WO2008072354A1 - コンパイルドメモリ、asicチップおよびコンパイルドメモリのレイアウト方法 - Google Patents
コンパイルドメモリ、asicチップおよびコンパイルドメモリのレイアウト方法 Download PDFInfo
- Publication number
- WO2008072354A1 WO2008072354A1 PCT/JP2006/325093 JP2006325093W WO2008072354A1 WO 2008072354 A1 WO2008072354 A1 WO 2008072354A1 JP 2006325093 W JP2006325093 W JP 2006325093W WO 2008072354 A1 WO2008072354 A1 WO 2008072354A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- memory
- compiled
- control unit
- word
- memory block
- Prior art date
Links
Classifications
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11C—STATIC STORES
- G11C5/00—Details of stores covered by group G11C11/00
- G11C5/06—Arrangements for interconnecting storage elements electrically, e.g. by wiring
- G11C5/063—Voltage and signal distribution in integrated semi-conductor memory access lines, e.g. word-line, bit-line, cross-over resistance, propagation delay
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11C—STATIC STORES
- G11C11/00—Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor
- G11C11/21—Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using electric elements
- G11C11/34—Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using electric elements using semiconductor devices
- G11C11/40—Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using electric elements using semiconductor devices using transistors
- G11C11/41—Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using electric elements using semiconductor devices using transistors forming static cells with positive feedback, i.e. cells not needing refreshing or charge regeneration, e.g. bistable multivibrator or Schmitt trigger
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11C—STATIC STORES
- G11C11/00—Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor
- G11C11/21—Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using electric elements
- G11C11/34—Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using electric elements using semiconductor devices
- G11C11/40—Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using electric elements using semiconductor devices using transistors
- G11C11/41—Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using electric elements using semiconductor devices using transistors forming static cells with positive feedback, i.e. cells not needing refreshing or charge regeneration, e.g. bistable multivibrator or Schmitt trigger
- G11C11/413—Auxiliary circuits, e.g. for addressing, decoding, driving, writing, sensing, timing or power reduction
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11C—STATIC STORES
- G11C8/00—Arrangements for selecting an address in a digital store
- G11C8/10—Decoders
Definitions
- the present invention relates to a compiled memory formed in a chip such as an ASIC and a layout method of the compiled memory.
- LSIs such as ASIC (Application Specific IC) may be equipped with memory such as SRAM accessed by functional blocks in the ASIC chip.
- This type of memory often has a variable storage capacity depending on the user's system specifications.
- Layout data for a memory having an arbitrary storage capacity specified by the user is generated using a layout design tool such as a compiler.
- the memory is automatically generated by inputting an arbitrary number of bits and words into the compiler.
- Memory that is automatically generated by the compiler is called compiled memory or compiled.
- a technique in which the bit line is composed of a local bit line and a global bit line has been proposed.
- the local bit line is wired in the memory cell array and directly connected to the memory cell.
- the global bit line is connected to a local bit line on the memory cell array and connected to a data input / output circuit.
- Patent Document 1 Japanese Patent Laid-Open No. 2006-32577
- the local bit line Since the local bit line is wired corresponding to the memory cell, the local bit line is wired near the semiconductor substrate and using a thin / thin wiring layer, and therefore has a high wiring resistance. Also, since the local bit line is connected to a large number of memory cells, the parasitic capacitance is large. For this reason, the length of the local bit line Greatly affects the memory access time. On the other hand, since the Guronobit line is wired using a thick wiring layer away from the semiconductor substrate, the wiring resistance is low. Since the global bit line is not connected to the memory cell, the parasitic capacitance is small. Therefore, the influence of the length of the global bit line on the access time of the memory is smaller than that of the local bit line. Therefore, in a memory having a hierarchical bit line structure, it is important to shorten the length of the local bit line directly connected to the memory cell in order to shorten the access time.
- An object of the present invention is to shorten the length of a bit line connected to a memory cell and shorten the access time of compiled memory. In particular, it shortens the access time of compiled memory formed in LSIs such as ASIC chips.
- a compiled memory has a pair of memory blocks, a data control unit, a connection control unit, and a decoder unit.
- Each memory block has a word group composed of at least one word line, a memory cell connected to the word line, and a bit line connected to each memory cell.
- the data control unit inputs / outputs a data signal to / from the memory cell via the bit line.
- the connection control unit is provided corresponding to each memory block in order to selectively connect the bit lines of each memory block to the data control unit.
- the decoder unit is used to access the memory block to be accessed.
- the address signal is decoded to select whether the word group is!
- the logic of the decoder unit is configured by assigning bits of an address signal for identifying a memory block and a connection control unit to a lower order than bits of an address signal for identifying a word group. For example, when each word group is composed of a plurality of word lines, the logic of the decoder unit uses the bits of the address signal for identifying the word lines in each word loop as the addresses for identifying the memory blocks. It is configured by allocating lower than the signal bits. As a result, the number of word lines arranged in the memory block can be made equal to each other, and the length of the bit line can be shortened.
- layout data is generated using the following layout method. First, an input of the number of bits of the data signal and the number of words indicating the number of word lines connected to the memory cell is received. Number of input words When the maximum number of words that can be placed in a memory block is exceeded, the specified number of words is evenly distributed so that it is smaller than the maximum number of words, and an even number of memory blocks are laid out. Data is generated. Next, the bit of the address signal for identifying the memory block is assigned lower than the bit of the address signal for identifying the word group, and the logic of the decoder section for selecting the memory block and the word line is selected. Is generated. Then, according to the generated logic, contact portions for connecting the wirings to each other are arranged at the intersections of the decode signal lines wired in the decoder unit, and the layout data is completed.
- the access time of the compiled memory can be shortened.
- FIG. 1 is a block diagram showing a first embodiment of the present invention.
- FIG. 2 is a circuit diagram showing details of a main decoder of the compiled memory shown in FIG.
- FIG. 3 is a circuit diagram showing details of the memory block shown in FIG. 1.
- FIG. 4 is a circuit diagram showing a main part of a compiled memory before the present invention is made.
- FIG. 5 is a circuit diagram showing a main part of a compiled memory examined before the present invention is made.
- FIG. 6 is an explanatory diagram showing the relationship between the number of word lines laid out in the compiled memory and the access time.
- FIG. 7 is a flowchart showing a compiled memory layout method according to the first embodiment.
- FIG. 8 is a circuit diagram showing details of a main decoder in a second embodiment of the present invention.
- FIG. 9 is a circuit diagram showing details of a main part in a third embodiment of the present invention.
- FIG. 10 is a circuit diagram showing a compiled memory according to a fourth embodiment of the present invention.
- FIG. 11 is an explanatory diagram showing the logic of the main decoder of the fourth embodiment.
- FIG. 12 is a flowchart showing a compiled memory layout method according to the fourth embodiment.
- FIG. 13 is a circuit diagram showing a compiled memory according to a fifth embodiment of the present invention.
- FIG. 14 is an explanatory diagram showing the logic of the main decoder of the fifth embodiment.
- FIG. 15 is a circuit diagram showing a compiled memory according to a sixth embodiment of the present invention.
- FIG. 16 is an explanatory diagram showing the logic of the main decoder of the sixth embodiment.
- FIG. 1 shows a first embodiment of the present invention.
- the compiled memory CM is an SRAM core, for example, and is formed in the ASIC chip together with functional blocks such as the controller CNTL and the peripheral circuit PERI in order to hold data used by the controller CNTL.
- Compiled memory CM is accessed from controller CNTU such as CPU.
- controller CNTU such as CPU.
- the compiled memory is greatly drawn in order to make it easy to distribute force.
- many functional blocks other than those shown are mounted on the chip.
- the ASIC chip configures the system alone or with other semiconductor chips.
- the compiled memory CM is a sense having a sense amplifier SA shared by a pair of memory blocks MBLKO-1 and connection control units CCNTO-1 and memory blocks MBLKO-1 corresponding to the memory blocks MBLK 0-1 respectively. It has an amplifier area SAA, main decoder MD (decoder section), data input / output section IZO, and timing control section CPG.
- the local bit line pair LBL, ZLBL of each memory block MBLKO-1 is connected to the connection control unit C global bit line pair GBL, ZGBL.
- the global bit line pair GBL, / GBL is connected to the data input / output I / O.
- the main decoder MD decodes the address signal ADR (Fig. 2) output from the controller CNTL, outputs column selection signals CO and C1 to turn on the column switch according to the decoding result, and outputs the memory block MBLKO. — Select the word line WL ( Figure 2) routed in 1.
- the data input / output unit IZO outputs the read data on the global bit line pair GBL, ZGBL to the controller CNTL via the ASIC chip data bus, and is transmitted from the controller CNTL via the ASIC chip data bus. Write data to global bit line pair GBL and ZGBL.
- the sense amplifier SA and the data input / output unit IZO operate as a data control unit that transmits a data signal to the memory cell MC (FIG.
- the timing control unit CPG controls the operation of the main decoder MD, data input / output unit IZO, sense amplifier SA, etc. according to the access command output from the controller CNTL to access the compiled memory CM (timing Signal).
- the storage capacity of the compiled memory CM is designed according to the data size used by the controller CNTL. That is, the storage capacity of the compiled memory CM (MBLK0-1 size) is determined according to the user specifications of the ASIC chip. As will be described later, in the present invention, layout design of the main decoder MD is performed so that the sizes of the memory blocks MBLK0-1 are equal to each other. As a result, the lengths of the local bit lines LBL and ZLBL can always be the shortest regardless of the storage capacity (user specification) of the compiled memory CM. As a result, the access time can be minimized.
- FIG. 2 shows details of the main decoder MD shown in FIG.
- the main decoder MD is configured to connect the logic circuits LA and LB and the mesh-like wiring and wirings connecting the logic circuits LA and LB.
- a contact part CONT arranged in the difference part.
- the position of the contact part CONT is determined by the layout method described later (Fig. 7).
- the logic circuit LA decodes the address signal ADR (ADR0-2), and generates a block decode signal BAD0-1 and a word decode signal WAD0-2.
- the logic circuit LB outputs the word line signal WL (WL0 5) according to the decode signal lines BAD0-3 and WAD02.
- the column selection signal CO-1 is output in synchronization with the clock CK output from the timing control unit CPG in response to the access command.
- the memory blocks MBLK0-1 are laid out with a word group WLG (word step) having at least one word line WL as one unit.
- WLG word step
- one bottle A word group WLG is formed by the line WL. Therefore, the number of each word group WLG 0-5 is the same as the number of the word line WLO-5.
- the memory block MBLKO-1 and the connection control unit CCNTO-1 are identified by the least significant bit ADRO of the address signal ADR.
- the word line WL (word group WLG) in each memory block MBLKO-1 is identified by the upper bits AD R1-2 of the address signal ADR. That is, the logic of the main decoder MD is lower in the bits of the address signal ADR for identifying the memory block MB LKO-1 and the connection control unit CCNTO-1 than the bits of the address signal ADR for identifying the word group WLG. It is composed by assigning to.
- the number of words is set in the range of “4” to “1024”, for example. For example, if the maximum number of words that can be placed in the memory block MBLK is "512" and the number of words is set to "600", the number of words in the memory block MBLKO-1 is set to "300" respectively. Is done. Memory block MBLKO-1 is identified by address signal ADRO (least significant bit). The word line WL in each memory block MBLKO-1 is identified by the address signal ADR1-9. On the other hand, when the maximum number of words is “256” and the number of words is set to “600”, four memory blocks MBLKO-3 are laid out as shown in FIGS.
- ADRO least significant bit
- the number of words in each memory block MBLKO-3 is set to “150”.
- the length of the local bit lines LBL and / LBL orthogonal to the word line WL can be minimized.
- the length of the local bit line pair LBL, ZLBL is LO for both the memory block MBLKO-1.
- FIG. 3 shows details of the memory block MBLKO-1 shown in FIG.
- Each memory block MBLKO-1 is arranged in a matrix and has a plurality of static memory cells MC connected to complementary local bit lines LBL, / LBL and word line WL.
- Each memory cell MC has a latch having a pair of inverters and a pair of transfer transistors (nMOS transistors) each having one end connected to a complementary storage node (output node of each inverter) of the latch. have.
- the other end of the transfer transistor is connected to either the local bit line LBL or / LBL.
- the gate of the transfer transistor is connected to the word line WL.
- the column switch formed in the connection control unit CNT0-1 is configured by an nMOS transistor, and is turned on when the corresponding column selection signal CO or C1 is at a high logic level.
- the sense amplifier area SAA has a precharge circuit (not shown) for precharging the local bit line pair LBL, / LBL to a predetermined voltage while the memory block MBLKO-1 is not being accessed.
- a circuit for controlling input of a write data signal may be formed in the sense amplifier area SAA.
- the sense amplifier SA is shared by the memory block MBLKO-1.
- the sense amplifier SA amplifies the signal amount of the data signal read from the memory cell MC to the local bit line pair LBL, / LBL during the read access operation, and the amplified data signal is amplified to the global bit line pair GBL, Tell ZGBL.
- the write data signal supplied from the outside of the compiled memory CM via the Grono bit line pair GBL, ZGBL is connected to the local bit line pair LBL, / LB L is transmitted to the memory cell MC connected to the selected word line WL.
- the sense amplifier S A also operates.
- FIG. 4 shows a main part of the compiled memory CM before the present invention is made.
- the maximum number of words that can be placed in each memory block MBLK (MBLKO-1) is described as “4”.
- the connected memory CM connected the word line WL to the bottom of the figure. They were laid out by arranging them sequentially from the side.
- the length LO of the local bit lines LBL, / LBL (not shown) of the memory block MBLKO is always maximized.
- the length L1 of the local bit lines LBL, ZLBL (not shown) of the memory block MBLK1 is shorter than the length LO.
- the access time (timing specification) of the compiled memory CM is determined based on the worst operating time. Therefore, the access time is determined according to the maximum length (LO) of local bit lines LBL and ZLBL.
- the access time of the compiled memory CM composed of a plurality of memory blocks MBLK is set according to the maximum length of the local bit lines LBL and ZLBL regardless of the number of words. Is done.
- the access time is the time from when the read access command and the address signal are supplied to the connected memory CM until the read data is output from the connected memory CM.
- the access time is the time from when the write access command, the write address signal and the write data are supplied to the compiled memory CM and written to the memory cell.
- FIG. 5 shows a main part of the compiled memory CM studied before the present invention is made.
- the lengths of the local bit lines LBL and ZLBL of the memory block MBL KO—1 are made equal and the lengths of the local bit lines LBL and / LBL are shortened. That is, the number of node lines WL arranged in the memory block MBLKO-1 may be made equal. Therefore, word line WLO-2 in memory block MBLKO And word lines WL3-5 are arranged in the memory block MBLK1.
- the compiled memory CM malfunctions.
- FIG. 6 shows the relationship between the number of word lines WL (number of words) laid out in the compiled memory CM and the access time tAAC.
- the solid line in the figure shows the state after the application of the present invention, and the broken line in the figure shows the state before the application of the present invention.
- the maximum number of words that can be placed in the memory block MBLK is "512".
- the length of the local bit lines LBL and ZLBL is “1024”. Until the number of word lines WL increases by 2, the number increases.
- the access time tAAC increases every time the number of word lines WL increases by two until the number of words reaches “1024”.
- the word line WL is arranged for each memory block MBLK. For this reason, the length of the local bit lines LBL, / LBL increases each time the word line WL increases until the number of words reaches “512”. That is, the access time tAAC increases every time the word line WL force S1 increases until the number of words reaches “512”.
- the access time tAAC (specification) of the compiled memory CM at this time is T1 corresponding to the length of the local bit lines LBL and / LBL crossing the 300 word lines WL.
- 512 word lines WL are arranged in the memory block MBLKO, and the remaining 88 word lines WL are arranged in the memory block MBLK1.
- the access time tAAC (specification) of the compiled memory CM at this time is T2 corresponding to the length of the local bit lines LBL and ZLBL crossing the two word lines WL.
- the TO of the access time tAAC is a time used for a decoder or an input / output circuit that always requires a constant operation time regardless of the number of word lines WL.
- the word lines WL are always alternately arranged in the pair of memory blocks MBLKO-1.
- the access time tAAC increases every time two word lines WL increase. Therefore, the access time tAAC is smaller than that before the present invention even when the number of words is less than “512”. Shortened.
- FIG. 7 shows a layout method of the compiled memory CM according to the first embodiment.
- the flow shown in the figure is realized when the layout design tool of the compiled memory CM executes the layout program.
- the layout design tool is a computer such as a workstation, and the layout program (compiler) is executed by a controller such as a built-in CPU.
- the layout data of the compiled memory CM is generated by executing the flow shown in the figure.
- the layout design flow description of the circuit on the path through which the data signal is transmitted is omitted.
- the compiler receives input of the number of bits BLN of the data signal, which is the specification of the compiled memory CM, and the number of words WLN indicating the number of word lines WL.
- the word number WLN is specified as 1024.
- Number of bits When BLN is specified as 1 28 bits, the number of words WLN is specified as 512 bits.
- the number of bits BLN may be specified separately for the number of data terminals and the number of memory cells MC (bit lines LBL) connected to the word line WL. In this case, the number of bit lines LBL is treated as the number of bits BLN.
- step S12 it is determined whether or not the input word number WLN force exceeds the maximum word number WLNmax that can be arranged in one memory block MBLK. If the number of words WLN exceeds the number of words WLNmax, the process proceeds to step S16. If the number of words WLN is less than or equal to the number of words WLNmax, the process proceeds to step S14.
- step S 14 a pair of memory blocks MBLKO—1 is generated, and a word line WL that is half the number of words WLN is arranged in each memory block MBLKO—1.
- step S16 the word number WLN is divided by an even value (2, 4, 6,%) Until a word number WLN1 smaller than the word number WLNmax appears. Then, it is used as the number of obtained even value memory blocks MBLK, and the obtained word line WLN1 is arranged in each memory block MBLK. That is, the specified number of words WLN is evenly distributed so as to be smaller than the maximum number of words WLNmax, and layout data of even number of memory blocks MBLK is generated.
- step S18 a common sense amplifier is used between the pair of memory blocks MBLK.
- Area SAA is allocated and a sense amplifier SA is arranged.
- step S20 the connection control unit CCNT is arranged between the memory block MBLK and the sense amplifier area SAA.
- step S22 the number of bits of the address signal ADR necessary to identify the word line WL is obtained, and the bit MBLKb for identifying the memory block MBLK and the word line WL (word group WLG) is assigned in order from the lower bit of the bit WLGb and the power address signal ADR. Then, the logic of the main decoder MD is generated. Bits MBLKb and WLGb correspond to ADRO and ADR1 (or ADR2) shown in Figure 2, respectively.
- step S 24 according to the generated logic of the main decoder MD, the contact part CONT is arranged at the intersection of the mesh-like decode signal lines wired in the main decoder MD.
- step S26 the output of the main decoder MD is connected to the word line WL and the column switch of the connection control unit CCNT. Then, the layout data of the compiled memory CM is completed.
- the bit ADRO of the address signal ADR for identifying the memory block MBLK is the bit ADR1—2 of the address signal ADR for identifying the word line WL (word group WLG). Assigned lower.
- the number of word lines WL arranged in the memory block MBLK can be made equal to each other, and the lengths of the local bit lines LBL, / L BL can be shortened.
- the wiring delay of the local bit lines LBL and ZLBL can be minimized, and the access time tAAC of the compiled memory CM can be shortened.
- FIG. 8 shows details of the main decoder MD of the compiled memory CM in the second embodiment of the present invention.
- the same elements as those described in the first embodiment are denoted by the same reference numerals, and detailed description thereof will be omitted.
- the compiled memory CM is, for example, an SRAM core, and is formed in an A SIC chip together with other functional blocks as shown in FIG.
- a system is configured by an ASIC chip alone, or a system is configured with other semiconductor chips.
- the number of words is specified as “8” by the user specification, and eight word lines WL are arranged in the compiled memory CM.
- Each memory block MBLKO—1 has a word line WL Four are arranged alternately.
- the length of the local bit line pair LBL, ZLBL can be set to LO for both memory blocks MBLKO-1 and the access time tAAC can be minimized.
- the contact lines CONT indicated by the thick wiring and the large black circles shown in the figure indicate the elements added to Figure 2 above.
- the layout data of the compiled memory CM can be easily generated by adding the signal line wiring of the main decoder MD and the contact part CONT even when the designated number of words is different. it can.
- the layout method of the compiled memory CM is the same as the flow shown in FIG.
- the same effect as in the first embodiment described above can be obtained. Furthermore, even when the number of words is changed, the layout data of the connected memory CM can be easily generated by adding or deleting the wiring and the contact part CONT.
- FIG. 9 shows the details of the main parts in the third embodiment of the present invention.
- the same elements as those described in the first embodiment are denoted by the same reference numerals, and detailed description thereof will be omitted.
- the compiled memory CM is, for example, an SRAM core, and is formed in the ASIC chip together with other functional blocks as shown in FIG.
- a system is configured by an ASIC chip, or a system is configured with other semiconductor chips.
- the layout method of compiled memory CM is the same as the flow shown in Fig.7.
- the compiled memory CM has a dummy area DMY for optimizing the operation timing of the sense amplifier SA.
- the sense amplifier SA is a circuit that processes a read data signal from the memory cell MC, and operates as a data processing unit.
- dummy memory cells DMC, dummy local bit line pairs DLBL, / DLBL, dummy sense amplifier DSA, and dummy column switches DC0, DC1 are arranged in the dummy area DMY.
- the layout data and arrangement interval of the dummy memory cell DMC are the same as those of the real memory cell MC.
- the dummy memory cell DMCref has a latch storage node connected to the dummy local bit line / DLBL connected to the ground line.
- the gate of the transfer transistor of the dummy memory cell DMCref is an access to the memory cell MC of the compiled memory CM.
- the level of the dummy local bit line ZDLBL (dummy data signal) changes to a low level in synchronization with the active signal of the active signal ACT.
- the dummy local bit line pair DLBL, ZDLBL is precharged to a high level by the precharge circuit in the same manner as the memory cell MC before the active signal ACT is activated.
- the column switch DCO connected to the dummy local bit lines DLBL, / DLBL of the memory block MBLKO is always turned on by receiving the power supply voltage VDD at the gate.
- the column switch DC 1 connected to the dummy local bit lines DLBL, / DLBL of the memory block MBLK1 is always turned off by receiving the ground voltage VSS at the gate.
- the dummy sense amplifier DSA has, for example, a dummy local bit line / input via the column switch DCO.
- the dummy sense amplifier DSA outputs a sense amplifier enable signal SEN for starting the amplification operation of the sense amplifier SA in synchronization with the activation of the active signal ACT, that is, the output timing of the dummy data signal.
- the sense amplifier enable signal SEN is supplied to all sense amplifiers SA.
- the operation timing of the sense amplifier SA does not depend on the position of the memory cell MC to be accessed (the length of the local bit lines LBL and ZLBL) and needs to be set optimally. In other words, the operation timing of the sense amplifier SA needs to be set according to the memory cell MC having the slowest access speed. Therefore, the period from the activation of the active signal ACT to the output of the sense amplifier enable signal SEN is set using the dummy memory cell DMCref farthest from the dummy sense amplifier DSA.
- the dummy memory cell DMCref is also referred to as a self timing cell. This kind of timing setting method is called a self-timing method.
- the operation timing of the sense amplifier SA is set in accordance with the operation timing of the memory cell MC (DMCref) farthest from the sense amplifier area SAA.
- the access time tAAC is the same as the broken line shown in FIG.
- the access time tAAC is It can be the same as the solid line shown in Figure 6.
- the same effect as in the first embodiment described above can be obtained. Furthermore, the access time tAAC can be shortened even in a compiled memory CM that employs the self-timing method.
- FIG. 10 shows a connected memory CM according to the fourth embodiment of the present invention.
- the same elements as those described in the first embodiment are denoted by the same reference numerals, and detailed description thereof will be omitted.
- the compiled memory CM is an SRAM core, for example, and is formed in the ASIC chip together with other functional blocks as shown in FIG.
- a system is configured by an ASIC chip alone, or a system is configured with other semiconductor chips.
- a word group WLG (WLGO-5) is formed for every four word lines WL.
- the word line WL (WLO-23) is arranged with the word group WLG as one unit.
- Other configurations are the same as those in the first embodiment except that the number of word lines WL is different.
- FIG. 11 shows the logic of the main decoder MD of the fourth embodiment.
- Memory block MBLKO-1 and connection control unit CCNTO-1 are identified by bit ADR2 of address signal ADR.
- the word group WLG in each memory block MBLKO-1 is identified by the upper bits ADR3-4 of the address signal ADR.
- the word line WL in each word group WLG is identified by the lower bit ADRO-1 of the address signal ADR.
- the logic of the main decoder MD is based on the bit of the address signal ADR for identifying the memory block MBLKO-1 and the connection control unit CCNTO-1 and the address signal for identifying the word group W LG. It is configured by assigning it lower than the ADR bit. Further, the logic of the main decoder MD assigns the bit of the address signal ADR for identifying the word line WL in each word group WLG to a lower order than the bit of the address signal ADR for identifying the memory block MBLKO—1. It is constituted by. As the address signal ADR0—4 advances in sequence, the memory block MBLKO—1
- Loop WLG is selected and accessed alternately as a unit.
- FIG. 12 shows a layout method of the compiled memory CM according to the fourth embodiment. This In this embodiment, step S22a is performed instead of step S22 in FIG. The rest of the flow is the same as in Figure 7.
- step S22a the number of bits of the address signal ADR necessary for identifying the word line WL is obtained.
- the bit WLb for identifying the word line WL and the memory Bit MBLKb for identifying block MBLK and bit WLGb for identifying word group WLG are allocated.
- the logic of the main decoder MD is generated. Bits WLb, MBLKb, and WLGb correspond to ADRO-1, ADR2, and ADR4-5 shown in Figure 11, respectively.
- the same effect as in the first embodiment described above can be obtained. Furthermore, even when the word group WLG is composed of a plurality of word lines WL, layout data of the compiled memory CM that can shorten the access time tAAC can be easily generated.
- FIG. 13 shows a connected memory CM according to the fifth embodiment of the present invention.
- the same elements as those described in the first embodiment are denoted by the same reference numerals, and detailed description thereof will be omitted.
- the compiled memory CM is an SRAM core, for example, and is formed in the ASIC chip together with other functional blocks as shown in FIG.
- a system is configured by an ASIC chip alone, or a system is configured with other semiconductor chips.
- the layout method of the compiled memory CM is the same as the flow shown in Fig.7.
- the compiled memory CM is composed of a memory unit MUO-1 composed of a pair of memory blocks MBLK (MBLK 0-1 or MBLK2-3).
- the compiled memory CM has four memory blocks MBLKO-3.
- Each memory unit MUO-1 has the same configuration as the memory block pair shown in FIG.
- Each memory block MBLKO-3 is composed of three word lines WL (word group WLG).
- the memory block MBLK2 is connected to the sense amplifier SA in the sense amplifier area SAA via the connection control unit CCNT2.
- the memory block MBLK 3 is connected to the sense amplifier SA in the sense amplifier area SAA via the connection control unit CCNT3.
- the access time tAAC has priority over the layout size of the compiled memory CM, it is effective to increase the number of memory blocks MBLK. Also, if the number of words specified according to user specifications exceeds the maximum number of words that can be placed in the memory block MBLK, it is necessary to form four or more memory block MBLKs.
- FIG. 14 shows the logic of the main decoder MD of the fifth embodiment.
- Memory block MBLKO-3 and connection control unit CCNTO-3 are identified by bit ADRO-1 of address signal ADR.
- the word line WL (word group WLG) in each memory block MBLKO—1 is identified by the upper bits ADR2-3 of the address signal ADR.
- FIG. 15 shows a connected memory CM according to a sixth embodiment of the present invention.
- the same elements as those described in the first and fourth embodiments are denoted by the same reference numerals, and detailed description thereof will be omitted.
- the compiled memory CM is an SRA M core, for example, and is formed in the ASIC chip together with other functional blocks as shown in FIG.
- a system is configured by an ASIC chip, or a system is configured with other semiconductor chips.
- the layout method of the compiled memory CM is the same as the flow shown in Fig. 12.
- the compiled memory CM is configured by a memory unit MUO-1 including a pair of memory blocks MBLK (MBLK 0-1 or MBLK2-3).
- the compiled memory CM has four memory blocks MBLKO-3.
- Each memory unit MUO-1 has the same structure as the memory block pair MBLKO-1 shown in FIG. 10 except that the assigned address signal is different. Ie 4 words A word group WLG (WLGO-11) is established for each line WL.
- the word lines WL WLO-47
- WLG word group WLG as one unit.
- the memory block MBLK2 is connected to the sense amplifier SA in the sense amplifier area SAA via the connection control unit CCNT2.
- the memory block MBLK3 is connected to the sense amplifier SA in the sense amplifier area SAA via the connection control unit CCNT3.
- the word group WLGO-3, WLG4-7, WLG8-11 is sequentially assigned to the memory block MBLKO-3 by the main decoder MD shown in FIG. Not limited to FIG. 15, when the compiled memory CM has a plurality of memory blocks MB LK, the word groups WLG are sequentially assigned to the memory blocks MBLK.
- FIG. 16 shows the logic of the main decoder MD of the sixth embodiment.
- Memory block MBLKO-3 and connection control unit CCNTO-3 are identified by bits ADR2-3 of address signal ADR.
- the word group WLG in each memory block MBLKO-3 is identified by the upper bits ADR4-5 of the address signal ADR.
- the word line WL in each word group WLG is identified by the lower bit ADRO-1 of the address signal ADR.
- the access time tAAC can be shortened also when the compiled memory CM is configured by a plurality of memory units MUO-1 and the word group WLG is configured by a plurality of word lines WL.
- the layout data of compiled memory CM that can shorten the access time tAAC can be easily generated.
- the present invention is a semiconductor memory such as a DRAM core or a ferroelectric memory core mounted on an ASIC chip, and can be applied to a compiled memory CM whose number of words is changed according to user specifications.
- the chip on which the compiled memory CM is installed is not limited to the ASIC chip.
- the compiled memory CM of the present invention may be mounted on an ASSP (Application Specific Standard Product) chip, a single chip microcomputer computer chip, or the like.
- the operation of the sense amplifier SA is performed using a self-timing technique.
- Circuit operation timing may be set optimally.
- the present invention can be applied to a compiled memory formed in a chip such as an ASIC and a compiled memory layout method.
Abstract
Description
Claims
Priority Applications (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/JP2006/325093 WO2008072354A1 (ja) | 2006-12-15 | 2006-12-15 | コンパイルドメモリ、asicチップおよびコンパイルドメモリのレイアウト方法 |
JP2008549184A JP5018786B2 (ja) | 2006-12-15 | 2006-12-15 | コンパイルドメモリ、asicチップおよびコンパイルドメモリのレイアウト方法 |
KR1020097013850A KR101129147B1 (ko) | 2006-12-15 | 2006-12-15 | 컴파일드 메모리, asic 칩 및 컴파일드 메모리의 레이아웃 방법 |
US12/482,656 US7864621B2 (en) | 2006-12-15 | 2009-06-11 | Compiled memory, ASIC chip, and layout method for compiled memory |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/JP2006/325093 WO2008072354A1 (ja) | 2006-12-15 | 2006-12-15 | コンパイルドメモリ、asicチップおよびコンパイルドメモリのレイアウト方法 |
Related Child Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US12/482,656 Continuation US7864621B2 (en) | 2006-12-15 | 2009-06-11 | Compiled memory, ASIC chip, and layout method for compiled memory |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2008072354A1 true WO2008072354A1 (ja) | 2008-06-19 |
Family
ID=39511386
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2006/325093 WO2008072354A1 (ja) | 2006-12-15 | 2006-12-15 | コンパイルドメモリ、asicチップおよびコンパイルドメモリのレイアウト方法 |
Country Status (4)
Country | Link |
---|---|
US (1) | US7864621B2 (ja) |
JP (1) | JP5018786B2 (ja) |
KR (1) | KR101129147B1 (ja) |
WO (1) | WO2008072354A1 (ja) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2010192052A (ja) * | 2009-02-19 | 2010-09-02 | Hitachi Ulsi Systems Co Ltd | 半導体装置 |
JP2010198711A (ja) * | 2009-02-27 | 2010-09-09 | Renesas Electronics Corp | 半導体記憶装置及びその検査方法 |
WO2018124062A1 (ja) | 2016-12-26 | 2018-07-05 | 塩野義製薬株式会社 | 含量均一性を改善した製剤の製造方法 |
Families Citing this family (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2006052738A2 (en) * | 2004-11-04 | 2006-05-18 | Fabbrix, Inc. | A method and process for design of integrated circuits using regular geometry patterns to obtain geometrically consistent component features |
JP5666108B2 (ja) * | 2009-07-30 | 2015-02-12 | ピーエスフォー ルクスコ エスエイアールエルPS4 Luxco S.a.r.l. | 半導体装置及びこれを備えるシステム |
US9449692B2 (en) * | 2011-08-03 | 2016-09-20 | Micron Technology, Inc. | Functional data programming and reading in a memory |
US9564205B2 (en) * | 2014-11-13 | 2017-02-07 | Winbond Electronics Corp. | Memory apparatus and method for accessing memory |
WO2021173943A1 (en) * | 2020-02-27 | 2021-09-02 | Micron Technology, Inc. | Apparatuses and methods for address based memory performance |
US11551746B2 (en) | 2020-11-19 | 2023-01-10 | Micron Technology, Inc. | Apparatuses including memory regions having different access speeds and methods for using the same |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH04209392A (ja) * | 1990-11-30 | 1992-07-30 | Nec Corp | 半導体記憶装置 |
JPH0887885A (ja) * | 1994-09-14 | 1996-04-02 | Hitachi Ltd | 半導体記憶回路装置 |
JP2000156081A (ja) * | 1998-11-18 | 2000-06-06 | Matsushita Electric Ind Co Ltd | 半導体装置 |
JP2002230060A (ja) * | 2001-02-05 | 2002-08-16 | Matsushita Electric Ind Co Ltd | メモリ分割数計算方法及び記憶装置生成方法 |
JP2004095058A (ja) * | 2002-08-30 | 2004-03-25 | Renesas Technology Corp | 半導体記憶装置 |
Family Cites Families (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6603413B2 (en) * | 2001-02-07 | 2003-08-05 | Canon Kabushiki Kaisha | Variable-length decoding apparatus and method |
JP4437891B2 (ja) * | 2003-03-24 | 2010-03-24 | Okiセミコンダクタ株式会社 | 同期型dramのデータ書込方法 |
JP4149969B2 (ja) | 2004-07-14 | 2008-09-17 | 株式会社東芝 | 半導体装置 |
JP4471902B2 (ja) * | 2005-07-28 | 2010-06-02 | 株式会社ルネサステクノロジ | 半導体記憶装置 |
-
2006
- 2006-12-15 JP JP2008549184A patent/JP5018786B2/ja not_active Expired - Fee Related
- 2006-12-15 WO PCT/JP2006/325093 patent/WO2008072354A1/ja active Application Filing
- 2006-12-15 KR KR1020097013850A patent/KR101129147B1/ko active IP Right Grant
-
2009
- 2009-06-11 US US12/482,656 patent/US7864621B2/en not_active Expired - Fee Related
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH04209392A (ja) * | 1990-11-30 | 1992-07-30 | Nec Corp | 半導体記憶装置 |
JPH0887885A (ja) * | 1994-09-14 | 1996-04-02 | Hitachi Ltd | 半導体記憶回路装置 |
JP2000156081A (ja) * | 1998-11-18 | 2000-06-06 | Matsushita Electric Ind Co Ltd | 半導体装置 |
JP2002230060A (ja) * | 2001-02-05 | 2002-08-16 | Matsushita Electric Ind Co Ltd | メモリ分割数計算方法及び記憶装置生成方法 |
JP2004095058A (ja) * | 2002-08-30 | 2004-03-25 | Renesas Technology Corp | 半導体記憶装置 |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2010192052A (ja) * | 2009-02-19 | 2010-09-02 | Hitachi Ulsi Systems Co Ltd | 半導体装置 |
JP2010198711A (ja) * | 2009-02-27 | 2010-09-09 | Renesas Electronics Corp | 半導体記憶装置及びその検査方法 |
US8737118B2 (en) | 2009-02-27 | 2014-05-27 | Renesas Electronics Corporation | Semiconductor memory device and test method therefor |
WO2018124062A1 (ja) | 2016-12-26 | 2018-07-05 | 塩野義製薬株式会社 | 含量均一性を改善した製剤の製造方法 |
US11135217B2 (en) | 2016-12-26 | 2021-10-05 | Shionogi & Co., Ltd. | Manufacturing process of formulation having improved content uniformity |
Also Published As
Publication number | Publication date |
---|---|
KR101129147B1 (ko) | 2012-03-27 |
US20090244988A1 (en) | 2009-10-01 |
KR20090090366A (ko) | 2009-08-25 |
JP5018786B2 (ja) | 2012-09-05 |
US7864621B2 (en) | 2011-01-04 |
JPWO2008072354A1 (ja) | 2010-03-25 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US5377153A (en) | Virtual ground read only memory circuit | |
JP5018786B2 (ja) | コンパイルドメモリ、asicチップおよびコンパイルドメモリのレイアウト方法 | |
KR100822613B1 (ko) | 반도체 기억 장치 | |
JP4331966B2 (ja) | 半導体集積回路 | |
JPH08279602A (ja) | 半導体メモリ装置 | |
KR100902125B1 (ko) | 저전력 디램 및 그 구동방법 | |
JP4439082B2 (ja) | 半導体記憶装置 | |
US8027219B2 (en) | Semiconductor memory devices having signal delay controller and methods performed therein | |
JP2009230776A (ja) | マルチポートメモリ、および該マルチポートメモリを備えるコンピュータシステム | |
US6510094B2 (en) | Method and apparatus for refreshing semiconductor memory | |
US6456563B1 (en) | Semiconductor memory device that operates in sychronization with a clock signal | |
US6515927B2 (en) | Semiconductor memory having a wide bus-bandwidth for input/output data | |
KR20170065433A (ko) | 메모리 관리 메커니즘을 구비한 전자 시스템 및 그의 동작 방법 | |
US6339560B1 (en) | Semiconductor memory based on address transitions | |
JP3913451B2 (ja) | 半導体記憶装置 | |
JP4757607B2 (ja) | 半導体メモリ | |
CN106469566B (zh) | 半导体器件 | |
KR20190066685A (ko) | 메모리 장치, 이를 포함하는 시스템 온 칩 및 메모리 장치의 동작 방법 | |
JPH1117137A (ja) | 半導体記憶装置 | |
CN117935882A (en) | Memory controller operation method and memory device | |
JP2005251273A (ja) | 半導体記憶装置 | |
JP2008066744A (ja) | 半導体記憶装置 | |
JP2009176419A (ja) | データ処理装置 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 06834820 Country of ref document: EP Kind code of ref document: A1 |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2008549184 Country of ref document: JP |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
WWE | Wipo information: entry into national phase |
Ref document number: 1020097013850 Country of ref document: KR |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 06834820 Country of ref document: EP Kind code of ref document: A1 |