CN109840878A - It is a kind of based on SystemC towards GPU parameter management method - Google Patents
It is a kind of based on SystemC towards GPU parameter management method Download PDFInfo
- Publication number
- CN109840878A CN109840878A CN201811520556.8A CN201811520556A CN109840878A CN 109840878 A CN109840878 A CN 109840878A CN 201811520556 A CN201811520556 A CN 201811520556A CN 109840878 A CN109840878 A CN 109840878A
- Authority
- CN
- China
- Prior art keywords
- register
- variable
- graphfunbar
- graphdrawbar
- module
- 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.)
- Granted
Links
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Image Generation (AREA)
Abstract
The present invention relates to computer hardware modeling technique fields, offer it is a kind of based on SystemC towards GPU parameter management method, include: step 1: executing barrier/lock tube reason, new barrier state is calculated according to current barrier state and fifo status, judge when graphRegLock is 0, graphical management module (2) are output to by transaction-level interface, execute step 2, otherwise when graphDrawLock is 0 and graphFunLock is 0, it is output to register window management module (3) by transaction-level interface, executes step 3;Step 2: executing Graphic drawing commands parameter and function code parameter management;Step 3: executing the read-write management of register parameters.
Description
Technical field
The present invention relates to computer hardware modeling technique field more particularly to it is a kind of based on SystemC towards GPU parameter
Management method.
Background technique
As that graphically applies is continuously increased, the solution for depending merely on CPU progress graphic plotting in early days has been difficult to completely
The graphics process demand that sufficient achievement and technology increase, graphics processor (Graphic Processing Unit, GPU) meet the tendency of and
It is raw.So far from Nvidia publication first item GPU product in 1999, the development of GPU technology mainly experienced fixed function assembly line
Stage, separation stainer framework stage, unified stainer framework stage, graphics capability are constantly promoted, application field
General-purpose computations field is gradually expanded to from initial graphic plotting.GPU assembly line high speed, parallel feature and flexible programmable
Ability provides good operation platform for graphics process and universal parallel calculating.
Currently, China's GPU R&D capability is weak, largely using the commercialization of external import in each field display control program
GPU chip.Especially in military domain, there are safeties, reliability, protection etc. for external import commercialization GPU chip
Hidden danger, be unable to satisfy the demand of military environment;Moreover, for reasons such as politics, military affairs, economy, it is external that skill is carried out to China
Art " block " and product " monopolization ", it is difficult to the Floor layer Technology data for obtaining GPU chip, such as register data, the micro- frame of detailed inner
Structure, kernel software source code etc., cause GPU function, performance to be unable to give full play, and transplantability is poor;The above problem seriously restricts
The independent of China's display system is developed and independent development capability, and graphics processor key technology, development graph processor chips are broken through
It is extremely urgent.
GPU chip development hardware logic is huge, and complexity is higher and higher, and it is right on higher abstraction hierarchy to need
Design is described, to can be carried out the emulation of higher speed, the exploration of software/hardware collaborative simulation and architectural framework.Work as design
When being expressed as system-level model, selects different algorithms to carry out repeatedly attempting to be easily done to design, use difference instead
Structure tested and cracking can also be completed;If carrying out design of expression using Method at Register Transfer Level or gate-level model, scale is logical
It is often quite big, to sound out different design structures or doing some changes, even if be not it is too difficult, it is also fairly time consuming laborious.
SystemC pushes its exploitation as a kind of language and standardized key factor is can to carry out system level design,
And the framework of hardware and the algorithm of software can be described, support verifying is exchanged with IP's.Made on a system level using SystemC
It is more much easier than other language for the segmentation tradeoff of software and hardware, and emulation is carried out than carrying out emulation using multilingual
It is quickly more.Therefore the imitative of a complete standard can be established using the micro-structure for being designed and being described unit based on SystemC
True environment, the Direct Modeling on high abstraction hierarchy.
Summary of the invention
The problem of based on background technique, it is provided by the invention it is a kind of based on SystemC towards GPU parameter management
Method is able to solve the problem of RTL emulates GPU parameter management cell data precise alignment, can shift to an earlier date RTL to GPU parametron
The hardware micro-structure of unit is managed the problem of carrying out functional verification on TLM model.
The technical solution of the invention is as follows:
It is a kind of based on SystemC towards GPU parameter management method, comprising:
Step 1: executing barrier/lock tube reason, new barrier state is calculated according to current barrier state and fifo status, is judged
When graphRegLock is 0, graphical management module 2 is output to by transaction-level interface, executes step 2, otherwise when
When graphDrawLock is 0 and graphFunLock is 0, register window management module 3 is output to by transaction-level interface,
Execute step 3;
Step 2: executing Graphic drawing commands parameter and function code parameter management;
Step 3: executing the read-write management of register parameters.
Step 1 includes:
Step 11: when graphDrawBar be 0, graphFunBar 0, graphDrawFifoStatus be non-empty and
When graphFunFifoStatus is empty, graphFunBar is set to 1;It is no to then follow the steps 12;
Step 12: when graphDrawBar is 0, graphFunBar 0, graphDrawFifoStatus be it is empty and
When graphFunFifoStatus is non-empty, graphDrawBar is set to 1;It is no to then follow the steps 13;
Step 13: when graphDrawBar is 0, graphFunBar 1, graphDrawFifoStatus be it is empty and
When graphFunFifoStatus is non-empty, graphDrawBar is set to 1, graphFunBar and is set to 0;It is no to then follow the steps
14;
Step 14: when graphDrawBar be 1, graphFunBar 0, graphDrawFifoStatus be non-empty and
When graphFunFifoStatus is empty, graphDrawBar is set to 0, graphFunBar and is set to 1;It is no to then follow the steps 15;
Step 15: graphDrawLock, graphFunLock, graphRegLock are read by transaction-level interface, when
When graphRegLock is 0, step 16 is executed, when graphDrawLock is 0 and graphFunLock is 0, executes step
17;
Step 16: all variable informations and data fifo information are output to graphical management module 2 by things grade interface,
Execute graphic parameter management;
Step 17: all variable informations and register window data information are output to register window by things grade interface
Mouth management module 3, executes register parameters management.
Step 2 includes:
Step 21: test pattern module 21 is read from barrier/lock management module 1 variable graphRegLock, if
It locks, then circulation step 21;It is no to then follow the steps 22;
Step 22: test pattern module 21 read from barrier/lock management module 1 variable graphDrawBar,
GraphFunBar, graphDrawFifoStatus, graphFunFifoStatus, when graphDrawBar be 0,
GraphFunBar is 1, and graphDrawFifoStatus be non-empty when, by transaction-level interface be connected to rendering order processing
Module 22 executes step 23;Otherwise when graphDrawBar is 1, graphFunBar 0, and graphFunFifoStatus is
When non-empty, function code processing module 23 is connected to by transaction-level interface, executes step 24;Otherwise step 21 is returned to;
Step 23: graphDrawLock is first set 1 by rendering order processing module 22, then carries out graphic plotting parameter
Decoding and assembling, and assembled data are handed down to task scheduling modules 5 by things grade interface, it will after having executed
GraphDrawLock sets 0;
Step 24: graphFunLock is first set 1 by function code processing module 23, then carries out the decoding of graphing capability parameter
And assembling, and assembled data are handed down to corresponding graphics module 6 by things grade interface, it will after having executed
GraphFunLock sets 0.
Step 3 includes:
Step 31: graphRegLock is set to 1;
Step 32: judging register read-write type DataType, if reading, execute step 33;If writing, step is executed
34;
Step 33: external register module 7 being connected to by transaction-level interface, is executed according to register address
Corresponding register data DataStruct is written in DataAddr;
Step 34: external register module 7 being connected to by transaction-level interface, is executed according to register address
DataAddr reads corresponding register data DataStruct;
Step 35: graphRegLock is set to 0.
In this method:
Inside includes two FIFO, and receive the two class orders from command processing module 4: wherein rendering order is stored in figure
Shape rendering order FIFO, graphing capability order are stored in graphing capability order FIFO;
Inside includes data information: register window data specifically include register read-write type DataType, register
Data DataStruct, register address DataAddr;
Inside includes variable:
For indicating the variable of Graphic drawing commands barrier state, it is denoted as graphDrawBar, is initialized as 0;
For indicating the variable of graphing capability order barrier state, it is denoted as graphFunBar and is initialized as 0;
For indicating the variable of Graphic drawing commands fifo status, it is denoted as graphDrawFifoStatus;
For indicating the variable of graphing capability order fifo status, it is denoted as graphFunFifoStatus;
For indicating the variable of Graphic drawing commands lock status, it is denoted as graphDrawLock;
For indicating the variable of graphing capability order lock status, it is denoted as graphFunLock;
For indicating the variable of register window lock status, it is denoted as graphRegLock.
The beneficial effects of the present invention are:
1, it is provided by the invention based on SystemC towards GPU parameter management method, be internally integrated barrier/lock tube reason mould
Block 1, graphical management module 2 and register window management module Unit 3 can be realized distribution and the pipe of GPU various types parameter
Reason.
2, barrier/1 subelement of lock management module realizes the monitoring and management of lock and barrier, ensure that various types parameter
Succession journey in the management process.
3,2 subelement of graphical management module realizes the distribution and management function of graphic plotting parameter, has specifically included figure
Shape rendering parameter and the two kinds of parameter management of graphing capability parameter.
4,3 subelement of register window management module realizes the distribution and management function of window registers.
5, the present invention solves the problems, such as the comparison of GPU parameter management unit R TL simulation result model, solves towards GPU
The problem of TLM microstructure unit functional verification of parameter management unit hardware, and accelerate the speed of emulation.
Detailed description of the invention
Fig. 1 is towards GPU parameter management method block diagram;
Fig. 2 is the TLM micro-structure block diagram of 2 hardware of graphical management module.
Specific embodiment
It is a kind of based on SystemC towards GPU parameter management method, comprising:
Step 1: executing barrier/lock tube reason, new barrier state is calculated according to current barrier state and fifo status, is judged
When graphRegLock is 0, graphical management module 2 is output to by transaction-level interface, executes step 2, otherwise when
When graphDrawLock is 0 and graphFunLock is 0, register window management module 3 is output to by transaction-level interface,
Execute step 3;
Step 2: executing Graphic drawing commands parameter and function code parameter management;
Step 3: executing the read-write management of register parameters.
The step 1 includes:
Step 11: when graphDrawBar be 0, graphFunBar 0, graphDrawFifoStatus be non-empty and
When graphFunFifoStatus is empty, graphFunBar is set to 1;It is no to then follow the steps 12;
Step 12: when graphDrawBar is 0, graphFunBar 0, graphDrawFifoStatus be it is empty and
When graphFunFifoStatus is non-empty, graphDrawBar is set to 1;It is no to then follow the steps 13;
Step 13: when graphDrawBar is 0, graphFunBar 1, graphDrawFifoStatus be it is empty and
When graphFunFifoStatus is non-empty, graphDrawBar is set to 1, graphFunBar and is set to 0;It is no to then follow the steps
14;
Step 14: when graphDrawBar be 1, graphFunBar 0, graphDrawFifoStatus be non-empty and
When graphFunFifoStatus is empty, graphDrawBar is set to 0, graphFunBar and is set to 1;It is no to then follow the steps 15;
Step 15: graphDrawLock, graphFunLock, graphRegLock are read by transaction-level interface, when
When graphRegLock is 0, step 16 is executed, when graphDrawLock is 0 and graphFunLock is 0, executes step
17;
Step 16: all variable informations and data fifo information are output to graphical management module 2 by things grade interface,
Execute graphic parameter management;
Step 17: all variable informations and register window data information are output to register window by things grade interface
Mouth management module 3, executes register parameters management.
The step 2 includes:
Step 21: test pattern module 21 is read from barrier/lock management module 1 variable graphRegLock, if
It locks, then circulation step 21;It is no to then follow the steps 22;
Step 22: test pattern module 21 read from barrier/lock management module 1 variable graphDrawBar,
GraphFunBar, graphDrawFifoStatus, graphFunFifoStatus, when graphDrawBar be 0,
GraphFunBar is 1, and graphDrawFifoStatus be non-empty when, by transaction-level interface be connected to rendering order processing
Module 22 executes step 23;Otherwise when graphDrawBar is 1, graphFunBar 0, and graphFunFifoStatus is
When non-empty, function code processing module 23 is connected to by transaction-level interface, executes step 24;Otherwise step 21 is returned to;
Step 23: graphDrawLock is first set 1 by rendering order processing module 22, then carries out graphic plotting parameter
Decoding and assembling, and assembled data are handed down to task scheduling modules 5 by things grade interface, it will after having executed
GraphDrawLock sets 0;
Step 24: graphFunLock is first set 1 by function code processing module 23, then carries out the decoding of graphing capability parameter
And assembling, and assembled data are handed down to corresponding graphics module 6 by things grade interface, it will after having executed
GraphFunLock sets 0.
As shown in Fig. 2.
The step 3 includes:
Step 31: graphRegLock is set to 1;
Step 32: judging register read-write type DataType, if reading, execute step 33;If writing, step is executed
34;
Step 33: external register module 7 being connected to by transaction-level interface, is executed according to register address
Corresponding register data DataStruct is written in DataAddr;
Step 34: external register module 7 being connected to by transaction-level interface, is executed according to register address
DataAddr reads corresponding register data DataStruct;
Step 35: graphRegLock is set to 0.
In this method:
Inside includes two FIFO, and receive the two class orders from command processing module 4: wherein rendering order is stored in figure
Shape rendering order FIFO, graphing capability order are stored in graphing capability order FIFO;
Inside includes data information: register window data specifically include register read-write type DataType, register
Data DataStruct, register address DataAddr;
Inside includes variable:
For indicating the variable of Graphic drawing commands barrier state, it is denoted as graphDrawBar, is initialized as 0;
For indicating the variable of graphing capability order barrier state, it is denoted as graphFunBar and is initialized as 0;
For indicating the variable of Graphic drawing commands fifo status, it is denoted as graphDrawFifoStatus;
For indicating the variable of graphing capability order fifo status, it is denoted as graphFunFifoStatus;
For indicating the variable of Graphic drawing commands lock status, it is denoted as graphDrawLock;
For indicating the variable of graphing capability order lock status, it is denoted as graphFunLock;
For indicating the variable of register window lock status, it is denoted as graphRegLock.
It is a kind of based on SystemC towards GPU parameter management method, including barrier/lock management module 1, graphics management mould
Block 2, register window management module 3.
The barrier/lock management module 1 is connected to external command processing module 4 by transaction-level interface;Graphics management mould
3 two block 2, register window management module subelements in physics, be logically independent of each other, it is logical with barrier/lock management module 1
Transaction-level interface is crossed to be connected;Graphical management module 2 is connected to external tasks scheduler module 5 and drawing mould by transaction-level interface
Block 6;Register window management module 3 is connected to external register module 7 by transaction-level interface.
As shown in Fig. 1.
Finally it should be noted that the above embodiments are merely illustrative of the technical solutions of the present invention, rather than its limitations;Although
Present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that;It still may be used
To modify to the technical solution that foregoing embodiments are recorded or equivalent replacement of some of the technical features;And
These are modified or replaceed, the spirit and model of technical solution of various embodiments of the present invention that it does not separate the essence of the corresponding technical solution
It encloses.
Claims (5)
1. it is a kind of based on SystemC towards GPU parameter management method, it is characterised in that: include:
Step 1: executing barrier/lock tube reason, new barrier state is calculated according to current barrier state and fifo status, judgement is worked as
GraphRegLock be 0 when, be output to graphical management module (2) by transaction-level interface, execute step 2, otherwise when
When graphDrawLock is 0 and graphFunLock is 0, register window management module is output to by transaction-level interface
(3), step 3 is executed;
Step 2: executing Graphic drawing commands parameter and function code parameter management;
Step 3: executing the read-write management of register parameters.
2. it is according to claim 1 it is a kind of based on SystemC towards GPU parameter management method, it is characterised in that: step
1 includes:
Step 11: when graphDrawBar be 0, graphFunBar 0, graphDrawFifoStatus be non-empty and
When graphFunFifoStatus is empty, graphFunBar is set to 1;It is no to then follow the steps 12;
Step 12: when graphDrawBar is 0, graphFunBar 0, graphDrawFifoStatus be it is empty and
When graphFunFifoStatus is non-empty, graphDrawBar is set to 1;It is no to then follow the steps 13;
Step 13: when graphDrawBar is 0, graphFunBar 1, graphDrawFifoStatus be it is empty and
When graphFunFifoStatus is non-empty, graphDrawBar is set to 1, graphFunBar and is set to 0;It is no to then follow the steps
14;
Step 14: when graphDrawBar be 1, graphFunBar 0, graphDrawFifoStatus be non-empty and
When graphFunFifoStatus is empty, graphDrawBar is set to 0, graphFunBar and is set to 1;It is no to then follow the steps 15;
Step 15: graphDrawLock, graphFunLock, graphRegLock are read by transaction-level interface, when
When graphRegLock is 0, step 16 is executed, when graphDrawLock is 0 and graphFunLock is 0, executes step
17;
Step 16: all variable informations and data fifo information being output to graphical management module (2) by things grade interface, are held
Row graphic parameter management;
Step 17: all variable informations and register window data information are output to register window pipe by things grade interface
It manages module (3), executes register parameters management.
3. it is according to claim 1 it is a kind of based on SystemC towards GPU parameter management method, it is characterised in that: step
2 include:
Step 21: test pattern module (21), which is read, comes from barrier/lock management module (1) variable graphRegLock, if
It locks, then circulation step 21;It is no to then follow the steps 22;
Step 22: test pattern module (21) read from barrier/lock management module (1) variable graphDrawBar,
GraphFunBar, graphDrawFifoStatus, graphFunFifoStatus, when graphDrawBar be 0,
GraphFunBar is 1, and graphDrawFifoStatus be non-empty when, by transaction-level interface be connected to rendering order processing
Module (22) executes step 23;Otherwise when graphDrawBar is 1, graphFunBar 0, and graphFunFifoStatus
It when for non-empty, is connected to function code processing module (23) by transaction-level interface, executes step 24;Otherwise step 21 is returned to;
Step 23: graphDrawLock is first set 1 by rendering order processing module (22), then carries out translating for graphic plotting parameter
Code and assembling, and assembled data are handed down to task scheduling modules (5) by things grade interface, it will after having executed
GraphDrawLock sets 0;
Step 24: graphFunLock is first set 1 by function code processing module (23), then carry out graphing capability parameter decoding and
Assembling, and assembled data are handed down to corresponding graphics module (6) by things grade interface, it will after having executed
GraphFunLock sets 0.
4. it is according to claim 1 it is a kind of based on SystemC towards GPU parameter management method, it is characterised in that: step
3 include:
Step 31: graphRegLock is set to 1;
Step 32: judging register read-write type DataType, if reading, execute step 33;If writing, step 34 is executed;
Step 33: it is connected to external register module (7), is executed according to register address DataAddr by transaction-level interface,
Corresponding register data DataStruct is written;
Step 34: it is connected to external register module (7), is executed according to register address DataAddr by transaction-level interface,
Read corresponding register data DataStruct;
Step 35: graphRegLock is set to 0.
5. it is according to claim 1 it is a kind of based on SystemC towards GPU parameter management method, it is characterised in that: we
In method:
Inside includes two FIFO, and receive the two class orders for coming from command processing module (4): wherein rendering order is stored in figure
Rendering order FIFO, graphing capability order are stored in graphing capability order FIFO;
Inside includes data information: register window data specifically include register read-write type DataType, register data
DataStruct, register address DataAddr;
Inside includes variable:
For indicating the variable of Graphic drawing commands barrier state, it is denoted as graphDrawBar, is initialized as 0;
For indicating the variable of graphing capability order barrier state, it is denoted as graphFunBar and is initialized as 0;
For indicating the variable of Graphic drawing commands fifo status, it is denoted as graphDrawFifoStatus;
For indicating the variable of graphing capability order fifo status, it is denoted as graphFunFifoStatus;
For indicating the variable of Graphic drawing commands lock status, it is denoted as graphDrawLock;
For indicating the variable of graphing capability order lock status, it is denoted as graphFunLock;
For indicating the variable of register window lock status, it is denoted as graphRegLock.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811520556.8A CN109840878B (en) | 2018-12-12 | 2018-12-12 | GPU parameter management method based on SystemC |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811520556.8A CN109840878B (en) | 2018-12-12 | 2018-12-12 | GPU parameter management method based on SystemC |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109840878A true CN109840878A (en) | 2019-06-04 |
CN109840878B CN109840878B (en) | 2023-05-23 |
Family
ID=66883185
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811520556.8A Active CN109840878B (en) | 2018-12-12 | 2018-12-12 | GPU parameter management method based on SystemC |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109840878B (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111047504A (en) * | 2019-11-21 | 2020-04-21 | 中国航空工业集团公司西安航空计算技术研究所 | TLM microstructure for GPU sub-image processing based on SystemC |
CN111880916A (en) * | 2020-07-27 | 2020-11-03 | 长沙景嘉微电子股份有限公司 | Multi-drawing task processing method, device, terminal, medium and host in GPU |
CN112559139A (en) * | 2020-12-05 | 2021-03-26 | 西安翔腾微电子科技有限公司 | SystemC-based multi-GPU transaction-level model device and operation method |
CN112581581A (en) * | 2020-12-24 | 2021-03-30 | 西安翔腾微电子科技有限公司 | GPU window transformation module TLM device based on SysML view and operation method |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20150333957A1 (en) * | 2015-02-12 | 2015-11-19 | Advanced Micro Devices, Inc. | Centralized distribution of configuration parameters for a cluster server |
US20150378956A1 (en) * | 2014-06-27 | 2015-12-31 | Advanced Micro Devices, Inc. | Memory physical layer interface logic for generating dynamic random access memory (dram) commands with programmable delays |
GB201704528D0 (en) * | 2016-04-07 | 2017-05-03 | Imagination Tech Ltd | Processors supporting atomic writes to multiword memory locations & methods |
CN106648547A (en) * | 2016-12-12 | 2017-05-10 | 中国航空工业集团公司西安航空计算技术研究所 | Distributed unified management method for GPU graphic state parameters |
US9876729B1 (en) * | 2015-06-24 | 2018-01-23 | Cadence Design Systems, Inc. | Method and system for efficient data streaming in an emulation system |
-
2018
- 2018-12-12 CN CN201811520556.8A patent/CN109840878B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20150378956A1 (en) * | 2014-06-27 | 2015-12-31 | Advanced Micro Devices, Inc. | Memory physical layer interface logic for generating dynamic random access memory (dram) commands with programmable delays |
US20150333957A1 (en) * | 2015-02-12 | 2015-11-19 | Advanced Micro Devices, Inc. | Centralized distribution of configuration parameters for a cluster server |
US9876729B1 (en) * | 2015-06-24 | 2018-01-23 | Cadence Design Systems, Inc. | Method and system for efficient data streaming in an emulation system |
GB201704528D0 (en) * | 2016-04-07 | 2017-05-03 | Imagination Tech Ltd | Processors supporting atomic writes to multiword memory locations & methods |
CN106648547A (en) * | 2016-12-12 | 2017-05-10 | 中国航空工业集团公司西安航空计算技术研究所 | Distributed unified management method for GPU graphic state parameters |
Non-Patent Citations (1)
Title |
---|
阙恒: "嵌入式图形处理器设计", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111047504A (en) * | 2019-11-21 | 2020-04-21 | 中国航空工业集团公司西安航空计算技术研究所 | TLM microstructure for GPU sub-image processing based on SystemC |
CN111047504B (en) * | 2019-11-21 | 2023-06-13 | 中国航空工业集团公司西安航空计算技术研究所 | TLM microstructure for GPU sub-image processing based on SystemC |
CN111880916A (en) * | 2020-07-27 | 2020-11-03 | 长沙景嘉微电子股份有限公司 | Multi-drawing task processing method, device, terminal, medium and host in GPU |
CN112559139A (en) * | 2020-12-05 | 2021-03-26 | 西安翔腾微电子科技有限公司 | SystemC-based multi-GPU transaction-level model device and operation method |
CN112559139B (en) * | 2020-12-05 | 2022-12-13 | 西安翔腾微电子科技有限公司 | SystemC-based multi-GPU transaction-level model device and operation method |
CN112581581A (en) * | 2020-12-24 | 2021-03-30 | 西安翔腾微电子科技有限公司 | GPU window transformation module TLM device based on SysML view and operation method |
CN112581581B (en) * | 2020-12-24 | 2024-04-26 | 西安翔腾微电子科技有限公司 | GPU window transformation module TLM device based on SysML view and operation method |
Also Published As
Publication number | Publication date |
---|---|
CN109840878B (en) | 2023-05-23 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109840878A (en) | It is a kind of based on SystemC towards GPU parameter management method | |
US5493508A (en) | Specification and design of complex digital systems | |
US6289498B1 (en) | VDHL/Verilog expertise and gate synthesis automation system | |
US5526517A (en) | Concurrently operating design tools in an electronic computer aided design system | |
US7523023B1 (en) | Automatic generation of component interfaces for computational hardware implementations generated from a block diagram model | |
US7472361B2 (en) | System and method for generating a plurality of models at different levels of abstraction from a single master model | |
US9881119B1 (en) | Methods, systems, and computer program product for constructing a simulation schematic of an electronic design across multiple design fabrics | |
US7584456B1 (en) | Method and apparatus for debugging embedded systems having read only memory | |
US10417355B1 (en) | Systems and methods for constructing and modifying computer models | |
US20210287120A1 (en) | Machine learning-based prediction of metrics at early-stage circuit design | |
US7437282B2 (en) | Method and apparatus to provide alternative stimulus to signals internal to a model actively running on a logic simulation hardware emulator | |
US20160011920A1 (en) | Message-based modeling | |
CN109669769A (en) | One kind is based on SystemC towards GPU vertex coloring method for scheduling task | |
US11734480B2 (en) | Performance modeling and analysis of microprocessors using dependency graphs | |
JP2002099584A (en) | System and method for verifying design and computer- readable medium with program for design verification recorded thereon | |
CN109727186B (en) | SystemC-based GPU (graphics processing Unit) fragment coloring task scheduling method | |
CN102184290A (en) | Cycle accurate and bit accurate system level model for embedded micro-processor | |
US7131099B2 (en) | Method, apparatus, and computer program product for RTL power sequencing simulation of voltage islands | |
US20230315964A1 (en) | Design aware adaptive mixed-signal simulation | |
CN110765715B (en) | GPU chip rendering output unit performance simulation method and platform | |
CN111028128A (en) | GPU (graphics processing Unit) -oriented vertex output control method and unit based on SystemC | |
US7761832B2 (en) | Method for incremental, timing-driven, physical-synthesis optimization under a linear delay model | |
US9875329B1 (en) | Method and system for import of mask layout data to a target system | |
CN111028132A (en) | SystemC-based GPU command processor unit hardware TLM microstructure | |
CN110930494A (en) | GPU chip-oriented rendering output device and modeling method |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |