Nikoli Dryden
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Nikoli Dryden has a total of 23 co-authors. Here are the top co-authors by number of shared publications:
- Hariharan Devarajan
Lawrence Livermore National Laboratory, Livermore, CA, USA
3 shared publications - Chen Zhuang
Institute of Science Tokyo, Tokyo, Japan and Riken Center for Computational Science, Kobe, Japan
2 shared publications - Lingqi Zhang
RIKEN Center for Computational Science, Kobe, Japan
2 shared publications - Mohamed Wahib
RIKEN Center for Computational Science, Kobe, Japan
2 shared publications - Peng Chen
RIKEN Center for Computational Science, Kobe, Japan
2 shared publications - Ray A. O. Sinurat
University of Chicago
2 shared publications
Latest Publications
- Open Access
- 11 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-12
WADO: A Distributed WORM Storage Service for Asynchronous Data Operations
AI-driven scientific workloads increasingly depend on data-intensive input pipelines, where deep learning frameworks must ingest and transform large datasets from hierarchical HPC storage. Existing system-centric data services improve movement and locality between the parallel file system (PFS), node-local storage, and memory. However, they do not directly optimize how input pipeline operations execute across scopes, stage overlap, and resource-specific parallelism. As scale grows, this gap causes worker stalls, contention, and poor hardware utilization. We present WADO, a distributed write-once-read-many (WORM) object-store runtime for data-centric workloads that closes this gap through three coordinated mechanisms: scope-centric processing, explicit pipeline decomposition, and interference-aware explicit parallelism. WADO dynamically maps operations to execution scopes, overlaps stages such as I/O, communication, and transformations, and applies contention-aware concurrency control to match hardware behavior at runtime. Our evaluation shows three main findings: (1) scope-centric processing preserves throughput under scale, improving mixed-operation throughput by up to 1.65 × ; (2) explicit pipeline decomposition converts serialized wait into overlapped progress, delivering up to 2.16 × higher sustained bandwidth; and (3) interference-aware explicit parallelism improves effective bandwidth by up to 4.4 × by avoiding oversubscription collapse. On Unet3D model training, these mechanisms translate to end-to-end gains, improving data loading performance by 4.1 × compared to baseline PyTorch on Lustre, and 1.51 × compared to DYAD, enabled by deeper pipelining, adaptive parallelism, and near-data transformation offloading.
- 11 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-12
Open AccessWADO: A Distributed WORM Storage Service for Asynchronous Data Operations
AI-driven scientific workloads increasingly depend on data-intensive input pipelines, where deep learning frameworks must ingest and transform large datasets from hierarchical HPC storage. Existing system-centric data services improve movement and locality between the parallel file system (PFS), node-local storage, and memory. However, they do not directly optimize how input pipeline operations execute across scopes, stage overlap, and resource-specific parallelism. As scale grows, this gap causes worker stalls, contention, and poor hardware utilization. We present WADO, a distributed write-once-read-many (WORM) object-store runtime for data-centric workloads that closes this gap through three coordinated mechanisms: scope-centric processing, explicit pipeline decomposition, and interference-aware explicit parallelism. WADO dynamically maps operations to execution scopes, overlaps stages such as I/O, communication, and transformations, and applies contention-aware concurrency control to match hardware behavior at runtime. Our evaluation shows three main findings: (1) scope-centric processing preserves throughput under scale, improving mixed-operation throughput by up to 1.65 × ; (2) explicit pipeline decomposition converts serialized wait into overlapped progress, delivering up to 2.16 × higher sustained bandwidth; and (3) interference-aware explicit parallelism improves effective bandwidth by up to 4.4 × by avoiding oversubscription collapse. On Unet3D model training, these mechanisms translate to end-to-end gains, improving data loading performance by 4.1 × compared to baseline PyTorch on Lustre, and 1.51 × compared to DYAD, enabled by deeper pipelining, adaptive parallelism, and near-data transformation offloading.
- Open Access
- 11 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-12
HORATIO: Bridging Management and Analysis of Traces at Scale
Modern scientific and deep learning workloads on HPC systems rely on profiling and tracing across multiple software and hardware layers, generating diagnostic traces that often reach terabyte scale. Existing approaches manage these traces along two axes: trace formats and analysis tooling. Practitioners often convert raw traces into queryable formats, but doing so nearly doubles storage when raw files are retained for compatibility and requires upfront schema discovery that profiling tools cannot guarantee. A cleaner path is to make raw traces efficient in place, but this requires overcoming three limitations: lack of selective querying, analysis throughput bottlenecks, and lack of physical clustering. To address these limitations jointly, we developed Horatio, a raw trace management framework that indexes, analyzes, and physically clusters raw traces directly. Three findings emerge from our work. First, Horatio stores a lightweight RocksDB-backed auxiliary index alongside the raw trace, including gzip checkpoints, per-chunk bloom filters, and chunk-level statistics, delivering selective queries up to 75 × faster than naive Parquet at ∼ 1.01 × raw storage, with the highest cross-query mean throughput (232 M events/s) among state-of-the-art formats. Second, offloading event-level computation to a native C++ backend while keeping Dask for orchestration yields 80–83 × speedup over the original Dask-based DFAnalyzer. Third, lossless trace clustering that preserves the same input format yields a further 1.8–5.5 × end-to-end speedup across four AI and scientific workloads, and up to 230 × on h5bench where the preset aligns tightly with the cluster boundary, all with original layouts reconstructible on demand. Across five AI and scientific workloads, Horatio completes pipelines that DFAnalyzer cannot finish within 8 hours. On a 2.2 TB uncompressed trace, Horatio’s MPI-based mode scales to 16 × at 32 nodes on the full event set and its Dask-based path peaks at 4 × on a preset-filtered workload, both completing where DFAnalyzer hits OOM at every scale.
- 11 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-12
Open AccessHORATIO: Bridging Management and Analysis of Traces at Scale
Modern scientific and deep learning workloads on HPC systems rely on profiling and tracing across multiple software and hardware layers, generating diagnostic traces that often reach terabyte scale. Existing approaches manage these traces along two axes: trace formats and analysis tooling. Practitioners often convert raw traces into queryable formats, but doing so nearly doubles storage when raw files are retained for compatibility and requires upfront schema discovery that profiling tools cannot guarantee. A cleaner path is to make raw traces efficient in place, but this requires overcoming three limitations: lack of selective querying, analysis throughput bottlenecks, and lack of physical clustering. To address these limitations jointly, we developed Horatio, a raw trace management framework that indexes, analyzes, and physically clusters raw traces directly. Three findings emerge from our work. First, Horatio stores a lightweight RocksDB-backed auxiliary index alongside the raw trace, including gzip checkpoints, per-chunk bloom filters, and chunk-level statistics, delivering selective queries up to 75 × faster than naive Parquet at ∼ 1.01 × raw storage, with the highest cross-query mean throughput (232 M events/s) among state-of-the-art formats. Second, offloading event-level computation to a native C++ backend while keeping Dask for orchestration yields 80–83 × speedup over the original Dask-based DFAnalyzer. Third, lossless trace clustering that preserves the same input format yields a further 1.8–5.5 × end-to-end speedup across four AI and scientific workloads, and up to 230 × on h5bench where the preset aligns tightly with the cluster boundary, all with original layouts reconstructible on demand. Across five AI and scientific workloads, Horatio completes pipelines that DFAnalyzer cannot finish within 8 hours. On a 2.2 TB uncompressed trace, Horatio’s MPI-based mode scales to 16 × at 32 nodes on the full event set and its Dask-based path peaks at 4 × on a preset-filtered workload, both completing where DFAnalyzer hits OOM at every scale.
- Open Access
- 13 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 807-816
CoDL: A Framework for Studying Cross-Component Interference in Deep Learning Training Pipelines
- 13 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 807-816
Open AccessCoDL: A Framework for Studying Cross-Component Interference in Deep Learning Training Pipelines
- Open Access
- 27 June 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Architecture and Code Optimization
- Vol. 23 (2) , 1-25
STen: Productive and Efficient Sparsity in PyTorch
As deep learning models grow, sparsity is becoming an increasingly critical component of deep neural networks, enabling improved performance and reduced storage. However, existing frameworks offer poor support for sparsity. Specialized sparsity engines focus exclusively on sparse inference, while general frameworks primarily focus on sparse tensors in classical formats and neglect the broader sparsification pipeline necessary for using sparse models, especially during training. Further, existing frameworks are not easily extensible: adding a new sparse tensor format or operator is challenging and time-consuming. To address this, we propose STen, a sparsity programming model and interface for PyTorch whose key design insight is the decoupling of sparsity layouts, operators, and sparsifiers into composable, first-class abstractions that users can independently define and combine. An automatic dispatch mechanism selects the best available sparse implementation and transparently falls back to dense execution, allowing STen to support virtually all sparsification methods while enabling rapid prototyping without sacrificing performance for optimized paths. We demonstrate the versatility of STen by expressing existing sparsification techniques within its abstraction, achieving a code size reduction of over 2×. Finally, we develop a novel, high-performance grouped n : m sparsity layout for CPU inference at moderate sparsity, accelerating end-to-end BERT BASE inference by up to 3.2×. STen brings high performance and ease of use, making sparsity readily accessible for existing PyTorch models.
- 27 June 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Architecture and Code Optimization
- Vol. 23 (2) , 1-25
Open AccessSTen: Productive and Efficient Sparsity in PyTorch
As deep learning models grow, sparsity is becoming an increasingly critical component of deep neural networks, enabling improved performance and reduced storage. However, existing frameworks offer poor support for sparsity. Specialized sparsity engines focus exclusively on sparse inference, while general frameworks primarily focus on sparse tensors in classical formats and neglect the broader sparsification pipeline necessary for using sparse models, especially during training. Further, existing frameworks are not easily extensible: adding a new sparse tensor format or operator is challenging and time-consuming. To address this, we propose STen, a sparsity programming model and interface for PyTorch whose key design insight is the decoupling of sparsity layouts, operators, and sparsifiers into composable, first-class abstractions that users can independently define and combine. An automatic dispatch mechanism selects the best available sparse implementation and transparently falls back to dense execution, allowing STen to support virtually all sparsification methods while enabling rapid prototyping without sacrificing performance for optimized paths. We demonstrate the versatility of STen by expressing existing sparsification techniques within its abstraction, achieving a code size reduction of over 2×. Finally, we develop a novel, high-performance grouped n : m sparsity layout for CPU inference at moderate sparsity, accelerating end-to-end BERT BASE inference by up to 3.2×. STen brings high performance and ease of use, making sparsity readily accessible for existing PyTorch models.
- Open Access
- 8 June 2025
- Published by Association for Computing Machinery (ACM)
- p. 57-72
Scaling Large-scale GNN Training to Thousands of Processors on CPU-based Supercomputers
- 8 June 2025
- Published by Association for Computing Machinery (ACM)
- p. 57-72
Open AccessScaling Large-scale GNN Training to Thousands of Processors on CPU-based Supercomputers
- Open Access
- 28 February 2025
- Published by Association for Computing Machinery (ACM)
- p. 566-568
A General and Scalable GCN Training Framework on CPU Supercomputers
Graph Convolutional Networks (GCNs) are widely used in various domains. However, training distributed full-batch GCNs on large-scale graphs poses challenges due to inefficient memory access patterns and high communication overhead. This paper presents a general and efficient GCN training framework on CPU supercomputers. It comprises a general aggregation kernel designed to optimize irregular memory access and a quantization method with label propagation to reduce communication overhead. Experimental results show that our method achieves a speedup of up to 4.1× compared with the SoTA implementations.
- 28 February 2025
- Published by Association for Computing Machinery (ACM)
- p. 566-568
Open AccessA General and Scalable GCN Training Framework on CPU Supercomputers
Graph Convolutional Networks (GCNs) are widely used in various domains. However, training distributed full-batch GCNs on large-scale graphs poses challenges due to inefficient memory access patterns and high communication overhead. This paper presents a general and efficient GCN training framework on CPU supercomputers. It comprises a general aggregation kernel designed to optimize irregular memory access and a quantization method with label propagation to reduce communication overhead. Experimental results show that our method achieves a speedup of up to 4.1× compared with the SoTA implementations.