2019 Proceeding- General Chairs:
- Bogdan Nicolae,
- Alexandru Costan,
- Dmitry Duplyakin,
- Program Chairs:
- Ana Gainaru,
- Alexandru Uta,
- Carlos Costa
- Sponsor:
- sigarch
It is our pleasure to welcome you to the ScienceCloud'19: the 10th Workshop on Scientific Cloud Computing. ScienceCloud continues to provide the scientific community with the premier forum for discussing new research, development, and deployment efforts in hosting scientific computing workloads on cloud computing infrastructures. The focus of the workshop is on the use of cloudbased technologies to meet the new scientific challenges of converging HPC, big data and machine learning workloads that are currently not being well served by traditional supercomputing and datacenter technologies. ScienceCloud provides a unique opportunity for interaction and crosspollination between researchers and practitioners developing applications, algorithms, software, hardware and networking, emphasizing scientific computing for such cloud platforms.
The call for papers attracted submissions from across the world. The program committee reviewed and accepted 3 paper submissions: two full papers and one short paper, all of which received accept or weak accept scores. We will also include a keynote given by a leading expert on cloud computing, which will open our workshop and highlight emerging trends and challenges on modern cloud.
In this year's edition, ScienceCloud hosts a guest session on Converged Computing Infrastructure (CCIW). This session makes the case for deep convergence between big data, HPC and AI/deep learning/machine learning. It aims to define common software and hardware ecosystems that enable: (1) efficient operation of software and hardware, which is currently available only to the HPC community, for a wider audience, including the big data and AI/DL/ML communities, (2) sustainable development of new solutions, addressing the human-resource scarcities affecting both communities, and (3) innovation, especially for more efficient software-and-hardware solutions. The program committee reviewed and accepted one short paper for this session that focuses on the parallelization aspects of deep learning in particular.
Proceeding Downloads
Horizontal or Vertical?: A Hybrid Approach to Large-Scale Distributed Machine Learning
Data parallelism and model parallelism are two typical parallel modes for distributed machine learning (DML). Traditionally, DML mainly leverages data parallelism, which maintains one model instance for each node and synchronizes the model parameters at ...
ElasticPipe: An Efficient and Dynamic Model-Parallel Solution to DNN Training
Traditional deep neural network (DNN) training is executed with data parallelism, which suffers from significant communication overheads and GPU memory consumption. Considering this, recent pioneering works have attempted to train DNN with model ...
Towards a Smart, Internet-Scale Cache Service for Data Intensive Scientific Applications
Data and services provided by shared facilities, such as large-scale observing facilities, have become important enablers of scientific insights and discoveries across many science and engineering disciplines. Ensuring satisfactory quality of service ...
Deconstructing the 2017 Changes to AWS Spot Market Pricing
The Amazon Web Services spot market sells excess computing capacity at a reduced price and with reduced reliability guarantees. The low cost nature of the spot market has led to widespread adoption in industry and science. However, one of the challenges ...
Index Terms
- Proceedings of the 10th Workshop on Scientific Cloud Computing
Recommendations
Acceptance Rates
| Year | Submitted | Accepted | Rate |
|---|---|---|---|
| ScienceCloud '19 | 106 | 22 | 21% |
| ScienceCloud '16 | 8 | 4 | 50% |
| ScienceCloud '15 | 6 | 3 | 50% |
| ScienceCloud '14 | 17 | 8 | 47% |
| Science Cloud '13 | 14 | 7 | 50% |
| Overall | 151 | 44 | 29% |