CN108460719A - A kind of graphics workstation system based on Domestic Platform - Google Patents
A kind of graphics workstation system based on Domestic Platform Download PDFInfo
- Publication number
- CN108460719A CN108460719A CN201810203607.8A CN201810203607A CN108460719A CN 108460719 A CN108460719 A CN 108460719A CN 201810203607 A CN201810203607 A CN 201810203607A CN 108460719 A CN108460719 A CN 108460719A
- Authority
- CN
- China
- Prior art keywords
- domestic
- graphics workstation
- system based
- cpu
- workstation system
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T1/00—General purpose image data processing
- G06T1/20—Processor architectures; Processor configuration, e.g. pipelining
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F15/00—Digital computers in general; Data processing equipment in general
- G06F15/76—Architectures of general purpose stored program computers
- G06F15/78—Architectures of general purpose stored program computers comprising a single central processing unit
- G06F15/7839—Architectures of general purpose stored program computers comprising a single central processing unit with memory
Abstract
The graphics workstation system based on Domestic Platform that the present invention relates to a kind of, including domestic CPU, domestic GPU, complex programmable logic device (CPLD), baseboard management controller, power module, gigabit Ethernet driver, SATA3.0 controllers, USB3.0 controllers, the domestic CPU is separately connected the domestic GPU, the complex programmable logic device (CPLD), the baseboard management controller, and the domestic GPU integrates the gigabit Ethernet driver, the SATA3.0 controllers and the USB3.0 controllers.
Description
Technical field
The graphics workstation system based on Domestic Platform that the present invention relates to a kind of, belongs to computer workstation technical field.
Background technology
Information security is an important indicator in Future Information war, in the environment of network-enabled operation, Di Renke
The place for initiating information attack is ubiquitous, impossible to guard against.The graphics process of modern computer becomes more and more important, needs one
A special graphics processor undertakes display task, to adapt to aerospace navigation, satellite data processing, biological medicine research
Etc. demand of the every field to high-performance GPU.
Foreign countries are very deep for the research of high performance graphics work station, and ripe software and hardware product is also relatively more,
And the research of China in this respect is started late, the especially high performance graphics work station based entirely on domestic CPU and domestic GPU
Or blank out.
The comprehensive Combat Command System of a new generation is core commander's hinge of army, needs to complete the comprehensive collection to various operation elements
At with display, the requirement to presentation of information is high.Firstly, it is necessary to which the information of display is wide, destination number is more.Secondly, presentation of information needs
That wants is computationally intensive.System needs battlefield surroundings to be shown it can be seen from the characteristics of by integrating Combat Command System presentation of information above
It is extremely complex with situation information, and requirement of real-time is high, it is therefore necessary to there is autonomous controllable high performance graphics work of new generation
Station is supported.
Although can ensure the normal use of various functions on general information processing platform on autonomous graphics workstation at present,
But it is barely satisfactory in performance；Since increasing for number of concurrent leads to fluency not in the case of audio and video stream media multichannel program request
It is high；Refreshing frequency is relatively low when three-dimensional geographic information loaded and displayed image/elevation/vector data and general particle effect；It is multiple batches of
Refresh time is longer when creating two-dimentional moving-target military symbol；Situation synthesis system shows and overflows in static, the upper figure of dynamic object
There is response time length in trip scaling；Three-dimensional label is influenced when refreshing display by monokaryon CPU computing capabilitys and video card capabilities,
Refresh performance is very poor.
These problems all drastically influence user experience and service efficiency present on, it is therefore desirable to for existing autonomous
The problem of graphics workstation display performance is exposed carries out the mechanism study of display performance in a deep going way, is further promoted at present certainly
The display performance of main graphic work station.
Invention content
In view of the deficiencies of the prior art, the graphics workstation system based on Domestic Platform that the present invention provides a kind of；
The present invention provides a kind of high performance graphics work station frameworks based on the domestic GPU of domestic CPU+, have filled up and have been based on
The blank of Domestic Platform high performance graphics work station.
The technical scheme is that：
A kind of graphics workstation system based on Domestic Platform, including domestic CPU, domestic GPU, complicated programmable logic device
Part CPLD, baseboard management controller, power module, gigabit Ethernet driver, SATA3.0 controllers, USB3.0 controllers,
PCIE3.0 extended chips, the domestic CPU connections PCIE3.0 extended chips, the Complex Programmable Logic Devices
CPLD, baseboard management controller, power module, gigabit Ethernet driver, SATA3.0 controllers, USB3.0 controllers connect
Connect the PCIE3.0 extended chips.
According to currently preferred, the model FT2000+ of the domestic CPU.
CPU is that criterion carries out type selecting with " domestic, multinuclear, high-performance ".Currently, independently controllable high-performance CPU is main for mainstream
Have：Soar FT2000+ (64 core), Shen Wei SW1621 (16 core) and Godson 3B3000 (4 core).It is final to select by Comprehensive Correlation
With FT2000+ processors.FT2000+ has been internally integrated 64 64 bit processor cores, and typical operating frequency 2.2GHz is being weighed
In the Spec2006 tests of prestige, integral point 672, floating-point 585,5-6 times of the performance boost with FT1500A-16 compared with, it is sufficient to English spy
Your middle-and high-end server chip compares favourably.
FT2000+ is internally integrated 8 DDR4 Memory Controller Hub, and each controller supports four main memory access extensions, most
32 channels standard DDR4 extensions and the memory size of highest 2TB can be supported greatly.Consider mainboard and chassis space, simultaneously
To improve single Memory Controller Hub memory read-write performance, each Memory Controller Hub draws 1 dimm socket respectively on mainboard, and whole plate is total
Count 8 memory slots；
GPU carries out type selecting in line with the criterion of " domestic peak performance ".Current domestic autonomous GPU technologies are praised micro- the most prominent with scape
Go out, wherein JM7200 is domestic peak performance and supports double video card interconnections.According to currently preferred, the type of the domestic GPU
Number be JM7200.
According to currently preferred, the model of model 88SE9230, the USB3.0 controller of SATA3.0 controllers
uPD720210。
Other function components select the PCIE conversion chips being adapted in platform maturation of soaring, network-control respectively
Device selects I350 to control chip.In order to improve the manageability of work station, intelligent management system is devised on mainboard, is realized to work
Make the functions such as the outband management stood and remote on-off, voltage temperature monitoring and data access.Intelligent management control chip is selected
AST2400 managing chips.
According to currently preferred, the model PEX8796 of the PCIE3.0 extended chips, support in total 24Port,
The expansion interface of 96Lane；The uplink port of the PCIE3.0 extended chips is connected by PCIE3.0X16 and the domestic CPU
It connects, the downlink port of the PCIE3.0 extended chips expands USB3.0 interfaces, gigabit networking respectively by PCIE conversion chips
Interface and SATA3.0 interfaces；
According to currently preferred, the PCIE3.0 extended chips also expand two PCIE3.0X16 slots and 1
PCIE3.0X8 slots, PCIE3.0X8 slots are for connecting other PCIE interfaces, and two PCIE3.0X16 slots are for connecting
Two domestic video card JM7200 are met, and two video cards are directly interengageable for transmitting data by JJWLINK.
Video card based on the interconnection of domestic high-performance GPU double cards realizes that technology can be cooperateed with using two pieces of independent video cards, makes two
Block video card concurrent working respectively undertakes different graphic plotting tasks, to greatly improve the whole figure accelerating ability of platform.
Under this operating mode, two independent video card systems cooperate under the control of the same CPU.Double card interconnects display technology
At work, same frame picture can be divided into two drawing subtasks (upper layer driving journey automatically by CPU according to drawing complexity
Sequence can be divided according to task load dynamic), it distributes to two pieces of different video cards and carries out parallel processing.Monolithic video card renders in this way
Pel number relative reduction be artwork shape half, per frame render time accordingly shorten, with this reach raising draw render
The purpose of efficiency.
Beneficial effects of the present invention are：
1, the present invention has filled up the blank of the domestic GPU high performance graphics work stations of domestic CPU+, meets IT application facilities certainly
Main controllable demand；
2, with artificial intelligence, the development of virtual/augmented reality, GPU is no longer limited to 3D graphics process.In floating-point
Operation, parallel computation etc., GPU can provide the performance of decades of times or even up to a hundred times of CPU.The present invention is based on domestic flat
The high performance graphics work station of platform can be applied to artificial intelligence, virtual/augmented reality field, carry out the war based on military field
The technical research such as field multidimensional data analysis deduction, simulated training system, are conducive to deep accumulated completion bend and overtake other vehicles, further carry
Rise the level of IT application and fight capability of my army.
Description of the drawings
Fig. 1 is that the present invention is based on the Organization Charts of the graphics workstation system of Domestic Platform；
Fig. 2 is that the present invention is based on the principle schematics of the graphics workstation system of Domestic Platform；
Specific implementation mode
The present invention is further qualified with embodiment with reference to the accompanying drawings of the specification, but not limited to this.
Embodiment
A kind of graphics workstation system based on Domestic Platform, as shown in Figure 1 and Figure 2, including it is domestic CPU, domestic GPU, multiple
Miscellaneous programmable logic device (CPLD), baseboard management controller, power module, gigabit Ethernet driver, SATA3.0 controllers,
USB3.0 controllers, the domestic CPU connections PCIE3.0Switch, the PCIE3.0Switch connections domestic GPU, complexity can
Programmed logic device CPLD, baseboard management controller, gigabit Ethernet driver, SATA3.0 controllers and USB3.0 controls
Device processed.
The model FT2000+ of domestic CPU.
CPU is that criterion carries out type selecting with " domestic, multinuclear, high-performance ".Currently, independently controllable high-performance CPU is main for mainstream
Have：Soar FT2000+ (64 core), Shen Wei SW1621 (16 core) and Godson 3B3000 (4 core).It is final to select by Comprehensive Correlation
With FT2000+ processors.FT2000+ has been internally integrated 64 64 bit processor cores, and typical operating frequency 2.2GHz is being weighed
In the Spec2006 tests of prestige, integral point 672, floating-point 585,5-6 times of the performance boost with FT1500A-16 compared with, it is sufficient to English spy
Your middle-and high-end server chip compares favourably.
FT2000+ is internally integrated 8 DDR4 Memory Controller Hub, and each controller supports four main memory access extensions, most
32 channels standard DDR4 extensions and the memory size of highest 2TB can be supported greatly.Consider mainboard and chassis space, simultaneously
To improve single Memory Controller Hub memory read-write performance, each Memory Controller Hub draws 1 dimm socket respectively on mainboard, and whole plate is total
Count 8 memory slots；
GPU carries out type selecting in line with the criterion of " domestic peak performance ".Current domestic autonomous GPU technologies are praised micro- the most prominent with scape
Go out, wherein JM7200 is domestic peak performance and supports double video card interconnections, the model JM7200 of domestic GPU.
Other function components select the PCIE conversion chips being adapted in platform maturation of soaring respectively
PEX8796Switch chips, network controller selection I350 control chips, the model 88SE9230 of SATA3.0 controllers,
The model uPD720210 of USB3.0 controllers.Intelligent management control chip selects AST2400 managing chips.
In order to improve the manageability of work station, intelligent management system is devised on mainboard, is realized outside to the band of work station
The functions such as management and remote on-off, voltage temperature monitoring and data access.
Graphics workstation system further includes PCIE3.0 extended chips, the model PEX8796 of PCIE3.0 extended chips, always
The expansion interface of 24Port, 96Lane are supported altogether；The uplink port of PCIE3.0 extended chips by PCIE3.0X16 with it is domestic
The downlink port of CPU connections, PCIE3.0 extended chips expands USB3.0 interfaces, kilomega network respectively by PCIE conversion chips
Network interface and SATA3.0 interfaces；
PCIE3.0 extended chips also expand two PCIE3.0X16 slots and 1 PCIE3.0X8 slot, PCIE3.0X8
Slot is used to connect two domestic video card JM7200 for connecting other PCIE interfaces, two PCIE3.0X16 slots, and
And two video cards are directly interengageable for transmitting data by JJWLINK.
Video card based on the interconnection of domestic high-performance GPU double cards realizes that technology can be cooperateed with using two pieces of independent video cards, makes two
Block video card concurrent working respectively undertakes different graphic plotting tasks, to greatly improve the whole figure accelerating ability of platform.
Under this operating mode, two independent video card systems cooperate under the control of the same CPU.Double card interconnects display technology
At work, same frame picture can be divided into two drawing subtasks (upper layer driving journey automatically by CPU according to drawing complexity
Sequence can be divided according to task load dynamic), it distributes to two pieces of different video cards and carries out parallel processing.Monolithic video card renders in this way
Pel number relative reduction be artwork shape half, per frame render time accordingly shorten, with this reach raising draw render
The purpose of efficiency.
Claims (7)
1. a kind of graphics workstation system based on Domestic Platform, which is characterized in that can including domestic CPU, domestic GPU, complexity
Programmed logic device CPLD, baseboard management controller, power module, gigabit Ethernet driver, SATA3.0 controllers,
USB3.0 controllers, PCIE3.0 extended chips, the domestic CPU connections PCIE3.0 extended chips, the complexity can be compiled
Journey logic device (CPLD), baseboard management controller, power module, gigabit Ethernet driver, SATA3.0 controllers, USB3.0
Controller is all connected with the PCIE3.0 extended chips.
2. a kind of graphics workstation system based on Domestic Platform according to claim 1, which is characterized in that described domestic
The model FT2000+ of CPU.
3. a kind of graphics workstation system based on Domestic Platform according to claim 1, which is characterized in that described domestic
The model JM7200 of GPU.
4. a kind of graphics workstation system based on Domestic Platform according to claim 1, which is characterized in that SATA3.0
The model uPD720210 of model 88SE9230, the USB3.0 controller of controller.
5. according to a kind of any graphics workstation systems based on Domestic Platform of claim 1-4, which is characterized in that institute
The uplink port for stating PCIE3.0 extended chips is connect by PCIE3.0X16 with the domestic CPU, and the PCIE3.0 extends core
The downlink port of piece expands USB3.0 interfaces, gigabit networking interface and SATA3.0 interfaces respectively by PCIE conversion chips.
6. a kind of graphics workstation system based on Domestic Platform according to claim 5, which is characterized in that described
PCIE3.0 extended chips also expand two PCIE3.0X16 slots and 1 PCIE3.0X8 slot, and PCIE3.0X8 slots are used for
Other PCIE interfaces are connected, two PCIE3.0X16 slots are for connecting two domestic video card JM7200, and two aobvious
Card is directly interengageable for transmitting data by JJWLINK.
7. a kind of graphics workstation system based on Domestic Platform according to claim 6, which is characterized in that described
The model PEX8796 of PCIE3.0 extended chips.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810203607.8A CN108460719A (en) | 2018-03-13 | 2018-03-13 | A kind of graphics workstation system based on Domestic Platform |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810203607.8A CN108460719A (en) | 2018-03-13 | 2018-03-13 | A kind of graphics workstation system based on Domestic Platform |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108460719A true CN108460719A (en) | 2018-08-28 |
Family
ID=63219897
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810203607.8A Pending CN108460719A (en) | 2018-03-13 | 2018-03-13 | A kind of graphics workstation system based on Domestic Platform |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108460719A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111142951A (en) * | 2019-12-27 | 2020-05-12 | 中国电子科技集团公司第十五研究所 | Double-independent-display-card synchronous display device and method of Feiteng platform |
| CN111158625A (en) * | 2019-12-31 | 2020-05-15 | 中国电子科技集团公司第十五研究所 | Double-independent display card dynamic switching device and method based on Feiteng platform |
| CN112711503A (en) * | 2020-12-28 | 2021-04-27 | 北京同有飞骥科技股份有限公司 | Storage testing method based on Feiteng 2000+ CPU |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN202548731U (en) * | 2011-12-31 | 2012-11-21 | 曙光信息产业股份有限公司 | Extension module for supporting graphic workstation |
| CN106569877A (en) * | 2016-11-14 | 2017-04-19 | 中国石油化工股份有限公司 | Method for sharing graphic workstation GPU by virtual machines in direct connection way |
| CN206147604U (en) * | 2016-07-26 | 2017-05-03 | 浪潮电子信息产业股份有限公司 | PCIE switches module based on automatic GPU that matches of openpower platform |
| CN206209522U (en) * | 2016-09-30 | 2017-05-31 | 长沙景嘉微电子股份有限公司 | A kind of home brewed computer mainboard |
| CN107632953A (en) * | 2017-09-14 | 2018-01-26 | 郑州云海信息技术有限公司 | A kind of GPU casees PCIE extends interconnection topology device |
-
2018
- 2018-03-13 CN CN201810203607.8A patent/CN108460719A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN202548731U (en) * | 2011-12-31 | 2012-11-21 | 曙光信息产业股份有限公司 | Extension module for supporting graphic workstation |
| CN206147604U (en) * | 2016-07-26 | 2017-05-03 | 浪潮电子信息产业股份有限公司 | PCIE switches module based on automatic GPU that matches of openpower platform |
| CN206209522U (en) * | 2016-09-30 | 2017-05-31 | 长沙景嘉微电子股份有限公司 | A kind of home brewed computer mainboard |
| CN106569877A (en) * | 2016-11-14 | 2017-04-19 | 中国石油化工股份有限公司 | Method for sharing graphic workstation GPU by virtual machines in direct connection way |
| CN107632953A (en) * | 2017-09-14 | 2018-01-26 | 郑州云海信息技术有限公司 | A kind of GPU casees PCIE extends interconnection topology device |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111142951A (en) * | 2019-12-27 | 2020-05-12 | 中国电子科技集团公司第十五研究所 | Double-independent-display-card synchronous display device and method of Feiteng platform |
| CN111158625A (en) * | 2019-12-31 | 2020-05-15 | 中国电子科技集团公司第十五研究所 | Double-independent display card dynamic switching device and method based on Feiteng platform |
| CN112711503A (en) * | 2020-12-28 | 2021-04-27 | 北京同有飞骥科技股份有限公司 | Storage testing method based on Feiteng 2000+ CPU |
| CN112711503B (en) * | 2020-12-28 | 2024-03-26 | 北京同有飞骥科技股份有限公司 | Memory test method based on Feiteng 2000+CPU |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105630441B (en) | A kind of GPU system based on unified staining technique | |
| CN105209960B (en) | System, the method and computer program product of image are generated for nearly eye field display | |
| US10217183B2 (en) | System, method, and computer program product for simultaneous execution of compute and graphics workloads | |
| US11106261B2 (en) | Optimal operating point estimator for hardware operating under a shared power/thermal constraint | |
| US10013731B2 (en) | Maximizing parallel processing in graphics processors | |
| CN108460719A (en) | A kind of graphics workstation system based on Domestic Platform | |
| CN100570562C (en) | Video card, the dummy machine system of using this video card and display processing method | |
| WO2008097437A2 (en) | Method and system for processing a volume visualization dataset | |
| US9378533B2 (en) | Central processing unit, GPU simulation method thereof, and computing system including the same | |
| US9471952B2 (en) | Method and system for coordinated data execution using a primary graphics processor and a secondary graphics processor | |
| EP4332906A2 (en) | Game engine on a chip | |
| DE112017003838T5 (en) | THREAD PRIORITY MECHANISM | |
| CN109992103A (en) | During rendering using watch attentively information adjust the angle sample rate | |
| CN103885903A (en) | Technique For Performing Memory Access Operations Via Texture Hardware | |
| CN106796536A (en) | Memory pool access method, device and electronic equipment for multiple operating system | |
| US7979683B1 (en) | Multiple simultaneous context architecture | |
| US20130187956A1 (en) | Method and system for reducing a polygon bounding box | |
| DE102020104651A1 (en) | Memory compression hash mechanism | |
| US8854379B2 (en) | Routing across multicore networks using real world or modeled data | |
| US20140253413A1 (en) | System, method, and computer program product for representing a group of monitors participating in a desktop spanning environment to an operating system | |
| Huang et al. | Interactive illustrative rendering on mobile devices | |
| US9019284B2 (en) | Input output connector for accessing graphics fixed function units in a software-defined pipeline and a method of operating a pipeline | |
| CN111813449A (en) | Operation method, device and related product | |
| US8947444B1 (en) | Distributed vertex attribute fetch | |
| CN114201727A (en) | Data processing method, processor, artificial intelligence chip and electronic equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180828 |