CN104281543B - Display controller and graphics accelerator is supported to access the framework method of internal memory simultaneously - Google Patents
Display controller and graphics accelerator is supported to access the framework method of internal memory simultaneously Download PDFInfo
- Publication number
- CN104281543B CN104281543B CN201310270956.9A CN201310270956A CN104281543B CN 104281543 B CN104281543 B CN 104281543B CN 201310270956 A CN201310270956 A CN 201310270956A CN 104281543 B CN104281543 B CN 104281543B
- Authority
- CN
- China
- Prior art keywords
- data
- display controller
- memory
- framework
- passage
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T1/00—General purpose image data processing
- G06T1/60—Memory management
-
- 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
Abstract
The present invention relates to a kind of while support display controller and graphics accelerator to access the framework method of internal memory, step is as follows:Step 1, create the framework of an access internal memory;The framework of the access internal memory set in SOC overall frameworks includes:Memory Controller Hub, X-Y scheme processor, data buffer, data selector, data selection control, display controller;Step 2, the multilayer source images according to superposition mixing whether there is renewal, and the framework for accessing internal memory provides flowing in and out for data:When there is renewal, opening/closing information is allocated to Memory Controller Hub by multi-core CPU, and data are overlapped mixed processing by X-Y scheme processor, and through data buffer, data selector, are directly output to display controller.When without renewal, display controller reads the result of multi-layer image mixing superposition directly from internal storage.Advantage of the present invention is as follows:Optimizing in framework has a case that extra bandwidth consumption, is always maintained at the minimum mode of operation of bandwidth consumption.
Description
Technical field:
The present invention relates to a kind of multi-layer graphical image blend framework field, and display control is supported simultaneously more particularly, to a kind of
Device and graphics accelerator access the framework method of internal memory.
Background technology:
In existing Android operation system, multi-layer graphical image blend be system show and graph and image processing during most
Video, figure or desktop background image are mainly read in an important step, multi-layer image mixing, and there is each figure layer to mix for output
Close the piece image of Overlay and be transferred to the process of display device.This figure layer mixing superposition and transmitting procedure generally have
Two kinds:
First method is to carry out multi-layer graphical image blend using X-Y scheme processor.In this approach, first
It will think that layer position information to be processed is sent to X-Y scheme processor by drive system, then be pressed by X-Y scheme processor
The layer position information sent according to operating system reads in each figure layer successively, and processing is overlapped according to reading order, handles
The data that Overlay is mixed with figure layer are written in internal storage again after finishing, last display controller is from internal memory
The middle output result for reading X-Y scheme processor, is shown.This method uses X-Y scheme processor and display control
Device respectively by respective Memory Controller Hub access internal storage, although realize it is fairly simple, it require that X-Y scheme
Mixing resultant is write internal storage by processor, and display controller will read the number of image blend from internal storage again
According to.The reading of internal memory and write and to consume very big bandwidth, by taking high definition 1080P video resolutions as an example, once read-write consumption
With a width of 16 Mbytes.If the refresh rate of one second display device is 60 frames, a width of 960,000,000 word of band of consumption in that second
Section, by taking the superposition of 3 layer patterns as an example, the bandwidth of more 960 Mbytes of consumption 39% bandwidth equivalent to the system wastes.It is so huge
Big bandwidth consumption can increase the burden and power consumption of system, influence systematic entirety energy.System also will when data are read
Power consumption is consumed, wastes the power consumption that 39% bandwidth can also make DDR systems consume 39% more.
Second method is to carry out multi-layer graphical image blend superposition using display controller.This display controller has
The function of imaging importing, mainly completed by being superimposed the passage of figure layer.The source images for being first superimposed needs by driver
To display controller, display controller is read in each source images by being superimposed figure layer passage information configuration, then right internally
The source data of each figure layer is overlapped mixed processing, and the overlay order of each figure layer and stacked system are by drive control, finally
The result that image blend is superimposed is exported online and shown to display device.The software control flow of this implementation compares
Simply, but need to carry out mixing superposition to every two field picture of output.It is frequently very much this side not to be when the image change of display
Method consumes very big to system bandwidth.Exemplified by HD video resolution ratio so that refresh rate is 60 frames as an example, if the display of rear 30 frame
Image is constant, and this 30 image is formed by stacking by the mixing of the width source images of identical 3, and that each frame does more and is once superimposed
Mixing will consume 16 Mbytes of bandwidth more, and that this scheme will consume 480,000,000 bandwidth at one second more, and consume 480 more
Mbytes bandwidth equivalent to the system wastes 32% bandwidth.System will also consume power consumption when data are read, wave
The bandwidth of expense 32% can also make the power consumption of DDR systems consumption 32% more.Display controller is due to the limitation of framework and function, it is impossible to
The function of all multi-layer graphical imaging importing mixing is completed, for example image rotation, scaling, this method use upper in function
Limitation be present.
Moving multimedia equipment and home entertaining multimedia equipment just develop towards high-resolution at present, multimedia hard
In part solution, bandwidth resources become more and more nervous and precious, save the common objective that power consumption is multimedia SOC.How
Reduce the consumption of system bandwidth and power consumption as far as possible while user's visual experience is improved, become multimedia SOC design field
One difficult point and important topic.Therefore the present invention is a kind of saves system with high-performance low-power-consumption completion multi-layer image mixing superposition and can
The architecture design of system bandwidth consumption, turn into the development priority and Main way in present multimedia SOC design field.
The content of the invention:
The purpose of the present invention provides a kind of reduction system bandwidth and power consumption consumption for prior art weak point
Display controller and graphics accelerator is supported to access the framework method of internal memory simultaneously.
The purpose of the present invention is realized by following measures:It is a kind of to support display controller and graphics accelerator to visit simultaneously
Ask the framework method of internal memory, it is characterised in that step is as follows:
Step 1, create the framework of an access internal memory;The framework bag of the access internal memory set in SOC overall frameworks
Include:Memory Controller Hub, X-Y scheme processor, data buffer, data selector, data selection control, display controller;
The X-Y scheme processor is connected by data buffer, data selector with display controller, data selection control with
X-Y scheme processor, data selector, display controller are connected, and Memory Controller Hub is provided with data application passage 1, data return
Passage 1 and data application passage 2 and data backward channel 2;The data application passage 2 and data backward channel 2 of Memory Controller Hub
It is connected with X-Y scheme processor, the data application passage 1 of Memory Controller Hub is connected with display controller, data backward channel 1
Be connected by data selector with display controller, Memory Controller Hub by the internal storage of bus and SOC overall frameworks,
Multi-core CPU is connected;
Step 2, the multilayer source images according to superposition mixing whether there is renewal, and the framework for accessing internal memory provides the inflow of data
And outflow:
When the multilayer source images for needing to be superimposed mixing have renewal, multi-core CPU returns to data application passage 2 and data logical
Opening/the closing information in road 2 is allocated to Memory Controller Hub, and X-Y scheme processor sends internal memory control by data application passage 2
Device read-write requests processed, Memory Controller Hub send read-write requests to multi-core CPU, then returned data are transferred at X-Y scheme
Reason, the data of the mixing overlap-add procedure device collection multilayer source images inside X-Y scheme processor, and data are overlapped mixed
Conjunction is handled, and the result that multiple-layer stacked is mixed is sent to data buffer, while data selection control opens data selection
Device, the result of the superposition mixing of data buffer are exported to display controller;
When needing to be overlapped the multilayer source images of mixing without renewal, multi-core CPU open X-Y scheme processor with it is interior
The passage of memory controller, the result that multiple-layer stacked is mixed write internal storage, and multi-core CPU opens the data of Memory Controller Hub
Apply for passage 1, display controller reads the result of multilayer source images mixing superposition directly from internal storage, if multilayer source
Image does not update, then reads the result of superposition mixing always, until multi-core CPU detects source images renewal next time.
Memory Controller Hub in the step 1 is used for the selection of control data application passage and opening for data backward channel
Closing, Memory Controller Hub opens or closes corresponding passage according to the instruction configuration information of multi-core CPU, while by opened passage
Data application information is sent to internal storage by bus, and after internal storage returned data confirms, Memory Controller Hub is again
Data are conveyed to corresponding passage.
X-Y scheme processor in the step 1 is used for the superposition mixed processing for carrying out multilayer source images, according to multinuclear
By data application passage 2, internally memory sends read-write requests for CPU instruction, and returned data then is overlapped into mixing
Processing, and send the result for being superimposed mixing to internal data buffer module or write-in internal storage.
Data buffer in the step 1 is used to adjust the data between X-Y scheme processor and display controller
Transmission speed.
Data selector in the step 1 is used for the flow direction for connecting returned data or by X-Y scheme processor with showing
Show controller connection or be connected internal storage with display controller.
Data selection control in the step 1 is used for the passage that flows to for opening data selector, and receives display control
The timing information is sent to X-Y scheme processor after timing information that device processed is sent, processing.
Display controller in the step 1 is used to the multilayer source images received being superimposed the result of mixing according to display
Sequential required for equipment sends display device to, and the synchronizing information of each frame is sent into data selection control;Foundation
By data application passage 1, internally memory sends read-write requests for the instruction of multi-core CPU, and data letter is read to internal storage
Breath.
Compared with prior art, employ proposed by the present invention a kind of while support display controller and graphics accelerator to visit
The framework method of internal memory is asked, is had the following advantages that:1) present invention is directed to multi-layer graphical imaging importing in existing SOC systems and mixed
The problem of Time Bandwidth anxiety, can minimum degree consumption system in all application scenarios while having reached guarantee Consumer's Experience
Regiment commander's width and power consumption.Eliminate the extra consumption to bandwidth and power consumption in traditionally two kinds of image blend frameworks.No matter multilayer source
It is constant that image is that real-time update is also to maintain, and is attained by bandwidth consumption minimum.2) compared with traditional two kinds of image blend frameworks,
When image real-time update, 960 Mbytes are saved each second than the first framework, is the bandwidth and 39% that system saves 39%
Power consumption is identical with second of framework consumed bandwidth in this case.When image keeps constant, disappeared with the first framework
Consumption bandwidth is identical, and than second framework saves 480 Mbytes of bandwidth, and 32% bandwidth and 32% power consumption are saved for system,
Solve the problems, such as to rotate and scale in second of framework and can not support simultaneously.In other words, optimize and have in two kinds of conventional architectures
The situation for having extra bandwidth to consume, it is always maintained at the minimum mode of operation of bandwidth consumption.3) present invention is directed to the more matchmakers of existing movement
The SOC systems of body equipment and family multi-media equipment, the collaborative work for considering existing submodule each in systems well are closed
System, there is the advantages that framework is clear, and the division of labor clearly, is easily realized, software control flow is simple.
Brief description of the drawings:
Fig. 1 is one embodiment structural representation proposed by the present invention.
Fig. 2 is the flow chart of Fig. 1 embodiments.
Embodiment:
Embodiment is elaborated below in conjunction with the accompanying drawings:
Fig. 1~Fig. 2 shows one embodiment of the present of invention.It is a kind of to support display controller and graphics accelerator to visit simultaneously
Ask the framework method of internal memory, step is as follows:
Step 1, the framework of an access internal memory is created, as shown in Figure 1;The access internal memory set in SOC overall frameworks
Framework include:It is Memory Controller Hub, X-Y scheme processor, data buffer, data selector, data selection control, aobvious
Show controller;The X-Y scheme processor is connected by data buffer, data selector with display controller, data selection
Controller is connected with X-Y scheme processor, data selector, display controller, Memory Controller Hub be provided with data application passage 1,
Data backward channel 1 and data application passage 2 and data backward channel 2;The data application passage 2 and data of Memory Controller Hub return
Return passage 2 with X-Y scheme processor to be connected, the data application passage 1 of Memory Controller Hub is connected with display controller, and data are returned
Return passage 1 by data selector with display controller to be connected, Memory Controller Hub passes through bus and the inside of SOC overall frameworks
Memory, multi-core CPU are connected.
Memory Controller Hub in this structure is used for the selection of control data application passage and the switch of data backward channel, interior
Memory controller opens or closes corresponding passage according to the instruction configuration information of the multi-core CPU of SOC overall frameworks, while will beat
The data application information of open walk is sent to internal storage by bus, after internal storage returned data confirms, internal memory
Data are conveyed to the passage being connected with X-Y scheme processor or display controller by controller again.
X-Y scheme processor in this structure is used for the superposition mixed processing for carrying out multilayer source images, according to multi-core CPU
Instruction by data application passage 2 internally memory send read-write requests, then by the data of the multilayer source images of return
Mixed processing is overlapped, and sends the result for being superimposed mixing to internal data buffer module or write-in internal storage.
Data buffer in this structure is used to adjust the data transfer between X-Y scheme processor and display controller
Speed.X-Y scheme processor and display controller are to work at different clock frequencies, so reading and writing data speed has
Institute is different, and in order to ensure that display controller can get the data that it needs in real time, X-Y scheme processor can be in advance by one
Partial data is write in data buffer storage to ensure data transmission bauds.
Data selector in this structure is used for the flow direction for connecting returned data, or X-Y scheme processor and display are controlled
Device connection processed connects internal storage and display controller.When there is multilayer source images to have renewal, data selector is by two dimension
Image processor connects with display controller;When having renewal without multilayer source images, data selector is by Memory Controller Hub with showing
Show that controller connects.
Data selection control in this structure is used for the passage that flows to for opening data selector, and receives display controller
The timing information is sent to X-Y scheme processor after the timing information of transmission, processing.
Display controller in this structure is used to the multilayer source images received being superimposed the result of mixing according to display device
Required sequential sends display device to, and the synchronizing information of each frame is sent into data selection control;According to multinuclear
By data application passage 1, internally memory sends read-write requests for CPU instruction, is read to internal storage at X-Y scheme
Reason device writes back the superposition mixing resultant image of internal storage.
Step 2, the multilayer source images according to superposition mixing whether there is renewal, and the framework for accessing internal memory provides the inflow of data
And outflow, shown in Fig. 2:
When the multilayer source images for needing to be superimposed mixing have renewal, multi-core CPU returns to data application passage 2 and data logical
Opening/the closing information in road 2 is allocated to Memory Controller Hub, and X-Y scheme processor sends internal memory control by data application passage 2
Device read-write requests processed, Memory Controller Hub send read-write requests to multi-core CPU, then returned data are transferred at X-Y scheme
Reason, the data of the mixing overlap-add procedure device collection multilayer source images inside X-Y scheme processor, and data are overlapped mixed
Conjunction is handled, and the result that multiple-layer stacked is mixed is sent to data buffer, while data selection control opens data selection
Device, the result of the superposition mixing of data buffer are exported to display controller.In the data flow, at X-Y scheme processor
The result of the multilayer source images superposition mixing of reason need not write back to internal storage, and display controller also without going to inside again
Data are read in memory, so as to reach the purpose for saving system bandwidth.By taking HD video resolution ratio as an example, once read-write consumption
A width of 16 Mbytes of band.If the refresh rate of one second display device is 60 frames, one second band a width of 960 saved of this framework
Mbytes.The power consumption because being consumed to internal memory read/write is also saved simultaneously.
When needing to be overlapped the multilayer source images of mixing without renewal, multi-core CPU open X-Y scheme processor with it is interior
The passage of memory controller, the result that multiple-layer stacked is mixed write internal storage, and multi-core CPU opens the data of Memory Controller Hub
Apply for passage 1, display controller reads the result of multilayer source images mixing superposition directly from internal storage, if multilayer source
Image does not update, then reads the result of superposition mixing always, until multi-core CPU detects source images renewal next time.At this
In data flow, if multilayer source images do not update, the result of superposition mixing can be read always, until driving detects down
Source images renewal.Exemplified by HD video resolution ratio so that refresh rate is 60 frames as an example, if the display image of rear 30 frame is
Constant, and this 30 image is formed by stacking by the mixing of the width source images of identical 3, that each frame does once superposition mixing just more
To consume 16 Mbytes of bandwidth, that this scheme was in one second bandwidth that will save 480 Mbytes more.Also save simultaneously
Because the power consumption consumed to internal memory read/write.
Embodiments presented above is not construed as limiting the invention, and those skilled in the art in the art are in appended power
Profit makes various deformations or amendments in protection domain in the range of requiring.
Claims (7)
1. framework method that is a kind of while supporting display controller and graphics accelerator access internal memory, it is characterised in that step is such as
Under:
Step 1, create the framework of an access internal memory;The framework of the access internal memory set in SOC overall frameworks includes:It is interior
Memory controller, X-Y scheme processor, data buffer, data selector, data selection control, display controller;It is described
X-Y scheme processor is connected by data buffer, data selector with display controller, data selection control and two dimension
Graphics processor, data selector, display controller are connected, and Memory Controller Hub is provided with data application passage 1, data backward channel
1 and data application passage 2 and data backward channel 2;The data application passage 2 and data backward channel 2 and two of Memory Controller Hub
Dimension graphics processor is connected, and the data application passage 1 of Memory Controller Hub is connected with display controller, and data backward channel 1 passes through
Data selector is connected with display controller, and Memory Controller Hub passes through bus and internal storage, the multinuclear of SOC overall frameworks
CPU is connected;
Step 2, the multilayer source images according to superposition mixing whether there is renewal, and the framework for accessing internal memory provides inflow and the stream of data
Go out:
When the multilayer source images for needing to be superimposed mixing have renewal, multi-core CPU is by data application passage 2 and data backward channel 2
Opening/closing information be allocated to Memory Controller Hub, X-Y scheme processor passes through data application passage 2 and sends Memory Controller Hub
Read-write requests, Memory Controller Hub send read-write requests to multi-core CPU, and returned data then is transferred into X-Y scheme processing,
Mixing overlap-add procedure device inside X-Y scheme processor collects the data of multilayer source images, and data are overlapped at mixing
Reason, and the result that multiple-layer stacked is mixed is sent to data buffer, while data selection control opens data selector, number
The result mixed according to the superposition of buffer is exported to display controller;
When needing to be overlapped the multilayer source images of mixing without renewal, multi-core CPU opens X-Y scheme processor and internal memory control
The passage of device processed, the result that multiple-layer stacked is mixed write internal storage, and multi-core CPU opens the data application of Memory Controller Hub
Passage 1, display controller reads the result of multilayer source images mixing superposition directly from internal storage, if multilayer source images
Do not update, then read the result of superposition mixing always, until multi-core CPU detects source images renewal next time.
2. framework method that is according to claim 1 while supporting display controller and graphics accelerator access internal memory, its
It is characterised by, the Memory Controller Hub in the step 1 is used for the selection of control data application passage and opening for data backward channel
Closing, Memory Controller Hub opens or closes corresponding passage according to the instruction configuration information of multi-core CPU, while by opened passage
Data application information is sent to internal storage by bus, and after internal storage returned data confirms, Memory Controller Hub is again
Data are conveyed to corresponding passage.
3. framework method that is according to claim 1 while supporting display controller and graphics accelerator access internal memory, its
It is characterised by, the X-Y scheme processor in the step 1 is used for the superposition mixed processing for carrying out multilayer source images, according to more
By data application passage 2, internally memory sends read-write requests for core CPU instruction, is then overlapped returned data mixed
Conjunction is handled, and sends the result for being superimposed mixing to internal data buffer module or write-in internal storage.
4. framework method that is according to claim 1 while supporting display controller and graphics accelerator access internal memory, its
It is characterised by, the data buffer in the step 1 is used to adjust the data between X-Y scheme processor and display controller
Transmission speed.
5. framework method that is according to claim 1 while supporting display controller and graphics accelerator access internal memory, its
Be characterised by, the data selector in the step 1 be used for connect returned data flow direction or by X-Y scheme processor with
Display controller connects or is connected internal storage with display controller.
6. framework method that is according to claim 1 while supporting display controller and graphics accelerator access internal memory, its
It is characterised by, the data selection control in the step 1 is used for the passage that flows to for opening data selector, and receives display
The timing information is sent to X-Y scheme processor after timing information that controller is sent, processing.
7. framework method that is according to claim 1 while supporting display controller and graphics accelerator access internal memory, its
It is characterised by, the display controller in the step 1 is used to the multilayer source images received being superimposed the result of mixing according to display
Sequential required for equipment sends display device to, and the synchronizing information of each frame is sent into data selection control;Foundation
By data application passage 1, internally memory sends read-write requests for the instruction of multi-core CPU, and data letter is read to internal storage
Breath.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310270956.9A CN104281543B (en) | 2013-07-01 | 2013-07-01 | Display controller and graphics accelerator is supported to access the framework method of internal memory simultaneously |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310270956.9A CN104281543B (en) | 2013-07-01 | 2013-07-01 | Display controller and graphics accelerator is supported to access the framework method of internal memory simultaneously |
Publications (2)
Publication Number | Publication Date |
---|---|
CN104281543A CN104281543A (en) | 2015-01-14 |
CN104281543B true CN104281543B (en) | 2017-12-26 |
Family
ID=52256435
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201310270956.9A Active CN104281543B (en) | 2013-07-01 | 2013-07-01 | Display controller and graphics accelerator is supported to access the framework method of internal memory simultaneously |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN104281543B (en) |
Families Citing this family (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107169917B (en) * | 2017-04-21 | 2020-06-16 | 图芯芯片技术(上海)有限公司 | Device for vector graphics processor to finish real-time drawing of graphics image without DDR |
CN111028360B (en) * | 2018-10-10 | 2022-06-14 | 芯原微电子(上海)股份有限公司 | Data reading and writing method and system in 3D image processing, storage medium and terminal |
TWI685772B (en) * | 2018-11-23 | 2020-02-21 | 國立臺灣大學 | Image operation method for acquisition of eye-tracking gaze position and system thereof |
CN112565585B (en) * | 2019-09-26 | 2021-12-24 | 珠海格力电器股份有限公司 | Image display circuit, image display method, and image display apparatus |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101013408A (en) * | 2007-02-15 | 2007-08-08 | 威盛电子股份有限公司 | Data processing system and data processing method |
CN100357913C (en) * | 2002-03-20 | 2007-12-26 | 精工爱普生株式会社 | Data transfer device, transfer program and transfer method of multidimensional memory |
CN102184720A (en) * | 2010-06-22 | 2011-09-14 | 上海盈方微电子有限公司 | A method and a device for image composition display of multi-layer and multi-format input |
CN102831571A (en) * | 2011-07-08 | 2012-12-19 | 图芯芯片技术(上海)有限公司 | Design method of five-order filter for realizing graphic image resizing and rotation in one step in flow-line manner |
CN103049407A (en) * | 2012-12-27 | 2013-04-17 | 华为技术有限公司 | System, device and method for data storage |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2005242497A (en) * | 2004-02-25 | 2005-09-08 | Matsushita Electric Ind Co Ltd | Data transfer method and data storage medium |
-
2013
- 2013-07-01 CN CN201310270956.9A patent/CN104281543B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN100357913C (en) * | 2002-03-20 | 2007-12-26 | 精工爱普生株式会社 | Data transfer device, transfer program and transfer method of multidimensional memory |
CN101013408A (en) * | 2007-02-15 | 2007-08-08 | 威盛电子股份有限公司 | Data processing system and data processing method |
CN102184720A (en) * | 2010-06-22 | 2011-09-14 | 上海盈方微电子有限公司 | A method and a device for image composition display of multi-layer and multi-format input |
CN102831571A (en) * | 2011-07-08 | 2012-12-19 | 图芯芯片技术(上海)有限公司 | Design method of five-order filter for realizing graphic image resizing and rotation in one step in flow-line manner |
CN103049407A (en) * | 2012-12-27 | 2013-04-17 | 华为技术有限公司 | System, device and method for data storage |
Also Published As
Publication number | Publication date |
---|---|
CN104281543A (en) | 2015-01-14 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104281543B (en) | Display controller and graphics accelerator is supported to access the framework method of internal memory simultaneously | |
DE112006003194B4 (en) | Integrate ad control into a low power processor | |
US11615284B2 (en) | Efficient transferring of human experiences to robots and other autonomous machines | |
US6078339A (en) | Mutual exclusion of drawing engine execution on a graphics device | |
KR101713177B1 (en) | System and method for virtual displays | |
US20070091098A1 (en) | Transparent multi-buffering in multi-GPU graphics subsystem | |
US9105112B2 (en) | Power management for image scaling circuitry | |
CN112104819B (en) | Multi-channel video synchronous switching system and method based on FPGA | |
DE102019109858A1 (en) | Width and frequency conversion with PHY layer devices | |
US20140132611A1 (en) | System and method for data transmission | |
WO2022007473A1 (en) | Display screen driving circuit and method, electronic device, and storage medium | |
CN108256643A (en) | A kind of neural network computing device and method based on HMC | |
CN114416635A (en) | Graphic image superposition display method and chip based on SOC | |
CN101236740B (en) | Display data transmission method and apparatus | |
US20230335049A1 (en) | Display panel fps switching | |
CN109688462A (en) | Reduce method, apparatus, electronic equipment and the storage medium of equipment power dissipation | |
CN106796535A (en) | Image display method, device, electronic equipment and computer program product | |
CN1318939C (en) | Method and device for handling requests for changing system mode and computer system | |
CN111694532B (en) | Display control method of single-chip heterogeneous system and wearable device | |
EP4229586A1 (en) | Optimizing gpu kernel with simo approach for downscaling utilizing gpu cache | |
CN103680402B (en) | A kind of asynchronous all-colour LED display control program based on LEON3 soft nucleus CPU and control method thereof | |
EP4241230A1 (en) | Methods and apparatus for rasterization of compute workloads | |
Śniegowski et al. | Vitrall: web-based distributed visualization system for creation of collaborative working environments | |
US11705091B2 (en) | Parallelization of GPU composition with DPU topology selection | |
US11373267B2 (en) | Methods and apparatus for reducing the transfer of rendering information |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
CB03 | Change of inventor or designer information |
Inventor after: Zhang Huiming Inventor after: Wang Zhenyu Inventor after: MIKE.CAI Inventor before: Zhang Huimin Inventor before: Wang Zhenyu Inventor before: MIKE.CAI |
|
COR | Change of bibliographic data | ||
GR01 | Patent grant |