule templates or hand-optimized kernels. We address the scalability challenge of extending search-based automatic scheduling to map large real-world programs to the deep hierarchies of memory and parallelism on GPU architectures in reasonable compile time. We achieve this using (1) a two-phase search algorithm that first ‘freezes’ decisions for the lowest cost sections of a program, allowing relatively more time to be spent on the important stages, (2) a hierarchical sampling strategy that groups schedules based on their structural similarity, then samples representatives to be evaluated, allowing us to explore a large space with few samples, and (3) memoization of repeated partial schedules, amortizing their cost over all their occurrences. We guide the process with an efficient cost model combining machine learning, program analysis, and GPU architecture knowledge. We evaluate our method’s performance on a diverse suite of real-world imaging and vision pipelines. Our scalability optimizations lead to average compile time speedups of 49x (up to 530x). We find schedules that are on average 1.7x faster than existing automatic solutions (up to 5x), and competitive with what the best human experts were able to achieve in an active effort to beat our automatic results.

show abstract

Section: Discussionmentioning

confidence: 89%

“…No benchmarking is involved in selecting the schedule. This approach is directly comparable to Sioutas et al [Sioutas et al 2020]'s autoscheduler, and can run in seconds.…”

Section: Resultsmentioning

confidence: 99%

Section: Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 88%

Section: States Evaluatedmentioning

confidence: 99%

See 3 more Smart Citations

Efficient automatic scheduling of imaging and vision pipelines for the GPU

Anderson

Adams

et al. 2021

Proc. ACM Program. Lang.

View full text Add to dashboard Cite

show abstract

A compression-based memory-efficient optimization for out-of-core GPU stencil computation

et al. 2023

View full text Add to dashboard Cite

Stencil computation is an extensively-utilized class of scientific-computing applications that can be efficiently accelerated by graphics processing units (GPUs). Out-of-core approaches enable a GPU to handle large stencil codes whose data size is beyond the memory capacity of the GPU. However, current research on out-of-core stencil computation primarily focus on minimizing the amount of data transferred between the CPU and GPU. Few studies consider simultaneously optimizing data transfer and kernel execution. To fill the research gap, this work presents a synergy between on-and off-chip data reuse for out-of-core stencil codes, termed SO2DR. First, overlapping regions between data chunks are shared in the off-chip memory to eliminate redundant CPU-GPU data transfer. Secondly, redundant computation at the off-chip memory level is intentionally introduced to decouple kernel execution from region sharing, hence enabling data reuse in the onchip memory. Experimental results demonstrate that SO2DR significantly enhances the kernel-execution performance while reducing the CPU-GPU data-transfer time. Specifically, SO2DR achieves average speedups of 2.78× and 1.14× for five stencil benchmarks, compared to an out-of-core stencil code which is free of redundant transfer and computation, and an in-core stencil code which is free of data transfer, respectively.

show abstract

Approximation-Aware Design of an Image-Based Control System

Mohamed

Goswami

et al. 2020

IEEE Access

View full text Add to dashboard Cite

Image-based control (IBC) systems are common in many modern applications. In such systems, image-based sensing imposes massive compute workload, making them challenging to implement on embedded platforms. Approximate image processing is a way to handle this challenge. In essence, approximation reduces the workload at the cost of additional sensor noise. In this work, we propose an approximation-aware design approach for optimizing the energy, memory and performance of an IBC system, making it suitable for embedded implementation. First, we perform compute-and data-centric approximations and evaluate its impact on the energy efficiency, memory utilization and closed-loop qualityof-control (QoC) of the IBC system. We observe that the workload reductions due to approximations allow mapping these lighter approximated IBC tasks to embedded platforms with lower power consumption while still ensuring proper system functionality. Therefore, we explore the interplay between approximations and platform mappings to improve the energy-efficiency of IBC systems. Further, an IBC system operates under several environmental scenarios e.g., weather conditions. We evaluate the sensitivity of the IBC system to our approximation-aware design approach when operated under different scenarios and perform a failure probability (FP) analysis using Monte-Carlo simulations to analyze the robustness of the approximate system. Finally, we design an optimal approximation-aware controller that models the approximation error as sensor noise and show QoC improvements. We demonstrate the effectiveness of our approach using a concrete case-study of a lane keeping assist system (LKAS) using a heterogeneous NVIDIA AGX Xavier embedded platform in a hardware-in-the-loop (HiL) framework. We show energy and memory reduction of up to 92% and 88% respectively, for 44% QoC improvements with respect to the accurate implementation. We show that our approximation-aware design approach has an FP (per km) ≤ 9.6 × 10 −6 %.

show abstract

Schedule Synthesis for Halide Pipelines on GPUs

Cited by 11 publications

References 24 publications

Efficient automatic scheduling of imaging and vision pipelines for the GPU

Efficient automatic scheduling of imaging and vision pipelines for the GPU

A compression-based memory-efficient optimization for out-of-core GPU stencil computation

Approximation-Aware Design of an Image-Based Control System

Contact Info

Product

Resources

About