Improve CPU Utilization, Open HPC Efficient Computing
21st, demand for oil and natural gas supply increased dramatically, making oil companies to increase production and demand for new reservoirs to increase quickly. In order to improve efficiency and increase exploration and development efficiency, reduce risk, new emerging technology of digital geological have a rapid development, require higher-performance computer hardware and software. How to improve CPU utilization, open HPC efficient computing? Intel High Performance Computing Symposium 2010 give us a guide.
Oil industry users conversion to HPC-based digital modeling approach from the traditional exploration way, also liberate the user to an open x86 environment from the UNIX mainframe platform, thus further promot a comprehensive upgrade of the related software of the Geological Survey. Four development trend of High-performance computing 1, x86 high-performance computer clusters gradually occupy the dominant position. Since 2000, all previous top 500 ranking statistics, of which 82% of the system uses the x86 cluster (as of November 2009), has become a mainstream high-performance computers.
2, the global high-tech computer system CPU scale growing. Multi-CPU multi-core computing with powerful computing performance makes high-performance computing centers develop to large scale multi-core systems.
3, high-performance computer applications in more fields. In 2008 after the financial crisis, the number of high-performance computers decrease in financial sector, and super-computer used in the field of scientific research continues to increase, information processing and geophysical also on the top.
4, the cluster of x86 servers is becoming the mainstream application platform of oil and gas exploration data processing . It is understood that, in the Top500 list of the world (November 2009), the number of ultra-count used in the geophysical field came in after R & D, Finance, 2005 -2009, respectively, 9.4%, 4.6%, 8.6%, 9.8 %, 6.2% and the number reached 31 units (2009).
Three aspects should be noticed on improving low CPU utilization, storage performance and data security, high energy and high refrigeration in geophysical high performance computing. Bottlenecks of storage system I / O. Using a parallel shared file system to improve data transmission speed, and progress through the RAID array. Develop energy saving software to solve the high energy consumption problem.