USRE39529E1

Movatterモバイル変換

Info

Publication number: USRE39529E1
Application number: US09/536,646
Authority: US
Inventors: Koyo Katsura; Shinichi Kojima; Noriyuki Kurakami
Original assignee: Renesas Technology Corp
Current assignee: Renesas Technology Corp
Priority date: 1988-04-18
Filing date: 2000-03-28
Publication date: 2007-03-27
Anticipated expiration: 2009-01-27

Abstract

A Memory Interface and Video Attribute Controller (MIVAC) is inserted between a dynamic RAM (DRAM) capable of a consecutive data read operation, such as the operation associated with the static column mode, page mode, or nibble mode, and a graphic processor to provide a parallel data processing. A serial data transfer is executed on each data bus between the MIVAC and the DRAM, whereas parallel data transfer is conducted between the MIVAC and the graphic processor. As a result, the graphic processor can be configured with a reduced number of DRAMs so that the graphic processor operates without paying attention to the consecutive data read mode of the DRAM.

Description

This is a continuation of Reissue application Ser. No.07/985,141, filed Dec.3,1992 now U.S. Pat. No. RE37103.

BACKGROUND OF THE INVENTION

The present invention relates to a graphic processing apparatus for processing graphic data stored in a memory, and in particular, to a graphic processing apparatus in which the number of memories to be employed can be reduced so as to minimize the size of the processing apparatus.

For example, the Japanese Patent Publication JP-A-60-136793 describes a graphic processing apparatus in which characters and graphic data are generated in a display memory (frame buffer) so as to be delivered to output devices such as a display and a printer. In this conventional example, a high-speed graphic drawing operation is achieved by use of a method in which data bits constituting at least one pixel are packed in a word so as to be stored in the memory. In contrast with the prior method in which information of a pixel requires a plurality of words, this method allows accessing of the memory in the unit of a word (16 bits); in consequence, by packing information of a pixel in a single word, at least one pixel can be updated through one access, which therefore increases the processing speed.

In the conventional example above, although the memory is connected to a 16-bit data bus, the dynamic random access memory (DRAM) generally possesses a 1-bit or 4-bit data bus, and hence at least four to 16 memory elements are required, which prevents the apparatus from being miniturized.

In addition, the Japanese Patent Publication JP-A-60-225888 describes an apparatus including a dynamic random access memory (DRAM) having a nibble function (one of consecutive data read functions); however, description has not been given of a combination with a graphic processor in which data are accessed in a parallel fashion.

Moreover, in the Japanese Patent Publication JP-A-55-129387, there is described a system for transferring serial data between a processor and an external device; however, parallel data access is carried out between the processor and a memory.

SUMMARY OF THE INVENTION

It is therefore an object of the present invention to provide a small-sized graphic processing apparatus in which data transfer is enabled through a data bus having a reduced bit width so as to minimize the number of memory elements employed.

In order to achieve the object above, according to the present invention, there is disposed data converting means between processor means processing parallel data and a memory so as to enable the data bus width of the memory to be smaller than that of the processor means. The data converting means includes a latch for temporarily storing read data and a multiplexer for writing data. The present invention is characterized in that a memory having a successive data read function is applied to a processor effecting parallel data processing.

In the graphic processing apparatus according to the present invention, the memory is accessed in a time shared fashion such that data is converted by the converting means into parallel data. That is, in a data reading operation, data sequentially read out in a time shared fashion is temporarily stored in a latch so as to be supplied as parallel data to the processor. Moreover, in a data writing operation, parallel data supplied from the processor is sequentially written through the multiplexer into the memory in a time shared fashion.

BRIEF DESCRIPTION OF THE DRAWINGS

The present invention will be apparent from the following detailed description taken in conjunction with the accompanying drawings in which:

FIG. 1 is a schematic diagram showing an embodiment according to the present invention;

FIGS. 2,3a, and3b are diagrams for explaining a component of the embodiment ofFIG. 1;

FIG. 4 is a diagram schematically showing an internal configuration of the component;

FIGS. 5a,5b, and5c are explanatory diagrams showing in detail the embodiment ofFIG. 1;

FIGS. 6 and 7 are diagrams for explaining the embodiment ofFIG. 1;

FIGS. 8 to14 are explanatory diagrams useful for explaining operation modes;

FIGS. 15a to26 are detailed timing charts of the operation;

FIG. 27 is a diagram showing in detail the circuit configuration of the embodiment ofFIG. 1;

FIG. 28 is a diagram showing a gate circuit configuration; and

FIGS. 29a,29b, and29c are diagrams for explaining address outputs.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

Referring now to the drawings, description will be given of an embodiment according to the present invention.

FIG. 1 shows a configuration of a graphic processing apparatus according to the present invention. The graphic processing apparatus includes a graphic processor, namely, Advanced Cathode Ray Tube (CRT) Controller (ACRTC, Hitachi HD63484)10, a Memory Interface and Video Attribute Controller (MIVAC, Hitachi HD63487)20, aframe buffer30, a digital to analog converter with built-in color pallete (CPLT, Hitachi HD153108)40, and a CRT50. TheMIVAC20 produces various control signals and addresses necessary for theACRTC10 to access theframe buffer30. The MIVAC20 also generates 2CLK as a basic signal for the ACRTC10. Furthermore, the MIVAC20 has a function of converting parallel data from theframe buffer30 into serial data for video signals.

On receiving control signals (AS, MCYCDRAW, MRD, etc.) from the ACRTC10, the MIVAC20 initiates the read and write operations on theframe buffer30. In the operation, control signals includingRAS,CS,OE, andWE for the DRAM control are generated to be used in association with theframe buffer30. In addition, an address received from the ACRTC10 for theframe buffer30 is multiplexed so as to produce row/column addresses. By use of the static column mode, theMIVAC20 sequentially outputs a plurality of column addresses after a row address. In this embodiment, although the static column mode is adopted, it is also possible to use other sequential access mode (for example, a page mode, or a nibble mode) in combination therewith.

Read/write data is transferred between the ACRTC10 and theframe buffer30 through theMIVAC20.

In the display operation, parallel data read from theframe buffer30 is fetched into theMIVAC20 to be converted into serial data by means of a parallel/serial converter integrated therein, thereby producing digital video signals. These digital signals are converted by theCPLT40 into analog video signals so as to be displayed on theCRT50. In this embodiment, although the CRT50 is used as the output device, other output equipment, such as a printer, may also be employed.

FIG. 2 shows the pin arrangement of the MIVAC20. In this embodiment, the MIVAC20 is manufactured by use of the High performance Bipolar CMOS (Hi-BiCMOS) technology in which the technology of the CMOS of low power consumption, thereby implementing a high-speed and high-performance logic circuit of a relatively low power consumption. Since the MIVAC20 includes a Plastic Leaded Chip Carrier (PLCC) 68-pin package, surface mounting thereof is possible, which enables the mounting board of the graphic processing apparatus to be minimized.

FIGS. 3a and 3b show various interface signals of theMIVAC20. The input/output signals of theMIVAC20 are briefly classified into operation control signals for controlling operations thereof, interface signals with respect to theACRTC10, interface signals for theframe buffer30, and interface signals for thedisplay50.

Terminal INCLK of the operation control signals is used to receive a clock for the operation basis of theMIVAC20. The interface signals for the ACRTC10 include the 2CLK as the basic clock of theACRTC10, control signals MRD andDRAW for controlling the read and write operations, and signals on the address/data buses MAD0 to MAD15 and address buses MA16 to MA19. The interface signals for theframe buffer30 includeRAS,CS,OE, andWE as control signals of the DRAM and signals related to row/column address FA0 to FA9. The interface signals for thedisplay50 include digital video signals attained through parallel/serial conversion effected on display data and DOTCK produced by dividing INCLK.

FIG. 4 shows an internal configuration of theMIVAC20. In theMIVAC20, an attribute code definable by the user stored in the ACRTC10 is latched by means of anattribute code latch2011 so as to be decoded by aVCF decoder2012 into a signal, which enables various operation modes to be effected.

The INCLK as the basis of the operation of theMIVAC20 is divided by 2, 4, 8 16, and 32 by INCLK2006 and an INCLKdivider2009. The results are combined in a state decoder2007 to generate a timing signal, which is used in the respective logic circuits.

The 2CLK as the basic clock of theACRTC10 is produced from a2CLK generator2008. In the2CLK2008, in order to effect a plurality of read and write operations in the memory cycle, the first half cycle is shorter than the second half cycle, i.e., this signal has an asymmetric shape.

For the DOTCLK, a multiplex operation is achieved on the signals attained by dividing INCLK by 1, 2, and 4 by means of amultiplexer2010 to produce a multiplexed signal. Selection of the divided signals is automatically achieved depending on the operation mode of theMIVAC20.

The frame buffer address MAD0 to MAD15 and MA16 to MA19 supplied from theACRTC10 is temporarily latched in alatch2001 so as to be then multiplexed through amultiplexer2003 into a row/column address, thereby generating a ten-bit address associated with the frame buffer address signals FA0 to FA9. In addition, there is integrated acolumn address counter2002 such that the value of this counter and the latched address are multiplexed by themultiplexer2003, so that the resultant signal is adopted as a portion of the column address, thereby effecting several read/write operations in a memory cycle.

The control signals from theACRTC10 are latched in alatch2004. Depending onDRAW and MRD, the memory cycle is determined to be a draw read cycle, a draw write cycle, or a display cycle. WhenDRAW and MRD are respectively at low and high levels, namely, in the draw read cycle, the signalsRAS,CS, andOE, produced in the memory control2005, are delivered so as to read drawing data from the memory. Data obtained through several read operations in a cycle is temporarily latched in an input data latch2015 so as to be transferred therefrom to a readdata latch2016 to be latched again. The latched data is then outputted to the data buses MAD0 to MAD15 in accordance with the timing of the data fetch operation of theACRTC10 under control of theMA output control2000.

In addition, whenDRAW and MRD are both at a low level, namely, in the draw write cycle, the signalsRAS,CS, andWE, generated in the memory control2005, are supplied so as to write drawing data in the memory. The drawing data to be written is multiplexed by amultiplexer2014 disposed at an output stage including FD0 to FD7 in synchronism with the address which has undergone a counting operation by thecolumn address counter2002, so that the resultant multiplexed signals are written in the memory through several write operations effected at separate times under control of anFD output control2013.

WhenDRAW and MRD are both at the high level, namely, in the display read cycle, the data obtained through several read operations in a cycle is latched by the input data latch2015 used in the draw read cycle. Thereafter, the data is transferred to and is latched in adisplay data latch2019. In a case of a 4-chip memory configuration, since data is supplied through MAD8 to MAD15, the data is multiplexed by amultiplexer2017 so as to be transferred to the display data latch2019. The data is then sent to a shifter2020 and is latched by alatch20202 in the shifter2020 under the control of alatch control20201. The latched data is multiplexed by amultiplexer20204 in response to a clock signal produced from ashift clock generator20203 so as to convert the parallel data into serial data, thereby generating 4-bit video signals.

The video signal is skewed by a skew circuit2022 so as to be synchronized with the control signal from theACRTC10. For the video signal, a superimposing operation of a cursor can be achieved by use of acursor blink2023, or the video signals can be multiplexed through a multiplexer2024 in response to a signal attained by dividingVSYNC by two. The video signal after having undergone these processing operations is finally masked by use of theDISP signal so as to be produced as a 4-bit digital video signal. The signal used for the video mask is delivered as SHFTEN. In addition, the signal attained by dividingVSYNC by two is produced asVSYNC/2.

By using BLINK2 of the attribute codes, aBL 2IRQ/output section2021 generatesBL 2IRQ/ . When BLINK2 is set to “1”, “LOW” is supplied as theBL 2IRQ/ signal. When “Low” is inputted to theIRQCLR signal, theBL 2IRQ/ signal turns to “High”. The BLINK2 supplied from theACRTC10 outputs timing signals in which “1” and “0” are repeated for the predetermined number of fields.

FIGS. 5a,5b, and5c show connection methods for the frame buffers depending on the number of memories employed. In the case of a one chip memory configuration ofFIG. 5a, four data terminals of FD0 to FD3 of theMIVAC20 are connected to data terminals of aframe buffer300. Terminals related to FD4 to FD7 are not used. In this case, 4-bit data is transferred at one time between theMIVAC20 and theframe buffer300. In the draw read cycle, theMIVAC20 effects the 4-bit data read operation four times so as to transfer 16-bit data to theACRTC10. In the draw write cycle, 16-bit data from theACRTC10 is time-shared into four portions to be transferred to theframe buffer300 through four transfer operations. In the display read cycle, 4-bit data is read four times in a memory cycle or 16 times in two memory cycles so as to be fetched as 16-bit and 64-bit display data items, respectively.

In the case of a two chip memory configuration ofFIG. 5b, eight data terminals are used in association with FD0 to FD7 of theMIVAC20. In operation, data terminals of theframe buffer300 are connected to FD0 to FD3 and data terminals of theframe buffer301 are linked to FD4 to FD7. Between theMIVAC20 and the

frame buffers

300 and301, 8-bit data is transferred at one time. In the draw read cycle, theMIVAC20 reads 8-bit data twice so as to supply 16-bit data to theACRTC10. In the draw write cycle, 16-bit data from theACRTC10 is time-shared to be supplied to the

frame buffers

300 and301 through two transfer operations. In the display read cycle, 8-bit data is read out four times in a memory cycle or 16 times in two memory cycles so as to fetch 32-bit and 128-bit display data times, respectively. As a consequence, the operation can be applied to a CRT which has a higher operation speed as compared with the case of FIG.5a.

In the case of a four chip memory configuration ofFIG. 5c, the connections of the

frame buffers

300 and301 are the same as for the case of the two chip configuration ofFIG. 5b, the remaining two chips, namely,

frame buffers

302 and303 are connected to eight high-order bits of MAD8 to MAD15 selected from the data buses MAD0 to MAD15 between theACRTC10 and theMIVAC20. In the draw read cycle, theMIVAC20 read 16-bit data at a time. Eight-bit data read from the

frame buffers

300 and301 is outputted via theMIVAC20 to MAD0 to MAD7. Data containing the eight high-order bits read from the

frame buffers

302 and303 is transferred, without using theMIVAC20, directly via the buses MAD8 to MAD15 to theACRTC10. In the draw write cycle, data containing the eight low-order bits read from theACRTC10 is transferred through theMIVAC20 via the buses MAD0 to MAD7 to FD0 to FD7. Data containing the eight high-order bits is transferred, without using theMIVAC20, directly to the

frame buffers

302 and303. In the display read cycle, data containing eight low-order bits is read four times in a memory cycle via FD0 to FD7, whereas data containing eight high-order bits is read four times in a memory cycle via MAD8 to MAD15 such that the resultant 64-bit display data is fetched into theMIVAC20. In the display cycle effected in the circuit connection ofFIG. 5c, four addresses are outputted so as to execute four read operations as shown in FIG.29c. Data including eight low-order bits and data including eight high-order bits are respectively sent via FD0 to FD7 and MAD8 to MAD15 to the input data latch2015 (FIG. 4) so as to be latched therein. The input data latch2015 is of a length of 64 bits and hence 16 bits×4=64 bits are attained as display data.

In this mode, since the data buses are employed to input display data, it is impossible to effect a read operation in which 16 read operations are achieved in two memory cycles; however, when comparison is conducted in the read mode associated with four read operations per memory cycle, the operation above is applicable to a CRT which develops a higher processing speed as compared with the cases ofFIGS. 5a and 5b.

FIG. 6 shows video output timings in the respective cycle modes. TheACRTC10 has memory access modes including a single access mode in which the display cycle appears successively and a dual access mode in which high-speed drawing is possible. As shown inFIG. 6, in the single access mode, during a display period of time (whereDISP is “Low”), the display cycle continues successively without effecting the drawing cycle. In contrast, in the dual access mode, also during the display period, the display cycle and the drawing cycle appear alternately. In the single access mode, the drawing cycle is restricted to be effected during the fly-back or retrace period, whereas in the dual access mode, the fly-back period and a half portion of the display period can be used as the drawing cycle, which enables the drawing operation to be accomplished at a higher speed. In theMIVAC20, in addition to these access modes, there is a 2MCYC mode in which two display cycles of the single access mode are treated as a cycle so as to achieve 16 memory read operations. In the single access mode, data fetched in the first display cycle is displayed in the subsequent cycle. Data fetched in the second display cycle is displayed in the subsequent cycle. Thereafter, these operations are repeatedly achieved. Data obtained in the last display cycle is to be outputted in the next drawing cycle; however, since theDISP signal of theACRTC10 is supplied only during the display cycle period, the end portion ofDISP is elongated by a cycle in theMIVAC20 so as to use the signal as a mask signal. In the dual access mode, data of the first display cycle is delivered through two subsequent cycles. As a consequence, the end portion ofDISP is elongated by two cycles so as to produce a mask signal. In the 2MCYC mode, 16 data read operations are achieved in two cycles, and the video output is also supplied through two cycles.

FIG. 7 shows the output timing of the attribute codes delivered from theACRTC10. The attribute codes are information items arbitrarily defined by the user. The attribute code is fed to MAD0 to MAD15 and MA16 to MA19 of theACRTC10 while 2CLK and MCYC are both at the high level during the last refresh period. When the attribute code is fetched and is then decoded, the operation mode of theMIVAC20 is set.

FIG. 8 shows the setting of attribute codes in theMIVAC20. TheMIVAC20 uses MAD0 to MAD7, which are freely defined by the user, and MA18 and MA19, usages of which are predetermined for theACRTC10. Four bits of MAD0 to MAD3 are used to set the display color, the shift amount of the shift register, the access mode, the number of memories employed, and the division ratio of the DOTCLK. MAD4 and MAD5 are used to set the display color of the cursor. MAD6 sets the depth of the memory employed. MAD7 sets whether or not the video output is multiplexed. MA18 is used to set the blinking operation of the cursor. MA19 sets theBR 2IRQ/ output.

FIG. 9 shows 16 operation modes defined by the four bits MAD0 to MAD3 of FIG.8. The display color, the shift amount of the shift register, the access mode, the number of memories employed, and the division ratio of the DOTCLK are automatically determined by setting one of the 16 operation modes.

(1) For the display color (color/gradation), there can be specified a monochrome display represented by 1 bit/pixel, a four-color display expressed by 2 bits/pixel, and 16-color display represented by 4 bits per pixel. In the case of 1 bit/pixel, a word of the memory is loaded with information of 16 consecutive pixels in the horizontal direction. In the case of 2 bits/pixel, a word of the memory is loaded with information of 8 consecutive pixels in the horizontal direction, and in the case of 4 bits/pixel, a word of the memory is loaded with information of 4 consecutive pixels in the horizontal direction.

(2) The shift length of the shift register may be set to 4, 8, 16, or 32 bits.

(3) The access modes include a single access mode, a dual access mode in which high-speed drawing is possible, and a 2MCYC mode in which 16 display accesses are conducted in two memory cycles. In themodes0 to5, the single access mode is employed, whereas in themodes6 to C, the dual access mode is used. In the modes D to F, the 2MCYC mode is adopted.

(4) The number of memories selectable is 1, 2, or 4. For the memory, there is utilized a memory such as one having a static column mode in which a plurality of read/write operations can be accomplished in a cycle.

(5) DOTCLK is generated by dividing INCLK by 1, 2, and 4. The division ratios are determined according to the respective operation modes. Based on the frequency, the screen layout of the CRT is determined for each operation mode.

FIG. 10 shows frequencies of DOTCLK applicable to the respective operation modes. In the

modes

0,3,5,8, B, D, and F, the division ratio is one, that is, the output of DOTCLK is identical to INCLK. In the

modes

1,4,6,9, C, and E, the division ratio is two; whereas in the

modes

2,7, and A, the division ratio is 4 for the DOTCLK output.

FIG. 11 shows cursor display colors set by use of MAD4 (CUR0) and MAD5 (CUR1).

(1) When CUR1 and CUR0 are both 0

The four bits of video outputs VIDEOA to VIDEOD are set to 0, and hence a black cursor is displayed.

(2) When CUR1 is 0 and CUR0 is 1

The four bits of video outputs VIDEOA to VIDEOD are set to 1 and hence a white cursor is displayed.

(3) When CUR1 is 1 and CUR0 is 0

For the four bits of video outputs VIDEOA to VIDEOD, the respective colors are reversed on the display.

(4) When CUR1 and CUR0 are both 1

For the three bits of video outputs VIDEOA to VIDEOC, the respective colors are reversed on the display, whereas VIDEOD is kept unchanged.

FIG. 12 shows depths t be specified by MAD6 (VMD) for the memory elements employed. For VMD=0, the depth is set to 256 k×4 bits; for VMD=1, the depth is set to 1 M×4 bits for the memory.

FIG. 13 shows the settings of MAD7 (MUXEN) specifying whether the video outputs are to be multiplexed or not. When MUXEN is 0, the multiplex operation is not achieved. When MUXEN is 1 and VSYNC/2 is 0, the video outputs are not multiplexed. When MUXEN and VSYNC/2 are both 1, data of VIDEOC is delivered as VIDEOA and data of VODEOD is supplied as VIDEOB. This function is primarily adopted for a display equipment using a color shutter.

FIG. 14 shows the setting of MA18 (BLINK1) for the graphic cursor display. In the case of

BLINK1=0, the cursor is not displayed, whereas for

BLINK1=1, the cursor is displayed.

FIGS. 15a to26 shows detailed timing charts in the respective operation states.

FIGS. 15a and 15b show in detail timing of the draw read cycle in the case where one memory is employed.

FIGS. 16a and 16b show in detail timing of the draw read cycle in the case where two memories are employed.

FIGS. 17a and 17b show in detail timings of the draw read cycle in the case where four memories are employed.

FIGS. 18a and 18b show in detail timing of the draw write cycle in the case where one memory is employed.

FIGS. 19a and 19b show in detail timing of the draw write cycle in the case where two memories are employed.

FIGS. 20a and 20b show in detail timing of the draw write cycle in the case where four memories are employed.

FIGS. 21a and 21b show in detail timing of the display read cycle in the case where a memory or two memories are employed.

FIGS. 22a and 22b show in detail timing of the display read cycle in the case where four memories are employed.

FIGS. 23a and 23b show in detail timing of the display read cycle in the 2MCYC mode in the case where one memory or two memories are employed.

FIGS. 24a and 24b show in detail timing of theCS beforeRAS refresh cycle of the DRAM. The refresh operation is executed in a period where the horizontal synchronization signal HSYNC is at the low level.

FIG. 25 shows in detail the output timing, for the

division ratios

1, 2, and 4, of DOTCLK, VSYNC/2, VIDEOA to VIDEOD, and SHFTEN.

FIG. 26 shows in detail output timings ofBL 2IRQ/ .

FIG. 27 shows an exemplary configuration of a graphic processingapparatus including ACRTC10,MIVAC20, andDRAMs300 to303. A clock signal generated by theclock oscillator80 is supplied as INCLK of theMIVAC20. Anexternal circuit70 is utilized as an interface with the microprocessor (not shown in FIG.27), and aninterface circuit60 is used forHSYNC andVSYNC.

FIG. 28 shows a circuit example including an NAND gate. The configuration includes a bipolar transistor, an n-channel MOS transistor, and a p-channel MOS transistor. In a portion where the logic of the preceding stage is to be reflected, a CMOS of a low power consumption is employed, whereas in the output side of the succeeding stage, a bipolar transistor is used.

FIGS. 29a to29c show in detail addresses supplied by theMIVAC20 to the FA terminal. Cases of a one chip memory, a two chip memory, and a 4-chip memory are shown inFIGS. 29a to29c, respectively. Signals (NC0 to NC2 and WC0 to WC2) enclosed with broken lines inFIGS. 29a to29c are produced by thecolumn address counter2002. NC0 to NC2 are counters, each effective within a word, andbits1 to2 of the counter are used in the respective operation modes. WC0 to WC2 are word counters and are employed to generate a display address. The bit numbers of the address are not necessarily consecutive. This is because the bits are to be commonly used in the respective operation modes so as to configure the circuit of themultiplexer2003 as simple as possible.

As described above, according to the present invention, the data bus width of the memory can be minimized, and hence the size of the graphic processing apparatus can be reduced.

Claims

1. A graphic processing apparatus comprising:

memory means, including a plurality of memory locations in an array of columns, having corresponding column addresses, and rows, having corresponding row addresses, for storing data;

data processing means for specifying a row address in said memory means for retrieval of data from the memory locations at the different column addresses within the specified row of memory locations and processing of the retrieved data to generate graphic signals;

memory control means;

a memory data bus having m lines and interconnecting the memory means and the memory control means to transmit m bits of data in parallel therebetween, where m is an integer; and

a processor data bus having n lines and interconnecting the data processing means and the memory control means to transmit n bits of data in parallel therebetween, where n is an integer and n>m;

said memory control means including storage means for temporarily storing data received serially on said memory data bus from memory locations at different column addresses of the memory means row corresponding with the specified row address, and transmitting the temporarily stored data in parallel on said processor data bus to said data processing means for processing thereof to generate graphic signals.

2. A graphic processing apparatus comprising:

data processing means for specifying a row address in said memory means for writing of data in the memory locations at the different column addresses within the specified row of memory locations;

memory control means;

said memory control means including multiplexer means for multiplexing data received in parallel on said processor data bus into serial data and applying the serial data to said memory data bus for writing thereof in memory locations at different column addresses of the memory means row corresponding with the specified row address.

3. A graphic processing apparatus comprising:

data processing means for specifying a row address of memory locations in said memory means for transfer of a data word therewith;

memory control means;

a processor data bus having n lines and interconnecting the data processing means and the memory control means to transmit n bits of data in parallel therebetween, where n is a multiple of m;

said memory control means including counter means, responsive to receipt on said processor data bus of a row address specified by said processor means to specify an n-bit data word in said memory means, for successively generating n column addresses, applying the received row address and m of the generated column addresses on said memory data bus to transfer data between said memory means and said data processor means, with the data transfer including transfer of m bits of data in parallel between said memory means and said memory control means, and transfer of n bits of data between said memory control means and said data processor means.

4. A graphic processing apparatus comprising:

memory means, including a plurality of memory locations in an array of columns, having corresponding column addresses, and rows, having corresponding row addresses, for storing pixel information;

data processing means for specifying addresses of memory locations in said memory means for retrieval of pixel information therefrom and processing of the retrieved pixel information to generate graphic signals;

memory control means coupled to said memory means an said data processing means for retrieving pixel information from said memory means and applying the retrieved pixel information to said data processing means for processing thereof; and

output means connected to said memory control means for outputting processed pixel information to generate graphics.

5. A graphic processing apparatus as claimed inclaim 4, wherein the pixel information comprises multi-bit pixel information units corresponding to one pixel.

6. A graphic processing apparatus as claimed inclaim 4, wherein the pixel information comprises pixel information units, and wherein said memory control means includes means for selecting the number of bits in each pixel information unit.

7. A graphic processing apparatus as claimed inclaim 4, wherein said memory control means includes storage means for temporarily storing pixel information retrieved from said memory means.

8. A graphic processing apparatus comprising:

data processing means for specifying a row address in said memory means for transfer of data between the data processing means and the memory locations at the different column addresses within the specified row of memory locations;

memory control means;

said memory control means including storage means for temporarily storing data received on said memory bus from memory locations at different column addresses of the memory location row corresponding with the specified row address and transmitting the temporarily stored data in parallel on said processor data bus to said data processing means for processing thereof, and multiplexer means for multiplexing data received in parallel on said processor data bus into serial data and applying the serial data to said serial memory data bus for writing thereof in memory locations at different column addresses of the memory location row corresponding with the specified row address.

9. A graphic processing apparatus comprising:

a memory which stores graphic data;

a data processor which executes a predetermined graphic processing to generate graphic data to be stored in said memory;

a memory controller which controls data transfer between said memory and said data processor in accordance with a request from said data processor;

a digital to analog converter(DAC), connected to said memory controller, which outputs said graphic data read out from said memory;

a first bus, having m(wherein m is an integer)bits width, connected between said memory and said memory controller, which transfers m bits of data in parallel; and

a second bus, having n(wherein n is an integer, n>m)bits width, connected between said memory controller and said data processor, which transfers n bits of data in parallel;

wherein said memory controller comprises:

a storage which temporarily stores graphic data read out from said memory in successive groups of m bits of data during a predetermined period of time through said first bus,

a circuit which forms n bits of data using said successive groups of m bits of data and supplies said n bits of data in parallel to said data processor through said second bus based on an indication from said data processor, and

a converter which converts said graphic data temporarily stored in said storage into serial data which is provided to said DAC based on an indication from said data processor.

10. An apparatus according toclaim 9, wherein said memory controller further comprises:

a multiplexer which outputs the n bits graphic data transferred from said data processor to said first bus having m bits width in a time shared fashion.

11. An apparatus according toclaim 9, wherein said memory controller further comprises:

means for generating an address signal for accessing said memory plural times, in response to a signal for accessing said memory supplied from said data processor.

12. An apparatus according toclaim 9, wherein graphic data to be transferred to said memory controller through said first bus is read out from said memory plural times within a unit transfer time in a time shared fashion, based on an access signal to said memory designated by said data processor.

13. An apparatus according toclaim 12, wherein the graphic data transferred to said memory controller is supplied to said data processor through said second bus within a time longer than twice said unit transfer time.

14. A graphic processing apparatus comprising:

a memory which stores graphic data;

a data processor which executes predetermined graphic processing to generate graphic data;

a memory controller which controls transfer of data between said memory and said data processor in response to a request from said data processor;

a first bus having an m-bit width(wherein m is an integer)and connected between said memory and said memory controller, which transfers data of m bits in parallel; and

a second bus having an n-bit width(wherein n is an integer and n>m)and connected between said memory controller and said data processor, which transfers data of n bits in parallel;

wherein said memory controller concludes:

a storage which temporarily stores graphic data read out from said memory successively in a predetermined period of time via said first bus,

a circuit which applies said temporarily stored graphic data to said data processor as n-bit parallel data based on an indication from said data processor, and

a converter which converts said temporarily stored graphic data into serial data and outputs the serial data to said DCA based on an indication from said data processor.

15. A graphic processing apparatus according toclaim 14, wherein said memory controller includes a multiplexer which outputs n-bit graphic data transferred from said data processor on said first bus having the m-bit width successively in a time-sharing manner.

16. A graphic processing apparatus according toclaim 15, wherein said memory controller includes a second circuit which generates address signals for accessing said memory plural times with respect to a signal for accessing said memory means applied from said data processor.

17. A graphic processing apparatus according toclaim 15, wherein graphic data to be transferred to said memory controller via said first bus are successively read out plural times within a transfer unit time in a predetermined period of time on the basis of an access signal to said memory designated by said data processor.

18. A graphic processing apparatus according toclaim 17, wherein graphic data transferred to said memory controller are applied to said data processor via said second bus within a time period more than two times said transfer unit time.

19. A graphic processing apparatus according toclaim 14, wherein said memory controller includes a second circuit which generates address signals for accessing said memory plural times with respect to a signal for accessing said memory applied from said data processor.

20. A graphic processing apparatus according toclaim 14, wherein graphic data to be transferred to said memory controller via said first bus are successively read out plural times within a transfer unit time in a predetermined period of time on the basis of an access signal to said memory designated by said data processor.

21. A graphic processing apparatus according toclaim 20, wherein graphic data transferred to said memory controller are applied to said data processor via said second bus within a time period more than two times said transfer unit time.

22. A graphic processing apparatus comprising:

a memory which stores graphic data, said memory being accessed by using a row address and a column address;

a memory controller which controls data transfer of data between said memory and said data processor in response to a request from said data processor;

a second bus having an n-bit width(wherein n is an integer and n>m)and connected between said memory controller and said data processor, which transfers data of n bits in parallel; and

wherein said memory controller includes:

a first circuit which reads out a plurality of graphic data at different column addresses at a same row address from said memory via said first bus successively in a predetermined period of time,

a second circuit which applies said read-out graphic data to said data processor as n-bit parallel data based on an indication from said data processor, and

a converter which converts said read-out graphic data into serial data and outputs the serial data to said DAC based on an indication from said data processor.

23. A graphic processing apparatus according toclaim 22, wherein said memory controller includes a third circuit which successively generates a plurality of column addresses based on a signal for accessing said memory applied from said data processor.

24. A memory controller for controlling transference of data between a memory and a processor, said memory controller comprising:

m bit terminals for coupling to said memory, wherein successive groups of m bits of data is transferred through said m bit terminals between said memory and said controller by performing plural read operations within a memory cycle(where m is an integer);

n bit terminals for coupling to said processor, wherein n bits of data is transferred in parallel through said n bit terminals between said controller and said processor(where n is an integer and n>m);

a storage which temporarily stores graphic data read out from said memory in successive groups of m bits of data during a predetermined period of time through said m bit terminals;

a first circuit which forms n bits of data by combining successive groups of m bits of data from said m bit terminals and supplies said n bits of data in parallel to said n bit terminals based on an indication from said processor; and

a converter which converts said graphic data temporarily stored in said storage into serial data which is supplied to a digital to analog converter(DAC), said DAC being connected to said memory controller.

25. A memory controller according toclaim 24, wherein said successive groups of m bits of data from said m bit terminals are read out of said memory by performing plural read operations within a memory cycle based on an address specified by said processor.

26. A memory controller according toclaim 25, wherein said n bits of data is applied to said processor through said n bit terminals in a unit of time more than two times said memory cycle.

27. A memory controller according toclaim 24, wherein said successive groups of m bits of data each includes an m bit portion of said n bits of data.

28. A graphic processing apparatus comprising:

a memory which stores graphic data;

wherein said memory controller comprises:

at least one output terminal;

a converter which converts said graphic data temporarily stored in said storage into serial data which is provided to said at least one output terminal based on an indication from said data processor.

29. An apparatus according toclaim 28, wherein said memory controller further comprises:

30. An apparatus according toclaim 28, wherein said memory controller further comprises:

a second circuit which generates an address signal for accessing said memory plural times, in response to a signal for accessing said memory supplied from said data processor.

31. An apparatus according toclaim 28, wherein graphic data to be transferred to said memory controller through said first bus is read out from said memory plural times within a unit transfer time in a time shared fashion, based on an access signal to said memory designated by said data processor.

32. An apparatus according toclaim 31, wherein the graphic data transferred to said memory controller is supplied to said data processor through said second bus within a time longer than twice said unit transfer time.

33. A graphic processing apparatus comprising:

a memory which stores graphic data;

a second bus having an n-bit width(wherein n is an integer and n>m)and connected between said memory controller and said data processor, which transfers data of n bits in parallel,

wherein said memory controller includes:

at least one output terminal;

a converter which converts said temporarily stored graphic data into serial data and outputs the serial data to said at least one output terminal based on an indication from said data processor.

34. A graphic processing apparatus according toclaim 33, wherein said memory controller includes a multiplexer which outputs n-bit graphic data transferred from said data processor on said first bus having the m-bit width successively in a time-sharing manner.

35. A graphic processing apparatus according toclaim 34, wherein said memory controller includes a second circuit which generates address signals for accessing said memory plural times with respect to a signal for accessing said memory means applied from said data processor.

36. A graphic processing apparatus according toclaim 34, wherein graphic data to be transferred to said memory controller via said first bus are successively read out plural times within a transfer unit time in a predetermined period of time on the basis of an access signal to said memory designated by said data processor.

37. A graphic processing apparatus according toclaim 36, wherein graphic data transferred to said memory controller are applied to said data processor via said second bus within a time period more than two times said transfer unit time.

38. A graphic processing apparatus according toclaim 33, wherein said memory controller includes a second circuit which generates address signals for accessing said memory plural times with respect to a signal for accessing said memory applied from said data processor.

39. A graphic processing apparatus according toclaim 33, wherein graphic data to be transferred to said memory controller via said first bus are successively read out plural times within a transfer unit time in a predetermined period of time on the basis of an access signal to said memory designated by said data processor.

40. A graphic processing apparatus according toclaim 39, wherein graphic data transferred to said memory controller are applied to said data processor via said second bus within a time period more than two times said transfer unit time.

41. A graphic processing apparatus comprising:

wherein said memory controller includes:

at least one output terminal;

a converter which converts said read-out graphic data into serial data and outputs the serial data to said at least one output terminal based on an indication from said data processor.

42. A graphic processing apparatus according toclaim 41, wherein said memory controller includes a third circuit which successively generates a plurality of column addresses based a signal for accessing said memory applied from said data processor.

43. A memory controller for controlling transference of data between a memory and a processor, said memory controller comprising:

a converter which converts said graphic data temporarily stored in said storage into serial data which is supplied to at least one output terminal, said at least one output terminal being connected to said memory controller.

44. A memory controller according toclaim 43, wherein said successive groups of m bits of data from said m bit terminals are read out of said memory by performing plural read operations within a memory cycle based on an address specified by said processor.

45. A memory controller according toclaim 44, wherein said n bits of data is applied to said processor through said n bit terminals in a unit of time more than two times said memory cycle.

46. A memory controller according toclaim 43, wherein said successive groups of m bits of data each includes an m bit portion of said n bits of data.