Clones defective in metabolic cooperation selected from a pluripotent feeder-dependent mouse embryonal carcinoma cell line.

T. Smith,K. Box,M. Hooper
DOI: https://doi.org/10.1016/0014-4827(86)90208-9
IF: 4.145
1986-11-01
Experimental Cell Research
Abstract:
What problem does this paper attempt to address?
The problem this paper attempts to address is the development of a parallel Tree-SPH code for simulating the formation and evolution of galaxies. Specifically, the authors aim to utilize parallel computing technology to improve the efficiency and accuracy of large-scale astrophysical simulations. ### Background and Problem 1. **Complexity of Galaxy Formation and Evolution**: - Galaxies and galaxy clusters are believed to form due to the gravitational instability of density fluctuations in the early universe. - These fluctuations grow linearly in the early stages, while numerical simulations are required to study them in the nonlinear or quasi-linear stages. - The standard approach is to consider the matter density distribution in different cosmological models and use numerical simulations to study the structure formation process. 2. **Limitations of Existing Methods**: - Semi-analytical models, although successful in reproducing many galaxy properties, are based on many assumptions and cannot accurately describe complex processes such as gas dynamics and star formation. - Pure grid codes (e.g., PM-PPM) have low resolution in high-density regions. - Particle methods (e.g., PP) have computation times that increase with the square of the number of particles, making it difficult to handle large-scale simulations. 3. **Need for Parallel Computing**: - Large-scale parallel computers (e.g., Cray T3E) can significantly enhance computational capabilities, but porting existing codes to these machines presents many challenges. - Effective data distribution and load balancing strategies are needed to ensure all processors work efficiently. ### Objectives of the Paper - **Develop Parallel Tree-SPH Code**: Combine the Smoothed Particle Hydrodynamics (SPH) method with the Barnes-Hut tree algorithm to achieve efficient parallel computing. - **Optimize Parallel Performance**: Ensure efficient operation of the code on multiprocessor systems through reasonable data distribution and load balancing strategies. - **Validate Code Effectiveness**: Verify the accuracy and scalability of the code through benchmark tests and actual simulations (e.g., formation of X-ray galaxy clusters). ### Main Contributions - **Parallelization Strategy**: Use the SHMEM library to handle communication, implementing an asynchronous communication scheme to reduce communication overhead. - **Data Distribution and Load Balancing**: Employ Recursive Orthogonal Bisection (ROB) for data distribution, ensuring that each processor handles approximately the same amount of work. - **Practical Application**: Demonstrate simulations of galaxy formation and evolution using the code, including physical processes such as radiative cooling, star formation, and supernova feedback. ### Conclusion The paper successfully develops a parallel Tree-SPH code capable of running efficiently on large-scale parallel computers, providing a powerful tool for the study of galaxy formation and evolution.