iMLBench: A Machine Learning Benchmark Suite for CPU-GPU Integrated Architectures
The DOI matches its Crossref record; 47 of the 50 references it cites were found in a registry or an index.
Abstract
Utilizing heterogeneous accelerators, especially GPUs, to accelerate machine learning tasks has shown to be a great success in recent years. GPUs bring huge performance improvements to machine learning and greatly promote the widespread adoption of machine learning. However, the discrete CPU-GPU architecture design with high PCIe transmission overhead decreases the GPU computing benefits in machine learning training tasks. To overcome such limitations, hardware vendors release CPU-GPU integrated architectures with shared unified memory. In this article, we design a benchmark suite for machine learning training on CPU-GPU integrated architectures, called iMLBench, covering a wide range of machine learning applications and kernels. We mainly explore two features on integrated architectures: 1) zero-copy, which means that the PCIe overhead has been eliminated for machine learning tasks and 2) co-running, which means that the CPU and the GPU co-run together to process a single machine learning task. Our experimental results on iMLBench show that the integrated architecture brings an average 7.1× performance improvement over the original implementations. Specifically, the zero-copy design brings 4.65× performance improvement, and co-running brings 1.78× improvement. Moreover, integrated architectures exhibit promising results from both performance-per-dollar and energy perspectives, achieving 6.50× performance-price ratio while 4.06× energy efficiency over discrete GPUs. The bench…
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“The computing time can be accelerated by using graphics processing units (GPUs) to implement the ANNs (see Fig. 1a), as these can perform multiple operations in parallel10–12.”
From Hardware implementation of memristor-based artificial neural networks (2024) · cites this paper as [11]
Citations
Cites
The PARSEC benchmark suite: characterization and architectural implications
Bienia, C., Kumar, S., Singh, J. P., & Li, K. (2008) · Proceedings of the 17th international conference on Parallel architectures and compilation techniques · doi:10.1145/1454115.1454128 · Cited by 3,444 · cited 3×
“Rodinia [6], PARSEC [7], and Parboil [8] are comprehensive benchmark suites targeting general HPC domains, but not for machine learning workloads.”
BackgroundMethodRodinia: A benchmark suite for heterogeneous computing
Che, S., Boyer, M., Meng, J., Tarjan, D., Sheaffer, J. W., Lee, S.-H., & Skadron, K. (2009) · 2009 IEEE International Symposium on Workload Characterization (IISWC) · doi:10.1109/iiswc.2009.5306797 · Cited by 3,018 · cited 10×
“Rodinia [6] is a spectrum of applications and kernels which target CPU and GPU platforms.”
BackgroundMethodInfluentialFast Algorithms for Convolutional Neural Networks
Lavin, A., & Gray, S. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.435 · Cited by 915 · cited 1×
“By employing Winograd minimal filtering algorithm, Winograd convolution reduces the multiplication complexity by at least 2.25 (cid:1) [54] compared to the original implementation, which shows great success in accelerating CNNs, such as VGG [54] and ResNet [55].”
MethodInfluentialA Survey of Model Compression and Acceleration for Deep Neural Networks
Cheng, Y., Wang, D., Zhou, P., & Zhang, T. (2017) · arXiv · doi:10.48550/arxiv.1710.09282 · Cited by 857 · cited 1×
“For example, large neural networks may be trained for several days [42].”
BackgroundMars: a MapReduce framework on graphics processors
He, B., Fang, W., Luo, Q., Govindaraju, N. K., & Wang, T. (2008) · Proceedings of the 17th international conference on Parallel architectures and compilation techniques · doi:10.1145/1454115.1454152 · Cited by 715 · cited 1×
“The Mars project [36] includes a series of data-intensive workloads for MapReduce on discrete GPUs.”
BackgroundSelf-citationParboil: A Revised Benchmark Suite for Scientific and Commercial Throughput Computing
Stratton, J. A., Rodrigues, C. I., Sung, I.-J., Obeid, N., Chang, L., Anssari, N., Liu, G. D., & Hwu, W. (2012) · Cited by 707
The Scalable Heterogeneous Computing (SHOC) benchmark suite
Danalis, A., Marin, G., McCurdy, C., Meredith, J. S., Roth, P. C., Spafford, K., Tipparaju, V., & Vetter, J. S. (2010) · Proceedings of the 3rd Workshop on General-Purpose Computation on Graphics Processing Units · doi:10.1145/1735688.1735702 · Cited by 671 · cited 1×
“SHOC benchmark [37] is comprised of programs, testing the performance and stability of systems containing GPUs and multi-core processors.”
BackgroundAn integrated GPU power and performance model
Hong, S., & Kim, H. (2010) · Proceedings of the 37th annual international symposium on Computer architecture · doi:10.1145/1815961.1815998 · Cited by 483 · cited 1×
“Fortunately, the integrated architectures usually have a good performance-per-Watt ratio [43].”
BackgroundPMLB: a large benchmark suite for machine learning evaluation and comparison
Olson, R. S., La Cava, W., Orzechowski, P., Urbanowicz, R. J., & Moore, J. H. (2017) · BioData Mining · doi:10.1186/s13040-017-0154-4 · Cited by 365 · cited 3×
“Currently, there are many benchmark suites targeting heterogeneous systems [5], [6], [7], [8], [9], [10], [11], [12].”
BackgroundAccelerating Binarized Neural Networks: Comparison of FPGA, CPU, GPU, and ASIC
Nurvitadhi, E., Sheffield, D., Sim, J., Mishra, A., Venkatesh, G., & Marr, D. (2016) · 2016 International Conference on Field-Programmable Technology (FPT) · doi:10.1109/fpt.2016.7929192 · Cited by 342 · cited 1×
“…benchmark suites, such as PMLB [9], MLPerf [10], MLbench [11], EEMBC MLMark Benchmark [12], and HeteroSpark [13], have been developed to explore the performance behavior on multi-core or many-core parallel systems, especially with accelerators like GPU, FPGA, and ASIC designs [13], [38], [39].”
Prediction by linear regression on a quantum computer
Schuld, M., Sinayskiy, I., & Petruccione, F. (2016) · Physical Review A · doi:10.1103/physreva.94.022342 · Cited by 274 · cited 1×
“Linear regression has been used in applications such as traf-fic prediction, price estimation, and performance prediction [51].”
MethodMicroarchitecture-Independent Workload Characterization
Hoste, K., & Eeckhout, L. (2007) · IEEE Micro · doi:10.1109/mm.2007.56 · Cited by 190 · cited 1×
“We use the Microarchitecture-Independent Workload Characterization (MICA) framework developed by Hoste and Eeckhout [57].”
MethodAccelerating recurrent neural networks in analytics servers: Comparison of FPGA, CPU, GPU, and ASIC
Nurvitadhi, E., Sim, J., Sheffield, D., Mishra, A., Krishnan, S., & Marr, D. (2016) · 2016 26th International Conference on Field Programmable Logic and Applications (FPL) · doi:10.1109/fpl.2016.7577314 · Cited by 175
MLPerf: An Industry Standard Benchmark Suite for Machine Learning Performance
Mattson, P., Reddi, V. J., Cheng, C., Coleman, C., Diamos, G., Kanter, D., Micikevicius, P., Patterson, D., Schmuelling, G., Tang, H., et al. (2020) · IEEE Micro · doi:10.1109/mm.2020.2974843 · Cited by 171 · cited 4×
“…work-load selection, we refer to the Top Ten machine learning algorithms [14], [15] as well as the implementation of other machine learning benchmarks including MLPerf [10], MLBench [11], DeepBench [46], AIBench [47], and Deep500 [48], which are popular and of interest to a wide range of domains.”
BackgroundMethodInfluentialA Survey on Spark Ecosystem: Big Data Processing Infrastructure, Machine Learning, and Applications
Tang, S., He, B., Yu, C., Li, Y., & Li, K. (2020) · IEEE Transactions on Knowledge and Data Engineering · doi:10.1109/tkde.2020.2975652 · Cited by 153 · cited 1×
“Parallel processing [16], [17], [18], [19], [20], [21], [22] is a hot research topic in recent years, which boost the application of machine learning.”
BackgroundSelf-citationRemote procedure calls and Java Remote Method Invocation
Waldo, J. (1998) · IEEE Concurrency · doi:10.1109/4434.708248 · Cited by 106 · cited 1×
“The HeteroSpark platform implements data communication between the CPU and the GPU by Java RMI technique [40], and conducts machine learning experiments only on discrete GPUs, which wastes the CPU computing resources.”
MethodAccelerating MapReduce on a coupled CPU-GPU architecture
Chen, L., Huo, X., & Agrawal, G. (2012) · International Conference for High Performance Computing, Networking, Storage and Analysis (SC) · doi:10.1109/sc.2012.16 · Cited by 105 · cited 1×
“The integrated architectures arebeing adopted to various applications, including data management systems [23], [24], stream processing systems [25], PlayStation [26], network packet processing [27], [28], and supercomputing [29], [30], [31], [32], [33].”
MethodAchieving Exascale Capabilities through Heterogeneous Computing
Schulte, M., Ignatowski, M., Loh, G. H., Beckmann, B. M., Brantley, W. C., Gurumurthi, S., Jayasena, N., Paul, I., Reinhardt, S. K., & Rodgers, G. P. (2015) · IEEE Micro · doi:10.1109/mm.2015.71 · Cited by 103 · cited 1×
“The integrated architectures arebeing adopted to various applications, including data management systems [23], [24], stream processing systems [25], PlayStation [26], network packet processing [27], [28], and supercomputing [29], [30], [31], [32], [33].”
MethodUnderstanding Co-Running Behaviors on Integrated CPU/GPU Architectures
Zhang, F., Zhai, J., He, B., Zhang, S., & Chen, W. (2016) · IEEE Transactions on Parallel and Distributed Systems · doi:10.1109/tpds.2016.2586074 · Cited by 99 · cited 10×
“We have analyzed many performance indicators in hardware profiler for machine learning tasks, and find that these performance counters are much different from those in traditional heterogeneous benchmark suites [5], [6], [7], [8].”
BackgroundMethodInfluentialSelf-citationHeteroSpark: A heterogeneous CPU/GPU Spark platform for machine learning algorithms
Li, P., Luo, Y., Zhang, N., & Cao, Y. K. (2015) · IEEE International Conference on Networking, Architecture and Storage (NAS) · doi:10.1109/nas.2015.7255222 · Cited by 96 · cited 4×
“…benchmark suites, such as PMLB [9], MLPerf [10], MLbench [11], EEMBC MLMark Benchmark [12], and HeteroSpark [13], have been developed to explore the performance behavior on multi-core or many-core parallel systems, especially with accelerators like GPU, FPGA, and ASIC designs [13], [38], [39].”
BackgroundMethodInfluentialAccelerating MapReduce on a coupled CPU-GPU architecture
Chen, L., Huo, X., & Agrawal, G. (2012) · IEEE International Conference on High Performance Computing, Data, and Analytics · doi:10.5555/2388996.2389030 · Cited by 88
On the Efficacy of a Fused CPU+GPU Processor (or APU) for Parallel Computing
Daga, M., Aji, A. M., & Feng, W.-C. (2011) · 2011 Symposium on Application Accelerators in High-Performance Computing · doi:10.1109/saahpc.2011.29 · Cited by 84 · cited 1×
“For example, Daga and others [41] studied the reduction in data transfer time on integrated”
BackgroundMining Hard Samples Globally and Efficiently for Person Reidentification
Sheng, H., Zheng, Y., Ke, W., Yu, D., Cheng, X., Lyu, W., & Xiong, Z. (2020) · IEEE Internet of Things Journal · doi:10.1109/jiot.2020.2980549 · Cited by 82 · cited 1×
“[24], stream processing systems [25], PlayStation [26], network packet processing [27], [28], and supercomputing [29], [30], [31], [32], [33].”
BackgroundDesign and Analysis of an APU for Exascale Computing
Vijayaraghavan, T., Karunanithi, A., Kayıran, O., Meswani, M. R., Paul, I., Poremba, M., Raasch, S. E., Reinhardt, S. K., Sadowski, G., Sridharan, V., et al. (2017) · IEEE International Symposium on High-Performance Computer Architecture (HPCA) · doi:10.1109/hpca.2017.42 · Cited by 75 · cited 1×
“[24], stream processing systems [25], PlayStation [26], network packet processing [27], [28], and supercomputing [29], [30], [31], [32], [33].”
BackgroundAPUNet: Revitalizing GPU as Packet Processing Accelerator
Go, Y., Jamshed, M. A., Moon, Y. S., Hwang, C. H., & Park, K. (2017) · Networked Systems Design and Implementation · Cited by 72
Optimizing batched winograd convolution on GPUs
Yan, D., Wang, W., & Chu, X. (2020) · ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming · doi:10.1145/3332466.3374520 · Cited by 69 · cited 1×
“By employing Winograd minimal filtering algorithm, Winograd convolution reduces the multiplication complexity by at least 2.25 (cid:1) [54] compared to the original implementation, which shows great success in accelerating CNNs, such as VGG [54] and ResNet [55].”
MethodInfluentialMT-MPI: multithreaded MPI for many-core environments
Si, M., Peña, A. J., Balaji, P., Takagi, M., & Ishikawa, Y. (2014) · International Conference on Supercomputing · doi:10.1145/2597652.2597658 · Cited by 47 · cited 1×
“Parallel processing [16], [17], [18], [19], [20], [21], [22] is a hot research topic in recent years, which boost the application of machine learning.”
BackgroundFinePar: Irregularity-aware fine-grained workload partitioning on integrated architectures
Zhang, F., Wu, B., Zhai, J., He, B., & Chen, W. (2017) · Proceedings of the CGO/Proceedings of the ... CGO · doi:10.1109/cgo.2017.7863726 · Cited by 44 · cited 1×
“[24], stream processing systems [25], PlayStation [26], network packet processing [27], [28], and supercomputing [29], [30], [31], [32], [33].”
BackgroundSelf-citationBenchmarking Contemporary Deep Learning Hardware and Frameworks: A Survey of Qualitative Metrics
Dai, W., & Berleant, D. (2019) · 2019 IEEE First International Conference on Cognitive Machine Intelligence (CogMI) · doi:10.1109/cogmi48466.2019.00029 · Cited by 43 · cited 2×
“Benchmark suites simulate executing applications on a given system for users to evaluate the properties of various systems and provide valuable information to developers, manufacturers, and customers [6], [35].”
AIBench: Towards Scalable and Comprehensive Datacenter AI Benchmarking
Gao, W., Luo, C., Wang, L., Xiong, X., Chen, J., Hao, T., Jiang, Z., Fan, F., Du, M., Huang, Y., et al. (2019) · Lecture notes in computer science · doi:10.1007/978-3-030-32813-9_1 · Cited by 40 · cited 1×
“…work-load selection, we refer to the Top Ten machine learning algorithms [14], [15] as well as the implementation of other machine learning benchmarks including MLPerf [10], MLBench [11], DeepBench [46], AIBench [47], and Deep500 [48], which are popular and of interest to a wide range of domains.”
MethodSelf-citationMLbench: benchmarking machine learning services against human experts
Liu, Y., Zhang, H., Zeng, L., Wu, W., & Zhang, C. (2018) · Proceedings of the VLDB Endowment · doi:10.14778/3231751.3231770 · Cited by 37 · cited 4×
“…work-load selection, we refer to the Top Ten machine learning algorithms [14], [15] as well as the implementation of other machine learning benchmarks including MLPerf [10], MLBench [11], DeepBench [46], AIBench [47], and Deep500 [48], which are popular and of interest to a wide range of domains.”
BackgroundMethodInfluentialSelf-citationA comprehensive performance analysis of HSA and OpenCL 2.0
Mukherjee, S., Sun, Y., Blinzer, P., Ziabari, A. K., & Kaeli, D. R. (2016) · IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS) · doi:10.1109/ispass.2016.7482093 · Cited by 35 · cited 1×
“After the data objects of shared virtual memory are updated by the GPU, they can be directly accessed as normal data structures by the CPU due to fine-grained sharing [56].”
MethodFinePar: irregularity-aware fine-grained workload partitioning on integrated architectures
Zhang, F., Wu, B., Zhai, J., et al. (2017) · Symposium on Code Generation and Optimization · doi:10.5555/3049832.3049836 · Cited by 35
Self-citationFineStream: Fine-Grained Window-Based Stream Processing on CPU-GPU Integrated Architectures.
Zhang, F., Yang, L., Zhang, S., He, B., Lü, W., & Du, X. (2020) · USENIX Annual Technical Conference · Cited by 31
Self-citationDeep-Aligned Convolutional Neural Network for Skeleton-Based Action Recognition and Segmentation
Hosseini, B., Montagné, R., & Hammer, B. (2019) · Data Science and Engineering · doi:10.1007/s41019-020-00123-3 · Cited by 28 · cited 1×
“CNN is a class of deep neural networks and is used extensively in image and video recognition [53], image classification, recommender systems, medical image analysis, natural language processing, and financial time series.”
BackgroundUsing Dynamic Duty Cycle Modulation to Improve Energy Efficiency in High Performance Computing
Bhalachandra, S., Porterfield, A., & Prins, J. F. (2015) · IEEE International Symposium on Parallel & Distributed Processing, Workshops and Phd Forum · doi:10.1109/ipdpsw.2015.144 · Cited by 27 · cited 1×
“Parallel processing [16], [17], [18], [19], [20], [21], [22] is a hot”
BackgroundAutomatic Irregularity-Aware Fine-Grained Workload Partitioning on Integrated Architectures
Zhang, F., Zhai, J., Wu, B., He, B., Chen, W., & Du, X. (2019) · IEEE Transactions on Knowledge and Data Engineering · doi:10.1109/tkde.2019.2940184 · Cited by 23 · cited 1×
“[24], stream processing systems [25], PlayStation [26], network packet processing [27], [28], and supercomputing [29], [30], [31], [32], [33].”
BackgroundSelf-citationDIDO: Dynamic Pipelines for In-Memory Key-Value Stores on Coupled CPU-GPU Architectures
Zhang, K., Hu, J., He, B., & Hua, B. (2017) · IEEE International Conference on Data Engineering (ICDE) · doi:10.1109/icde.2017.120 · Cited by 21
Self-citationUnderstanding and overcoming parallelism bottlenecks in ForkJoin applications
Pinto, G., Canino, A., Castor, F., Xu, G. H., & Liu, Y. D. (2017) · IEEE/ACM International Conference on Automated Software Engineering (ASE) · doi:10.1109/ase.2017.8115687 · Cited by 19
NVIDIA'S Tegra K1 system-on-chip
Ditty, M. J. T., Architecture, T., Montrym, J. S., & Wittenbrink, C. M. (2014) · IEEE Hot Chips Symposium · doi:10.1109/hotchips.2014.7478809 · Cited by 15 · cited 2×
“To address the limitation of discrete GPUs, hardware vendors develop the CPU-GPU integrated architectures, such as the AMD APU [2], Intel integrated graphics processors [3], and Nvidia Tegra K1 processors [4].”
BackgroundMethodSCALO: Scalability-Aware Parallelism Orchestration for Multi-Threaded Workloads
Georgakoudis, G., Vandierendonck, H., Thoman, P., de Supinski, B. R., Fahringer, T., & Nikolopoulos, D. S. (2017) · ACM Transactions on Architecture and Code Optimization · doi:10.1145/3158643 · Cited by 11
Transparent partial page migration between CPU and GPU
Zhang, S., Qin, Z., Yang, Y., Shen, L., & Wang, Z. (2019) · Frontiers of Computer Science · doi:10.1007/s11704-018-7386-4 · Cited by 10 · cited 1×
“However, the PCIe data transmission overhead can be an important factor of the overall performance of the GPU processing [1].”
BackgroundChen, X., Chen, Y., Ronak, B., He, J., He, B., Wong, W., & Chen, D. (2020) · National University of Singapore · Cited by 9
Self-citationMemory-Efficient and Skew-Tolerant MapReduce Over MPI for Supercomputing Systems
Gao, T., Guo, Y., Zhang, B., Cicotti, P., Lu, Y., Balaji, P., & Taufer, M. (2020) · IEEE Transactions on Parallel and Distributed Systems · doi:10.1109/tpds.2019.2932066 · Cited by 9 · cited 1×
“Parallel processing [16], [17], [18], [19], [20], [21], [22] is a hot”
BackgroundEnergy-Efficient Graphics and Multimedia in 28-nm Carrizo Accelerated Processing Unit
Krishnan, G., Bouvier, D., & Naffziger, S. (2016) · IEEE Micro · doi:10.1109/mm.2016.24 · Cited by 7 · cited 6×
“Both Intel and AMD have released their CPU-GPU integrated designs, e.g., AMD’s Accelerated Processing Unit (APU) and Intel’s HD/Iris/Iris Pro Graphics.”
MethodInfluentialFu, H., Tang, S., He, B., Yu, C., & Sun, J. (2018) · Proceedings of the International Conference on Parallel Processing · doi:10.1145/3225058.3225077 · Cited by 4 · cited 1×
“Parallel processing [16], [17], [18], [19], [20], [21], [22] is a hot”
BackgroundSelf-citationResource Centric Characterization and Classification of Applications Using KMeans for Multicores
Jain, P., & Surve, S. (2019) · International Conference on Information Networking · doi:10.1109/icoin.2019.8717981 · Cited by 4
Research works printed but not linked: 1
- imandoust, 2013, Application of k-nearest neighbor (KNN) approach for predicting economic events: Theoretical background, Int J Eng Res Appl, 3, 605
Cited by
Hardware implementation of memristor-based artificial neural networks
Aguirre, F., Sebastian, A., Le Gallo, M., Song, W., Wang, T., Yang, J. J., Lu, W., Chang, M.-F., Ielmini, D., Yang, Y., et al. (2024) · Nature Communications · doi:10.1038/s41467-024-45670-9 · Cited by 564
The growing memristor industry
Lanza, M., Pazos, S., Aguirre, F., Sebastian, A., Le Gallo, M., Alam, S. M., Ikegawa, S., Yang, J. J., Vianello, E., Chang, M.-F., et al. (2025) · Nature · doi:10.1038/s41586-025-08733-5 · Cited by 177
“Graphics processing units (GPUs) are specialized electronic circuits that excel in parallel processing tasks 10,11 ; some of them are able to carry out (on average) 100–1,000 trillion operations per second—namely, TOPS, or TFLOPS if the data are codified in floating-point format 12 .”
Breaking the Edge: Enabling Efficient Neural Network Inference on Integrated Edge Devices
Zhang, F., Zhang, C., Guan, J., Zhou, Q., Chen, K., Zhang, X., He, B., Zhai, J., & Du, X. (2025) · IEEE Transactions on Cloud Computing · doi:10.1109/tcc.2025.3559346 · Cited by 108
“…of machine learning workloads on integrated architectures, primarily within cloud environments [28], [29], [30], [31], [32], [33]. iMLBench [30] studies the training stage of typical neural networks on CPU-GPU integrated devices, leaving the inference stage on edge devices for further…”
InfluentialSelf-citationA Benchmark Suite for Improving Performance Portability of the SYCL Programming Model
Jin, Z., & Vetter, J. S. (2023) · IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS) · doi:10.1109/ispass57527.2023.00041 · Cited by 62
Exploring Query Processing on CPU-GPU Integrated Edge Device
Liu, J., Zhang, F., Li, H., Wang, D., Wan, W., Fang, X., Zhai, J., & Du, X. (2022) · IEEE Transactions on Parallel and Distributed Systems · doi:10.1109/tpds.2022.3177811 · Cited by 25
“[49] developed a machine learning benchmark, called iMLBench, for”
Self-citation
Metadata
- Authors
- Zhang, C., Zhang, F., Guo, X., He, B., Zhang, X., & Du, X.
- Affiliations
- Renmin University of China (CN)
- National University of Singapore (SG)
- Venue
- IEEE Transactions on Parallel and Distributed Systems
- Published
- 2021-07-01
- Volume
- 32
- Issue
- 7
- Pages
- 1740-1752
- ISSN
- 1045-9219 (ISSN-L)
- 1558-2183
- 2161-9883
- OpenAlex
- W3116059414
- Cited by
- 27
- References
- 50
- Citations by year
- 2025: 7 · 2024: 7 · 2023: 4 · 2022: 4 · 2021: 5
- FWCI
- 2.73
- Type
- journal article
- Language
- en
- Open access
- closed
- Licence
- Locations
- IEEE Transactions on Parallel and Distributed Systems (publishedVersion)
- Funders
- National Key Research and Development Program of China (2018YFB1004401)
- National Natural Science Foundation of China (61802412, 61732014, 61972403, 62072459, U1911203)
- State Key Laboratory of Computer Architecture (CARCHA202007)
- MoE AcRF Tier 1 (T1 251RES1824)
- Tier 2 (MOE2017-T2-1-122)
- Topics
- Physical Sciences › Computer Science › Hardware and Architecture › Parallel Computing and Optimization Techniques
- Physical Sciences › Computer Science › Hardware and Architecture › Embedded Systems Design Techniques
- Physical Sciences › Computer Science › Computer Networks and Communications › Interconnection Networks and Systems
- Keywords
- zero-copy, energy efficiency, heterogeneous accelerators
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.