CN102799708B - Graphic processing unit (GPU) high-performance calculation platform device applied to electromagnetic simulation - Google Patents
Graphic processing unit (GPU) high-performance calculation platform device applied to electromagnetic simulation Download PDFInfo
- Publication number
- CN102799708B CN102799708B CN201210200692.5A CN201210200692A CN102799708B CN 102799708 B CN102799708 B CN 102799708B CN 201210200692 A CN201210200692 A CN 201210200692A CN 102799708 B CN102799708 B CN 102799708B
- Authority
- CN
- China
- Prior art keywords
- gpu
- infiniband
- gigabit ethernet
- performance calculation
- line
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Abstract
The invention provides a graphic processing unit (GPU) high-performance calculation platform device applied to electromagnetic simulation. The GPU high-performance calculation platform device comprises a GPU calculator node and a parallel storage system, wherein a GPU calculator part consists of 58 1U GPU servers; a GPU calculator server is connected with a management node through three data lines; the three data lines of the 1U GPU server comprise an intelligent platform management interface (IPMI) remote management line, a 40Gb Infiniband calculation network line and a gigabit Ethernet line; the IPMI remote management line and the gigabit Ethernet line are connected with a gigabit Ethernet switchboard; the gigabit Ethernet switchboard is connected with the parallel storage system through the gigabit Ethernet line; a 10Gb Infiniband storage network line is connected with a 10Gb Infiniband switchboard, and another 10Gb Infiniband storage network line is connected with a 40Gb Infiniband switchboard; and the parallel storage system is connected with the 40Gb Infiniband switchboard through the gigabit Ethernet line. By the floating point calculation capacity of a GPU high-performance calculation platform, the electromagnetic calculation performance is improved, so that the best research effect can be achieved within the shortest time, and the production efficiency is improved.
Description
Technical field
The present invention relates to a kind of GPU high-performance calculation stage apparatus being applied to Electromagnetic Simulation.
Background technology
Large for computational resource quantity, kind is many, user is especially from the complex environment of multiple department and project, system can not realize effectively administering and maintaining simply by system manager artificially, and the open IT integrated application service platform building a unification is required.
In existing Electromagnetic Simulation, CST STUDIO SUITE is a simulation software towards the design of 3D electromagnetism, circuit, temperature and structural stress, the Numerical Simulation Analysis of complete system-level and component-level is provided, typical apply comprises that all kinds of Antenna/RCS, EMI/EMS, SI/PI, charged particle and electromagnetic field interact, field road, all kinds of collaborative simulation such as electromagnetism temperature and temperature deformation;
COMSOL Multiphysics is a multiple physical field coupling analysis software, is widely used in scientific research and the engineering calculation of every field, the various physical processes of simulation science and engineering field.
Summary of the invention
The object of the invention is the deficiency overcoming prior art existence, a kind of GPU high-performance calculation stage apparatus being applied to Electromagnetic Simulation is provided.
Object of the present invention is achieved through the following technical solutions:
Be applied to the GPU high-performance calculation stage apparatus of Electromagnetic Simulation, it is characterized in that: comprise GPU computer server and parallel memory system, described GPU computer node forms by 58 1U GPU servers are parallel, described 1U GPU server parallel join has the separate data circuit in three tunnels and the separate data circuit in three tunnels is connected with management node device, three circuit-switched data circuits of described 1U GPU server are IPMI telemanagement circuit, 40Gb Infiniband calculates netting twine road and gigabit ethernet line, described IPMI telemanagement circuit is connected with gigabit ethernet switch, described gigabit ethernet line is connected with gigabit ethernet switch, described gigabit ethernet switch is connected with parallel memory system by gigabit ethernet line, described parallel memory system is provided with the separate 10Gb Infiniband storage network route road of two-way, wherein a 10Gb Infiniband storage network route road, road is connected with 10Gb Infiniband switch, 10Gb Infiniband storage network route road, another road is connected with 40Gb Infiniband switch, described parallel memory system is connected with 40Gb Infiniband switch by gigabit ethernet line, described 40Gb Infiniband switch calculates netting twine road with 40Gb Infiniband and is connected.
Further, the above-mentioned GPU high-performance calculation stage apparatus being applied to Electromagnetic Simulation, wherein, management node device is 2U two-way server.
Further, the above-mentioned GPU high-performance calculation stage apparatus being applied to Electromagnetic Simulation, wherein, described GPU computer node and management node device are arranged in rack.
Again further, the above-mentioned GPU high-performance calculation stage apparatus being applied to Electromagnetic Simulation, wherein, parallel memory system is made up of data directory controller MDC, data access module IOM and smart storage devices SSU.
Substantive distinguishing features and the progress of technical solution of the present invention are mainly reflected in:
The present invention promotes the performance of the calculating of electromagnetism aspect by GPU high-performance calculation platform, reaches the highest scientific payoffs to reach in the shortest time, thus improves IT investment repayment, enhances productivity.
Accompanying drawing explanation
Below in conjunction with accompanying drawing, technical solution of the present invention is described further:
Fig. 1: structural representation of the present invention.
Embodiment
As shown in Figure 1, be applied to the GPU high-performance calculation stage apparatus of Electromagnetic Simulation, comprise GPU computer node 1 and parallel memory system 5, GPU computer node 1 forms by 58 1U GPU servers 11 are parallel, 1U GPU server 11 parallel join has the separate data circuit in three tunnels and the separate data circuit in three tunnels is connected with management node device 2, three circuit-switched data circuits of 1UGPU server 11 are IPMI telemanagement circuit 7, 40Gb Infiniband calculates netting twine road 9 and gigabit ethernet line 8, IPMI telemanagement circuit 7 is connected with gigabit ethernet switch 3, gigabit ethernet line 8 is connected with gigabit ethernet switch 3, gigabit ethernet switch 3 is connected with parallel memory system 5 by gigabit ethernet line 8, parallel memory system 5 is provided with the separate 10Gb Infiniband storage network route road 10 of two-way, wherein a 10Gb Infiniband storage network route road, road 10 is connected with 10Gb Infiniband switch 6, 10Gb Infiniband storage network route road, another road 10 is connected with 40Gb Infiniband switch 4, parallel memory system 5 is connected with 40Gb Infiniband switch 4 by gigabit ethernet line 8, 40Gb Infiniband switch 4 and 40Gb Infiniband calculate netting twine road 9 and are connected.
Management node device 2 adopts 2U two-way server, can increase security, and can also improve performance.
GPU computer node 1 and management node device 2 are arranged in rack, and rack wants to meet a whole machine cabinet machine good heat radiation, also want to have good space and winding displacement pipeline to be used for installing winding displacement to the machine in rack simultaneously and accomplish attractive in appearance and convenient management.Interior of equipment cabinet is also wanted to provide good electric power system, conveniently powers to machine.
GPU computing node 1 adopts 1U GPU server 11, and install 1 Nvidia Tesla C2075 in its 1UGPU server 11, Nvidia Tesla C2075 not only can provide larger video memory, and the performance of opening after ECC more improves nearly 20%.
Computational grid adopts 40Gb hypervelocity Infiniband for calculating data communication network, and adopt the application of 1Gb Ethernet load general network, two kinds of networks backup each other, and reach total system high availability simultaneously.
Parallel memory system 5 can easy expansion more than PB and more than the unification of 10TB/s handling capacity storage solution extending transversely.Parallel memory system 5 can also provide sufficient data redundancy to protect and advanced AUTOMATIC ZONING system simultaneously; the performance of whole storage system is significantly promoted by a limited number of high speed SSD solid state hard disc; make its space utilization rate reach 5/6, far exceed 1/2 service efficiency of other mirroring scheme., there is not Single Point of Faliure source in full redundancy framework, front network, based on ten thousand mbit ethernets, can configure the redundancy of gigabit networking as 10,000,000,000 networks in addition.
Parallel memory system 5 can also correctly identify the high speed storing layer be made up of high speed SSD, and can by suitable data placement on this high-velocity bed.Importantly, this parallel memory system 5 can set multiple different strategy, both can arrange according to access frequent degree, can arrange again according to user or application type.
Parallel memory system 5 is made up of three parts, comprise data directory controller MDC, data access module IOM and smart storage devices SSU, wherein data directory controller is responsible for processing the data directory request in parallel file system, two number of units are according to indexing controller redundancy each other, ensure the high reliability of parallel file system, be connected into Infiniband computational grid, ensure the high access bandwidth with computing node, data access module is responsible for processing the request of data in parallel file system, and multiple stage data access module ensure that the high-performance of computing node request of data, and every platform data access module installs two optical fiber network interface cards, each other redundancies, is connected with rear end smart storage devices, 2 smart storage devices achieve the combination in high speed storing space and massive store space, actual disposition naked capacity 120TB SATA hard disc is as massive store space, 1. 6TBSSD hard disk is as high speed storing space, massive store space continues readwrite performance and is not less than 3GBps, the random IOPS in high speed storing space is not less than 30K, in the configuration that classification stores, by being combined with job scheduling system, realize 80% operation need only reading some small documents when Run Script and start-up routine from disk array and can use high speed storing space, 20% operation having the file access of G level size can use massive store space.
This device adopts two cover networks, and this computational grid front-end storage networks is multiplexing, the Infiniband express network of the 40G of use, and supervising the network adopts gigabit Ethernet, and gigabit Ethernet and 40Gb Infiniband backup each other; The network of rear end parallel memory system 5 is with ten thousand mbit ethernets for core, and gigabit Ethernet is as backup, and the 40Gb Infiniband of front-end storage networks is connected by the gateway of Mellanox with before ten thousand mbit ethernets.
In sum, the present invention promotes the performance of the calculating of electromagnetism aspect by GPU high-performance calculation platform, farthest shares to realize software and hardware to reach in the shortest time scientific payoffs reaching the highest, thus improves IT investment repayment, enhances productivity.
It is emphasized that: be only preferred embodiment of the present invention above, not any pro forma restriction is done to the present invention, every above embodiment is done according to technical spirit of the present invention any simple modification, equivalent variations and modification, all still belong in the scope of technical solution of the present invention.
Claims (1)
1. be applied to the GPU high-performance calculation stage apparatus of Electromagnetic Simulation, it is characterized in that: comprise GPU computer node and parallel memory system, described GPU computer node forms by 58 1U GPU servers are parallel, described 1U GPU server parallel join has the separate data circuit in three tunnels and the separate data circuit in three tunnels is connected with management node device, three circuit-switched data circuits of described 1U GPU server are IPMI telemanagement circuit, 40Gb Infiniband calculates netting twine road and gigabit ethernet line, described IPMI telemanagement circuit is connected with gigabit ethernet switch, described gigabit ethernet line is connected with gigabit ethernet switch, described gigabit ethernet switch is connected with parallel memory system by gigabit ethernet line, described parallel memory system is provided with the separate 10Gb Infiniband storage network route road of two-way, wherein a 10Gb Infiniband storage network route road, road is connected with 10Gb Infiniband switch, 10Gb Infiniband storage network route road, another road is connected with 40Gb Infiniband switch, described parallel memory system is connected with 40Gb Infiniband switch by gigabit ethernet line, described 40Gb Infiniband switch calculates netting twine road with 40Gb Infiniband and is connected.
2. the GPU high-performance calculation stage apparatus being applied to Electromagnetic Simulation according to claim 1, is characterized in that: described management node device is 2U two-way server.
3. the GPU high-performance calculation stage apparatus being applied to Electromagnetic Simulation according to claim 1, is characterized in that: described GPU computer node and management node device are arranged in rack.
4. the GPU high-performance calculation stage apparatus being applied to Electromagnetic Simulation according to claim 1, is characterized in that: described parallel memory system is made up of data directory controller MDC, data access module IOM and smart storage devices SSU.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210200692.5A CN102799708B (en) | 2012-06-18 | 2012-06-18 | Graphic processing unit (GPU) high-performance calculation platform device applied to electromagnetic simulation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210200692.5A CN102799708B (en) | 2012-06-18 | 2012-06-18 | Graphic processing unit (GPU) high-performance calculation platform device applied to electromagnetic simulation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102799708A CN102799708A (en) | 2012-11-28 |
| CN102799708B true CN102799708B (en) | 2015-02-11 |
Family
ID=47198818
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210200692.5A Active CN102799708B (en) | 2012-06-18 | 2012-06-18 | Graphic processing unit (GPU) high-performance calculation platform device applied to electromagnetic simulation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102799708B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2015081515A1 (en) | 2013-12-04 | 2015-06-11 | 华为技术有限公司 | Data processing method, device, storage controller and equipment cabinet |
| CN108037995A (en) * | 2017-11-22 | 2018-05-15 | 西南电子技术研究所（中国电子科技集团公司第十研究所） | Distributed electromagnetic situation simulation computing system based on GPU |
| CN109447876A (en) * | 2018-10-16 | 2019-03-08 | 湖北三峡云计算中心有限责任公司 | A kind of burgher card system |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN201336145Y (en) * | 2008-12-17 | 2009-10-28 | 重庆神州龙芯科技有限公司 | Embedded multi-hard-disk network storage system |
| CN102185707A (en) * | 2011-04-15 | 2011-09-14 | 浪潮电子信息产业股份有限公司 | Method for expanding IPMI (Intelligent Platform Management Interface) protocol of server management |
| CN202720644U (en) * | 2012-06-18 | 2013-02-06 | 苏州超集信息科技有限公司 | Graphic processing unit (GPU) high-performance computing platform device applied to electromagnetic simulation |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20040107244A1 (en) * | 2002-12-02 | 2004-06-03 | Hung-Chi Kuo | Scalable and intelligent network platform for distributed system |
-
2012
- 2012-06-18 CN CN201210200692.5A patent/CN102799708B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN201336145Y (en) * | 2008-12-17 | 2009-10-28 | 重庆神州龙芯科技有限公司 | Embedded multi-hard-disk network storage system |
| CN102185707A (en) * | 2011-04-15 | 2011-09-14 | 浪潮电子信息产业股份有限公司 | Method for expanding IPMI (Intelligent Platform Management Interface) protocol of server management |
| CN202720644U (en) * | 2012-06-18 | 2013-02-06 | 苏州超集信息科技有限公司 | Graphic processing unit (GPU) high-performance computing platform device applied to electromagnetic simulation |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102799708A (en) | 2012-11-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Benner | Optical interconnect opportunities in supercomputers and high end computing | |
| CN102594602B (en) | A kind of location management design method of multi-node cloud computing server | |
| US9461768B2 (en) | Terabit top-of-rack switch | |
| CN102169448B (en) | Deployment method of cluster parallel computing environment | |
| CN102546813B (en) | A kind of High-Performance Computing Cluster computing system based on x86 PC framework | |
| Abali et al. | Disaggregated and optically interconnected memory: when will it be cost effective? | |
| JP3157935U (en) | server | |
| CN104135514B (en) | Fusion type virtual storage system | |
| CN102799708B (en) | Graphic processing unit (GPU) high-performance calculation platform device applied to electromagnetic simulation | |
| CN106686099A (en) | Method of realizing active-active mode across machine rooms of OracleRAC database based on infiniband network | |
| CN105099776A (en) | Cloud server management system | |
| CN102404201B (en) | Method of realizing maximum bandwidth of Lustre concurrent file system | |
| CN107480094A (en) | A kind of pond server system architecture of fusion architecture | |
| CN206532220U (en) | A kind of I/O expansion case system | |
| CN107220195A (en) | A kind of multichannel computer system architecture for supporting high density NVMe to store | |
| US9940280B1 (en) | Provisioning an enclosure with PCIe connectivity and storage devices | |
| Vahdat | Delivering scale out data center networking with optics—Why and how | |
| CN102937997A (en) | Data processing system | |
| CN104580527A (en) | Multiple-I/O high-density multi-node server system designing method oriented to cloud server application | |
| CN108090011A (en) | A kind of SAS Switch controllers extension framework and design method | |
| CN106919533B (en) | 4U high-density storage type server | |
| US11055252B1 (en) | Modular hardware acceleration device | |
| CN202720644U (en) | Graphic processing unit (GPU) high-performance computing platform device applied to electromagnetic simulation | |
| CN107491408B (en) | Computing server node | |
| CN104932837A (en) | Storage pool framework |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |