Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userBeeSwarm: Enabling Scalability Tests in Continuous Integration
102
0
0.0
(
0
)
Added by
Jieyang Chen
Publication date
2020
fields
Informatics Engineering
and research's language is
English
Ask ChatGPT about the research
No Arabic abstract
Testing is one of the most important steps in software development. It ensures the quality of software. Continuous Integration (CI) is a widely used testing system that can report software quality to the developer in a timely manner during the development progress. Performance, especially scalability, is another key factor for High Performance Computing (HPC) applications. Though there are many applications and tools to profile the performance of HPC applications, none of them are integrated into the continuous integration. On the other hand, no current continuous integration tools provide easy-to-use scalability test capabilities. In this work, we propose BeeSwarm, a scalability test system that can be easily applied to the current CI test environment enabling scalability test capability for HPC developers. As a showcase, BeeSwarm is integrated into Travis CI and GitLab CI to execute the scalability test workflow on Chameleon cloud.
rate research
Read More
As part of the Exascale Computing Project (ECP), a recent focus of development efforts for the SUite of Nonlinear and DIfferential/ALgebraic equation Solvers (SUNDIALS) has been to enable GPU-accelerated time integration in scientific applications at extreme scales. This effort has resulted in several new GPU-enabled implementations of core SUNDIALS data structures, support for programming paradigms which are aware of the heterogeneous architectures, and the introduction of utilities to provide new points of flexibility. In this paper, we discuss our considerations, both internal and external, when designing these new features and present the features themselves. We also present performance results for several of the features on the Summit supercomputer and early access hardware for the Frontier supercomputer, which demonstrate negligible performance overhead resulting from the additional infrastructure and significant speedups when using both NVIDIA and AMD GPUs.
The blockchain paradigm provides a mechanism for content dissemination and distributed consensus on Peer-to-Peer (P2P) networks. While this paradigm has been widely adopted in industry, it has not been carefully analyzed in terms of its network scaling with respect to the number of peers. Applications for blockchain systems, such as cryptocurrencies and IoT, require this form of network scaling. In this paper, we propose a new stochastic network model for a blockchain system. We identify a structural property called emph{one-endedness}, which we show to be desirable in any blockchain system as it is directly related to distributed consensus among the peers. We show that the stochastic stability of the network is sufficient for the one-endedness of a blockchain. We further establish that our model belongs to a class of network models, called monotone separable models. This allows us to establish upper and lower bounds on the stability region. The bounds on stability depend on the connectivity of the P2P network through its conductance and allow us to analyze the scalability of blockchain systems on large P2P networks. We verify our theoretical insights using both synthetic data and real data from the Bitcoin network.
To support the variety of Big Data use cases, many Big Data related systems expose a large number of user-specifiable configuration parameters. Highlighted in our experiments, a MySQL deployment with well-tuned configuration parameters achieves a peak throughput as 12 times much as one with the default setting. However, finding the best setting for the tens or hundreds of configuration parameters is mission impossible for ordinary users. Worse still, many Big Data applications require the support of multiple systems co-deployed in the same cluster. As these co-deployed systems can interact to affect the overall performance, they must be tuned together. Automatic configuration tuning with scalability guarantees (ACTS) is in need to help system users. Solutions to ACTS must scale to various systems, workloads, deployments, parameters and resource limits. Proposing and implementing an ACTS solution, we demonstrate that ACTS can benefit users not only in improving system performance and resource utilization, but also in saving costs and enabling fairer benchmarking.
The variational quantum Monte Carlo (VQMC) method received significant attention in the recent past because of its ability to overcome the curse of dimensionality inherent in many-body quantum systems. Close parallels exist between VQMC and the emerging hybrid quantum-classical computational paradigm of variational quantum algorithms. VQMC overcomes the curse of dimensionality by performing alternating steps of Monte Carlo sampling from a parametrized quantum state followed by gradient-based optimization. While VQMC has been applied to solve high-dimensional problems, it is known to be difficult to parallelize, primarily owing to the Markov Chain Monte Carlo (MCMC) sampling step. In this work, we explore the scalability of VQMC when autoregressive models, with exact sampling, are used in place of MCMC. This approach can exploit distributed-memory, shared-memory and/or GPU parallelism in the sampling task without any bottlenecks. In particular, we demonstrate the GPU-scalability of VQMC for solving up to ten-thousand dimensional combinatorial optimization problems.
Distributed optimization for solving non-convex Optimal Power Flow (OPF) problems in power systems has attracted tremendous attention in the last decade. Most studies are based on the geographical decomposition of IEEE test systems for verifying the feasibility of the proposed approaches. However, it is not clear if one can extrapolate from these studies that those approaches can be applied to very large-scale real-world systems. In this paper, we show, for the first time, that distributed optimization can be effectively applied to a large-scale real transmission network, namely, the Polish 2383-bus system for which no pre-defined partitions exist, by using a recently developed partitioning technique. More specifically, the problem solved is the AC OPF problem with geographical decomposition of the network using the Alternating Direction Method of Multipliers (ADMM) method in conjunction with the partitioning technique. Through extensive experimental results and analytical studies, we show that with the presented partitioning technique the convergence performance of ADMM can be improved substantially, which enables the application of distributed approaches on very large-scale systems.
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
