US20200341838A1 - Encoding data in a modified-memory system - Google Patents
Encoding data in a modified-memory system Download PDFInfo
- Publication number
- US20200341838A1 US20200341838A1 US16/871,540 US202016871540A US2020341838A1 US 20200341838 A1 US20200341838 A1 US 20200341838A1 US 202016871540 A US202016871540 A US 202016871540A US 2020341838 A1 US2020341838 A1 US 2020341838A1
- Authority
- US
- United States
- Prior art keywords
- data bits
- bus
- memory
- dbi
- original data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Abandoned
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
- G06F11/08—Error detection or correction by redundancy in data representation, e.g. by using checking codes
- G06F11/10—Adding special bits or symbols to the coded information, e.g. parity check, casting out 9's or 11's
- G06F11/1004—Adding special bits or symbols to the coded information, e.g. parity check, casting out 9's or 11's to protect a block of data words, e.g. CRC or checksum
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
- G06F11/08—Error detection or correction by redundancy in data representation, e.g. by using checking codes
- G06F11/10—Adding special bits or symbols to the coded information, e.g. parity check, casting out 9's or 11's
- G06F11/1008—Adding special bits or symbols to the coded information, e.g. parity check, casting out 9's or 11's in individual solid state devices
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
- G06F11/08—Error detection or correction by redundancy in data representation, e.g. by using checking codes
- G06F11/10—Adding special bits or symbols to the coded information, e.g. parity check, casting out 9's or 11's
- G06F11/1008—Adding special bits or symbols to the coded information, e.g. parity check, casting out 9's or 11's in individual solid state devices
- G06F11/1048—Adding special bits or symbols to the coded information, e.g. parity check, casting out 9's or 11's in individual solid state devices using arrangements adapted for a specific error detection or correction feature
Definitions
- Embodiments of this invention relate to implementation of data bus inversion in a memory system.
- FIG. 1 An example illustrating data transmission between high-speed components within a single semiconductor device, or between two devices on a printed circuit board, is represented by the system 1 shown in FIG. 1 .
- a transmitter 2 e.g., a microprocessor
- transmission channels 4 x are often referred to as a “data bus,” which allows one or more data signals to be transmitted from one device to another.
- a data bus is susceptible to cross talk, simultaneous switching noise, intersymbol interference, and draws power based on the state of the data and/or frequency of data transition.
- DBI Data Bus Inversion
- Implementation of DBI includes encoding circuitry at the transmitter that assesses the relationship between data bits to be transmitted across a data bus and then decides (based on a particular DBI algorithm) if it would be advantageous to invert some or all of the data bits prior to transmission. If the data bits are inverted, an additional signal, referred to as a DBI bit, is also set at the encoding circuitry to indicate that the data bits are inverted. Typically, as shown in FIG. 1 , an extra channel 7 is then needed so that the DBI bit may be transmitted in parallel with the data bits to inform the receiving circuitry which groups of data bits have been inverted. The receiver 6 then uses the DBI bit in conjunction with decoding circuitry to return the incoming group of data bits to their original state.
- One specific DBI algorithm, illustrated in FIGS. 2A and 2B is referred to as the “minimum transitions” algorithm. While there may be variations of this technique, in general the minimum transitions algorithm begins by computing how many bits will transition during an upcoming cycle. When more than a certain number of transitions are determined, encoding circuitry inverts the entire bus, sets the DBI bit to a specified state (high or low depending on the implementation), and drives the inverted data bits and the DBI bit in parallel across the transmission channels 4 x and 7 , where the DBI bit is used to decode (i.e., de-invert) the inverted data bits prior to use in the receiver 6 .
- the minimum transitions algorithm begins by computing how many bits will transition during an upcoming cycle. When more than a certain number of transitions are determined, encoding circuitry inverts the entire bus, sets the DBI bit to a specified state (high or low depending on the implementation), and drives the inverted data bits and the DBI bit in parallel across the transmission channels 4 x and 7
- the minimum transitions technique can be implemented in one embedment using the encoding circuitry 8 of FIG. 2A . Because this technique is discussed at length in the above-mentioned '779 application, it is explained only briefly here. As shown, two successive bytes of data, Din(0:7) (the current byte) and Dout(0:7) (the previous byte), are compared. at exclusive OR (XOR) gates 3 on a bit-by-bit basis to determine which bits in the data signals are changing. After this XOR comparison, and in accordance with DBI algorithm 9 , the XOR results are summed, and a determination is made as to whether the sum is greater than four (i.e., whether there are at least five transitions from the previous byte of data to the current byte).
- XOR exclusive OR
- the current byte is inverted before it is transmitted, and the DBI bit 7 is transmitted as ‘1.’
- the sum is four or less (i.e., there are no more than four transitions from one byte to the next), the data is transmitted unaltered, and the DBI bit 7 is transmitted as ‘0.’
- FIG. 2B shows how the minimum transitions DBI algorithm 9 works to reduce the number of transitions in a random sequence of bytes, such that no more than four transitions are permitted between successive bytes. Again, this reduction of transitions reduces dynamic current draw at the transmitter 2 and improves the reliability of data transfer by reducing cross talk and simultaneous switching noise.
- DBI algorithms exists, such as are discussed in the above-mentioned '779 application, and thus one skilled in the art will realize that the DBI technique illustrated in FIGS. 2A and 2B is merely exemplary.
- some DBI techniques may use more than one DBI bit to provide even finer control over which bits of data will be inverted, as mentioned in the '779 application.
- FIG. 1 illustrates a block diagram of a data transmission system.
- FIGS. 2A and 2B illustrates a block diagram and a data table illustrating a minimum transitions DBI technique.
- FIG. 3A illustrates a system in which modified RAM ICs are used in conjunction with a logic integrated circuit having circuitry implementing functions traditionally present on RAM ICs.
- FIG. 3B illustrates a system, like that depicted in FIG. 3A , in which the logic integrated circuit also decodes the addresses for the modified RAM ICs.
- FIG. 4 illustrates how the systems of FIGS. 3A and 3B can be implemented using a vertically-stacked memory module on a printed circuit board.
- FIGS. 5 and 6 illustrate an embodiment of the invention in which DBI encoding and decoding is included on the logic integrated circuit, and in which the modified RAM ICs include a storage area for DBI bits.
- FIGS. 7A and 7B illustrate a system, similar to that of FIGS. 5 and 6 , in which a modified RAM IC is dedicate to providing DBI storage.
- FIGS. 8A-8C illustrate another embodiment of the invention in which DBI encoding is included on the logic integrated circuit, but DBI decoding is included on the modified RAM ICs, allowing for one-way DBI implementation but obviating the need to store DBI bits within the modified RAM ICs.
- DBI Data Bus Inversion
- a set of random access memory (RAM) integrated circuits (ICs) is separated from a logic system by a bus.
- the logic system can contain many of the logic functions traditionally performed on conventional RAM ICs, and accordingly the RAM ICs can be modified to not include such logic functions.
- the logic system which can be a logic integrated circuit intervening between the modified RAM ICs and a traditional memory controller, additionally contains DBI encoding and decoding circuitry.
- data is DBI encoded and at least one DBI bit issued when writing to the modified RAM ICs.
- the RAM ICs in turn store the DBI bit(s) with the encoded data.
- the encoded data is read from the modified RAM ICs, it is transmitted across the bus in its encoded state along with the DBI bit(s).
- the logic integrated circuit then decodes the data using the DBI bit(s) to return it to its original state.
- DBI may be implemented only for data being transmitted in one direction over the bus, for example, when writing to the memory.
- the logic integrated circuit includes a DBI encoder, with the DBI decoder occurring on the modified RAM ICs. Data written to the modified RAM ICs is encoded and issued with at least one DBI bit on the bus. At the modified RAM ICs, the data is then decoded using the DBI bit(s) and stored in its original un-encoded state. When data is read from the modified RAM ICs, no DBI encoding occurs, and there is no need to have stored the DBI bit(s) in the modified RAM ICs.
- System 100 comprises a system for reading/writing from/to a memory set 25 , which in the illustrated example comprises modified RAM ICs 16 x , whose modification from otherwise standard RAM ICs is discussed further below.
- a microprocessor 10 reads from and writes to the memory set 25 , but this is merely one example, and any other device or system could be used to so query the memory.
- the modified RAM ICs 16 x in the memory set 25 comprise dynamic RAM (DRAM) arrays of cells 61 , but could also comprise static RAM (SRAM) cells, or various other types of write/erasable non-volatile memory (e.g., NAND Flash cells, Flash EPROM cells, etc.).
- DRAM dynamic RAM
- SRAM static RAM
- Memory controller 12 Intervening between the microprocessor 10 and the memory set 25 is a memory controller 12 .
- Memory controllers 12 are well known in the art and work to create a standard interface 20 with which the microprocessor 10 can predictably communicate.
- the memory controller 12 couples to the microprocessor's data (DQ), address (A), and control (cntl) busses 11 , and converts them to new busses 13 DQ′, A′, and cntl′ suitable for interfacing with a logic integrated circuit (IC) 14 , discussed further below.
- Memory controller 12 typically comprises an integrated circuit separate and independent from other components in the system 100 , but this is not strictly necessary, and the controller 12 could be integrated with other components if desired.
- the memory controller 12 can comprise any well known memory controller 12 used in the industry.
- a logic IC 14 intervenes between the memory controller 12 and the modified RAM ICs 16 x , and in this respect both the logic IC 14 and the modified RAM ICs 16 x differ from standard memory typically used. Departing from such standard solutions, the logic IC 14 contains much if not all of the logic circuitry 49 typically present on a standard RAM IC (not shown). For example, the logic IC 14 can contain command decode and queuing circuitry 50 .
- Such circuitry 50 interprets the various command signals on the cntl′ data bus (such as signals write enable (WE), row address strobe (RAS), column address strobe (CAS), and chip select (CS), assuming the modified RAM arrays 16 x comprise DRAM memory), and issues and organizes the commands as appropriate for distribution to the modified PAM ICs 16 x along a control bus cntl′′ intervening between the logic IC 14 and the RAM ICs 16 x .
- the logic IC 14 may also contain redundancy circuitry 52 for determining faulty memory addresses in the modified RAM ICs 16 x and for rerouting around such defective addresses to functioning memory cells using programmable fuses or antifuses, as is well known.
- Logic IC 14 may additionally contain error correction circuitry 54 , which can comprise well known circuitry for assessing and correcting faulty data in accordance with any number of error correction algorithms. Further, logic IC 14 may contain test mode circuitry 56 , which is typically used during manufacturing and/or under the application of special test commands to test the operation of the various modified RAM ICs 16 x . Typically, such circuits 50 - 56 are formed as part of the peripheral logic of a standard memory integrated circuit (not shown), but in the illustrated system such circuitry has been removed from the modified RAM ICs 16 x .
- Logic IC 14 may also contain additional integration circuitry 58 not traditionally contained on a conventional RAM ICs, but which are relevant more globally to the modular integration of a number of memory arrays.
- integration circuitry 58 can assess the operation of the various signals on bus 15 (DQ′′, A′′, cntl′′) intervening between the logic IC 14 and the modified RAM ICs, and if necessary can reroute around any connections deemed to be faulty.
- modified RAM ICs 16 x need not contain such circuits, and instead modified RAM ICs 16 x can comprise only the array of storage cells 61 and other minimal circuitry desirable for the functioning of such cells.
- Such other minimal circuitry can comprise the sense amplifiers 62 used to sense the data state of the cells, column decoder circuitry 64 , row decoder circuitry 66 , column driver circuitry 68 , and row driver circuitry 70 .
- the RAM ICs 16 x can be made smaller in area. Reducing the size of the RAM ICs 16 x helps the yield and reliability of such devices, and allows their assembly into a smaller package or packages, as discussed further below with reference to FIG. 4 .
- Bus 15 contains data (DQ′′), address (A′′), and control (cad′′) signals that might otherwise appear internally on a standard RAM IC.
- FIG. 3B illustrates a modification to the system 100 ′ in which the address decode circuitry (e.g., column and row decoders 64 and 66 ) has been moved off of the modified RAM ICs 16 x and onto the logic IC 14 ′.
- a new address decode block 75 is included which performs column and row decoding, and (in the illustrated example) decoding to choose an appropriate one of the modified RAM ICs 16 x .
- bus 15 ′ contains the already-decoded row and column driver signals (Cx i ; Ry i ) necessary to access any of the x columns and y rows of the various modified RAM ICs 16 x .
- the column and row driver circuits 68 and 70 still remain on the modified RAM ICs 16 x , along with the sense amplifiers 62 , but these circuit blocks could also be moved onto the logic IC 14 ′ in an alternative embodiment, making the modified RAM ICs smaller still. While FIG. 3B illustrates a full decoding of the addresses on bus A′, note that this is not strictly required and instead some or all of the address decoding could occur on the modified RAM ICs 16 x .
- FIG. 4 shows portions of the systems 100 or 100 ′ of FIGS. 3A and 3B as they could be manufactured.
- a standard memory controller 12 is affixed to a printed circuit board (PCB) 30 using solder bumps 80 as is typical. Some of these bumps 80 will correspond to signals on bus 11 between the memory controller 12 and the microprocessor 10 (not shown in FIG. 4 ), while other bumps will correspond to signals on bus 13 between the memory controller 12 and the logic IC 14 .
- the logic IC 14 and the modified RAM ICs 16 x are integrated in a multichip module 40 .
- the logic IC 14 and the modified RAM ICs 16 x are vertically stacked in the module 40 within a singular package.
- through-silicon vias (TSVs) 83 run though the logic IC 14 and the modified RAM ICs 16 x .
- the module 40 is also affixed to the printed circuit board using solder bumps 82 .
- the microprocessor 10 could be on the same or separate PCBs from the memory controller 12 and/or the module 40 .
- any combination of the microprocessor 10 , memory controller 12 , and module 40 could be integrated within its own package or multichip module.
- the functionality contained within the memory controller 12 might also be moved to the logic IC 14 , thus eliminating the need for the memory controller 12 and the bus 13 in the system.
- the assembled configuration shown in FIG. 4 is merely one example of how system 100 or 100 ′ could be implemented.
- DBI Data Bus Inversion
- system 200 has had its logic IC 14 modified to include a DBI encoder/decoder 210 .
- the DBI encoder/decoder 210 couples to the data bus DQ′′ intervening between the logic IC 14 and the modified RAM ICs 216 x .
- at least one DBI bit is included with data bus DQ′′.
- each of the modified RAM ICs 216 x includes a DBI storage area 225 .
- the modified RAM IC 216 x in this example comprises a X8 device, in which eight bits (i.e., a byte) is written/read to/from the modified RAM ICs 216 x at one time.
- the array of memory cells 61 in the modified RAM IC 216 x includes areas in which each of the eight bits DQ′′ 1 - 8 are stored. Additionally, and as just mentioned, a dedicated. DBI storage area 225 is also present for the DBI bit.
- the modified RAM IC 216 x comprises in this example a X9 device comprising at each address a data byte and its associated DBI bit.
- the areas dedicated to the data and the DBI bit 225 can occur in different locations on the modified RAM IC 216 x , and need not appear as depicted as contiguous rectangles.
- DBI encoding and decoding in the depicted example are implemented by encoding 210 a and decoding 210 b circuitry.
- Such circuitry 210 a and 210 b in this embodiment, is present only on the logic IC 14 as opposed to the modified RAM ICs 216 x , which is consistent with the desired intention in system 200 to move as much of the logic circuitry off of the modified RAM ICs 216 x and onto the logic IC 14 as possible.
- a command decoder 240 on the logic IC 14 interprets the write command and enables the encoder 210 a via a write enable signal, W/E.
- the encoder 210 a may be similar to that shown earlier in FIGS.
- the encoder 210 a assesses the incoming bytes of data on DQ′, and issues new data DQ′′ and a DBI bit in accordance with the DBI algorithm. Data DQ′′ may there be completely or partially inverted depending on the algorithm and the state of the given DBI bit. In any event, a given encoded data byte DQ′′ is stored with its DBI bit at the appropriate address in the modified RAM IC 216 x .
- Reading from the modified RAM IC 216 x requires decoding of the stored data in accordance with the DBI bit stored with that data. This occurs as follows.
- control decoder 240 on the logic IC interprets a read command, and enables the decoder 210 b via read enable signal, RIE. Encoded data is then read at the specified address from the modified.
- the decoder 210 b interprets the DBI bit, and decodes, i.e., completely or partially un-inverts the data as necessary, and passes the decoded data to the memory controller 12 via bus DQ′.
- the modified RAM ICs do not require an additional DBI storage area 225 , and hence can be the same as the modified RAM ICs 16 x discussed in conjunction with FIGS. 3A and 3B .
- one of the modified RAM ICs, specifically RAM IC 250 is dedicated to the storage of DBI bits.
- Such an architecture for system 200 ′ is logically implemented assuming the modified RAM ICs 16 x and 250 are x1 devices.
- each bit in the encoded byte is sent to a particular one of the modified RAM ICs 16 x, i.e., RAM IC 16 1 receives DQ′′ 1 , etc., and the DBI bit gets stored in parallel in modified RAM IC 250 .
- a given address in this example causes writing to/reading from all of the modified RAM ICs in the memory set 25 . While this implementation requires no further modifications to the modified RAM ICs 16 x and 250 , it will require an additional modified RAM IC 250 to be present in the module 40 if the same amount of data is to be stored, as shown in FIG. 7B .
- RAM IC 250 could be substituted for RAM residing outside of the module 40 for RAM residing outside of the module 40 .
- other RAM 260 shown in dotted lines in FIG. 7A , could be used for storage of the DBI bit, which other RAM 260 could appear outside of the module 40 , such as for example, in the logic IC 14 , in the memory controller 12 , or as a separate device in the system 200 ′. So long as such other RAM 260 is connected to the same address as its encoded data in the modified RAM ICs 16 x , proper DBI encoding and decoding can be established.
- FIGS. 8A-8C illustrate another embodiment of a system 300 implementing the disclosed DBI technique in which DBI is applied to the data in only one direction of the data flow between the logic IC 14 and the memory 25 .
- DBI is applied to reduce data transmissions only when writing data to the memory set 25 .
- logic IC 14 only contains a DBI encoder 210 a, as shown in FIGS. 8A and 8R
- the corresponding decoder for this one-way DBI implementation appears on the modified RAM ICs 316 x as decoder block 310 .
- encoded data DQ′′( 1 : 8 ) encoded via a suitable DBI algorithm at encoder 210 a on the logic IC 14 , is sent to modified RAM ICs 316 x via bus 15 .
- Each of the modified. RAM ICs 316 x contains a decoder 310 .
- command decoder 240 (which can be present on the logic IC 14 , the modified RAM ICs 316 x , or elsewhere), issues a write enable signal, W/E, which enables both the encoder 210 a on the logic IC 14 .
- This causes encoder 210 a to issue encoded data DQ′′ on bus 15 which may or may not be completely or partially inverted depending on the DBI algorithm being used.
- decoder 310 While writing (i.e., while W/E is asserted), decoder 310 will take the encoded data DQ′′ after transmission on the bus 15 and un-invert such data as mandated by the DBI bit at the modified RAM ICs 316 x .
- Such decoding can be accomplished by the simple utilization of XOR gates 315 for each of the data channels, DQ′′, as shown in FIG. 8C .
- decoder 310 can always remain enabled, as XOR gates only consume power during transitions.
- the W/E signal may still be used to enable other aspects of the receiving circuitry within memory set 25 .
- FIGS. 8A-8C there is no need to store the DBI bit, because the data is decoded by decoder 310 on the modified RAM ICs 316 x before it is written into the RAM ICs. Provisions therefore do not need to be made, as in earlier embodiments, to provide a DBI storage area 225 on the modified RAM ICs 316 x ( FIGS. 5 and 6 ), nor is a dedicated DBI storage modified. RAM ICs 250 necessary ( FIGS. 7A and 7B ).
- the modified RAM ICs 316 x include a DBI decoder 310 .
- the provision of such a decoder 310 on the modified RAM ICs 316 x is contrary to the desire that the modified RAM ICs contain as little excess circuitry as possible.
- such a DBI decoder 310 can require little more than the inclusion of XOR gates 315 or other simple digital logic gates on the modified RAM ICs 316 x , which additional decoding logic 310 is negligible in the context of the modified RAM IC 316 x , and takes up minimal layout area on the modified RAM ICs 316 x .
- XOR gates 315 or other simple digital logic gates on the modified RAM ICs 316 x
- additional decoding logic 310 is negligible in the context of the modified RAM IC 316 x
- takes up minimal layout area on the modified RAM ICs 316 x For clarity's sake, it is repeated here that although the implementations described herein function with only one DBI bit, more complex D
- the DBI encoder can be provided on the modified RAM ICs, while a DB 1 decoder is present on the logic IC 14 .
- provision for DBI encoder circuitry on the modified RAM ICs may be incompatible with the goals of the system. This is because encoding circuitry may take a significant amount of area (at least, compared to the decoder) to accommodate the analog circuitry (e.g., operational amplifiers, etc.) such encoders typically require. As such, encoding circuitry may be better relegated to the logic IC 14 in a system in which the modified RAM ICs 316 x are made as small as possible.
- FIGS. 5-8C can be implemented in a system in which the data is fully or partially decoded, as shown in FIG. 3B .
- FIG. 3B Such systems that in effect combine the system of FIG. 39 with the systems of FIGS. 5-8C are not illustrated because such an extension of the disclosed techniques should be apparent to one skilled in the art on the basis of this disclosure.
- the disclosed technique can be used more generally with any memory structure having storage cells, even if not a discrete integrated circuit.
- a plurality of memory sub-arrays could operate as the plurality of memory structures useable in the technique, even if such sub-arrays are themselves integrated on a single integrated circuit.
- the disclosed technique still has applicability when used with a single memory IC (or memory structure).
- DBI algorithms include the “minimum zeros” algorithm and the “minimum ones” algorithm.
- Minimum zeros algorithm
- minimum ones algorithm
- the purpose of these algorithms is, respectively, to minimize the number of binary zeros or binary ones transmitted across a channel.
- Such minimum zeros or ones algorithms conserve power when the driver or receiver circuits coupled to the transmission channels are not full CMOS and therefore will draw more power when transmitting or receiving a particular data state. For example, if a pull-up resistor connected to the voltage supply is used in a particular driver circuit, driving a logic ‘0’ will require more power than would driving a logic ‘1’.
- all of these DBI algorithms have the common feature of sequentially receiving groups of N original data bits and selectively encoding each group to form a corresponding group of N encoded data bits while issuing at least one encoding (DBI) indicator associated with each group of the N encoded data bits.
- DBI encoding
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Engineering & Computer Science (AREA)
- Quality & Reliability (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Security & Cryptography (AREA)
- Dram (AREA)
Abstract
Description
- Embodiments of this invention relate to implementation of data bus inversion in a memory system.
- An example illustrating data transmission between high-speed components within a single semiconductor device, or between two devices on a printed circuit board, is represented by the
system 1 shown inFIG. 1 . InFIG. 1 , a transmitter 2 (e.g., a microprocessor) sends data over one ormore transmission channels 4 x (e.g., conductive traces “on-chip” in a semiconductor device or on a printed circuit board) to a receiver 6 (e.g., another microprocessor or memory). As a group,such transmission channels 4 x are often referred to as a “data bus,” which allows one or more data signals to be transmitted from one device to another. - As discussed in U.S. patent application Ser. No. 11/873,779, filed Oct. 17, 2007, a data bus is susceptible to cross talk, simultaneous switching noise, intersymbol interference, and draws power based on the state of the data and/or frequency of data transition. One way to reduce these adverse effects and to prevent unnecessary power consumption is to encode the data. One specific form of data encoding that can be used is Data Bus Inversion (DBI).
- Implementation of DBI includes encoding circuitry at the transmitter that assesses the relationship between data bits to be transmitted across a data bus and then decides (based on a particular DBI algorithm) if it would be advantageous to invert some or all of the data bits prior to transmission. If the data bits are inverted, an additional signal, referred to as a DBI bit, is also set at the encoding circuitry to indicate that the data bits are inverted. Typically, as shown in
FIG. 1 , anextra channel 7 is then needed so that the DBI bit may be transmitted in parallel with the data bits to inform the receiving circuitry which groups of data bits have been inverted. Thereceiver 6 then uses the DBI bit in conjunction with decoding circuitry to return the incoming group of data bits to their original state. - One specific DBI algorithm, illustrated in
FIGS. 2A and 2B , is referred to as the “minimum transitions” algorithm. While there may be variations of this technique, in general the minimum transitions algorithm begins by computing how many bits will transition during an upcoming cycle. When more than a certain number of transitions are determined, encoding circuitry inverts the entire bus, sets the DBI bit to a specified state (high or low depending on the implementation), and drives the inverted data bits and the DBI bit in parallel across thetransmission channels receiver 6. - The minimum transitions technique can be implemented in one embedment using the
encoding circuitry 8 ofFIG. 2A . Because this technique is discussed at length in the above-mentioned '779 application, it is explained only briefly here. As shown, two successive bytes of data, Din(0:7) (the current byte) and Dout(0:7) (the previous byte), are compared. at exclusive OR (XOR)gates 3 on a bit-by-bit basis to determine which bits in the data signals are changing. After this XOR comparison, and in accordance withDBI algorithm 9, the XOR results are summed, and a determination is made as to whether the sum is greater than four (i.e., whether there are at least five transitions from the previous byte of data to the current byte). If the sum is greater than four, the current byte is inverted before it is transmitted, and theDBI bit 7 is transmitted as ‘1.’ Alternatively, if the sum is four or less (i.e., there are no more than four transitions from one byte to the next), the data is transmitted unaltered, and theDBI bit 7 is transmitted as ‘0.’ -
FIG. 2B shows how the minimumtransitions DBI algorithm 9 works to reduce the number of transitions in a random sequence of bytes, such that no more than four transitions are permitted between successive bytes. Again, this reduction of transitions reduces dynamic current draw at thetransmitter 2 and improves the reliability of data transfer by reducing cross talk and simultaneous switching noise. - Other DBI algorithms exists, such as are discussed in the above-mentioned '779 application, and thus one skilled in the art will realize that the DBI technique illustrated in
FIGS. 2A and 2B is merely exemplary. For example, some DBI techniques may use more than one DBI bit to provide even finer control over which bits of data will be inverted, as mentioned in the '779 application. - The inventor believes that advances in system integration are making the implementation of DBI, and other data encoding algorithms, more attractive. At the same time, the use of such algorithms is becoming more important as systems shrink and as it becomes increasingly important that such systems reduce their power consumption and operate at high speeds.
-
FIG. 1 illustrates a block diagram of a data transmission system. -
FIGS. 2A and 2B illustrates a block diagram and a data table illustrating a minimum transitions DBI technique. -
FIG. 3A illustrates a system in which modified RAM ICs are used in conjunction with a logic integrated circuit having circuitry implementing functions traditionally present on RAM ICs. -
FIG. 3B illustrates a system, like that depicted inFIG. 3A , in which the logic integrated circuit also decodes the addresses for the modified RAM ICs. -
FIG. 4 illustrates how the systems ofFIGS. 3A and 3B can be implemented using a vertically-stacked memory module on a printed circuit board. -
FIGS. 5 and 6 illustrate an embodiment of the invention in which DBI encoding and decoding is included on the logic integrated circuit, and in which the modified RAM ICs include a storage area for DBI bits. -
FIGS. 7A and 7B illustrate a system, similar to that ofFIGS. 5 and 6 , in which a modified RAM IC is dedicate to providing DBI storage. -
FIGS. 8A-8C illustrate another embodiment of the invention in which DBI encoding is included on the logic integrated circuit, but DBI decoding is included on the modified RAM ICs, allowing for one-way DBI implementation but obviating the need to store DBI bits within the modified RAM ICs. - Implementations of Data Bus Inversion (DBI) techniques within a memory system are disclosed. In one embodiment, a set of random access memory (RAM) integrated circuits (ICs) is separated from a logic system by a bus. The logic system can contain many of the logic functions traditionally performed on conventional RAM ICs, and accordingly the RAM ICs can be modified to not include such logic functions. The logic system, which can be a logic integrated circuit intervening between the modified RAM ICs and a traditional memory controller, additionally contains DBI encoding and decoding circuitry. In such a system, data is DBI encoded and at least one DBI bit issued when writing to the modified RAM ICs. The RAM ICs in turn store the DBI bit(s) with the encoded data. When the encoded data is read from the modified RAM ICs, it is transmitted across the bus in its encoded state along with the DBI bit(s). The logic integrated circuit then decodes the data using the DBI bit(s) to return it to its original state.
- In another embodiment, DBI may be implemented only for data being transmitted in one direction over the bus, for example, when writing to the memory. In this implementation, the logic integrated circuit includes a DBI encoder, with the DBI decoder occurring on the modified RAM ICs. Data written to the modified RAM ICs is encoded and issued with at least one DBI bit on the bus. At the modified RAM ICs, the data is then decoded using the DBI bit(s) and stored in its original un-encoded state. When data is read from the modified RAM ICs, no DBI encoding occurs, and there is no need to have stored the DBI bit(s) in the modified RAM ICs.
- A
system 100 benefitting from the implementation of DBI is shown inFIG. 3A .System 100 comprises a system for reading/writing from/to a memory set 25, which in the illustrated example comprises modifiedRAM ICs 16 x, whose modification from otherwise standard RAM ICs is discussed further below. In such a system, it can be assumed, for illustrative purposes, that amicroprocessor 10 reads from and writes to the memory set 25, but this is merely one example, and any other device or system could be used to so query the memory. In one embodiment, the modifiedRAM ICs 16 x in the memory set 25 comprise dynamic RAM (DRAM) arrays ofcells 61, but could also comprise static RAM (SRAM) cells, or various other types of write/erasable non-volatile memory (e.g., NAND Flash cells, Flash EPROM cells, etc.). - Intervening between the
microprocessor 10 and the memory set 25 is amemory controller 12.Memory controllers 12 are well known in the art and work to create astandard interface 20 with which themicroprocessor 10 can predictably communicate. Thememory controller 12 couples to the microprocessor's data (DQ), address (A), and control (cntl) busses 11, and converts them tonew busses 13 DQ′, A′, and cntl′ suitable for interfacing with a logic integrated circuit (IC) 14, discussed further below.Memory controller 12 typically comprises an integrated circuit separate and independent from other components in thesystem 100, but this is not strictly necessary, and thecontroller 12 could be integrated with other components if desired. Thememory controller 12 can comprise any well knownmemory controller 12 used in the industry. - In the disclosed embodiment, a
logic IC 14 intervenes between thememory controller 12 and the modifiedRAM ICs 16 x, and in this respect both thelogic IC 14 and the modifiedRAM ICs 16 x differ from standard memory typically used. Departing from such standard solutions, thelogic IC 14 contains much if not all of thelogic circuitry 49 typically present on a standard RAM IC (not shown). For example, thelogic IC 14 can contain command decode and queuingcircuitry 50.Such circuitry 50 interprets the various command signals on the cntl′ data bus (such as signals write enable (WE), row address strobe (RAS), column address strobe (CAS), and chip select (CS), assuming the modifiedRAM arrays 16 x comprise DRAM memory), and issues and organizes the commands as appropriate for distribution to the modifiedPAM ICs 16 x along a control bus cntl″ intervening between thelogic IC 14 and theRAM ICs 16 x. Thelogic IC 14 may also containredundancy circuitry 52 for determining faulty memory addresses in the modifiedRAM ICs 16 x and for rerouting around such defective addresses to functioning memory cells using programmable fuses or antifuses, as is well known.Logic IC 14 may additionally containerror correction circuitry 54, which can comprise well known circuitry for assessing and correcting faulty data in accordance with any number of error correction algorithms. Further,logic IC 14 may containtest mode circuitry 56, which is typically used during manufacturing and/or under the application of special test commands to test the operation of the various modifiedRAM ICs 16 x. Typically, such circuits 50-56 are formed as part of the peripheral logic of a standard memory integrated circuit (not shown), but in the illustrated system such circuitry has been removed from the modifiedRAM ICs 16 x. -
Logic IC 14 may also containadditional integration circuitry 58 not traditionally contained on a conventional RAM ICs, but which are relevant more globally to the modular integration of a number of memory arrays. For example,integration circuitry 58 can assess the operation of the various signals on bus 15 (DQ″, A″, cntl″) intervening between thelogic IC 14 and the modified RAM ICs, and if necessary can reroute around any connections deemed to be faulty. - As a result of circuits 50-58 being present on the
logic IC 14, the modifiedRAM ICs 16 x need not contain such circuits, and instead modifiedRAM ICs 16 x can comprise only the array ofstorage cells 61 and other minimal circuitry desirable for the functioning of such cells. Such other minimal circuitry can comprise thesense amplifiers 62 used to sense the data state of the cells,column decoder circuitry 64,row decoder circuitry 66,column driver circuitry 68, androw driver circuitry 70. - By moving the
logic circuitry 49 off of the modifiedRAM ICs 16 x, theRAM ICs 16 x can be made smaller in area. Reducing the size of theRAM ICs 16 x helps the yield and reliability of such devices, and allows their assembly into a smaller package or packages, as discussed further below with reference toFIG. 4 . - Because the
logic circuitry 49 is moved off of theRAM ICs 16 x and onto thelogic IC 14, abus 15 intervenes between thelogic IC 14 and the modifiedRAM ICs 16 x, as mentioned above.Bus 15 contains data (DQ″), address (A″), and control (cad″) signals that might otherwise appear internally on a standard RAM IC. -
FIG. 3B illustrates a modification to thesystem 100′ in which the address decode circuitry (e.g., column androw decoders 64 and 66) has been moved off of the modifiedRAM ICs 16 x and onto thelogic IC 14′. Specifically, a newaddress decode block 75 is included which performs column and row decoding, and (in the illustrated example) decoding to choose an appropriate one of the modifiedRAM ICs 16 x. Accordingly, in this embodiment,bus 15′ contains the already-decoded row and column driver signals (Cxi; Ryi) necessary to access any of the x columns and y rows of the various modifiedRAM ICs 16 x. As shown, the column androw driver circuits RAM ICs 16 x, along with thesense amplifiers 62, but these circuit blocks could also be moved onto thelogic IC 14′ in an alternative embodiment, making the modified RAM ICs smaller still. WhileFIG. 3B illustrates a full decoding of the addresses on bus A′, note that this is not strictly required and instead some or all of the address decoding could occur on the modifiedRAM ICs 16 x. -
FIG. 4 shows portions of thesystems FIGS. 3A and 3B as they could be manufactured. As shown, astandard memory controller 12 is affixed to a printed circuit board (PCB) 30 using solder bumps 80 as is typical. Some of these bumps 80 will correspond to signals onbus 11 between thememory controller 12 and the microprocessor 10 (not shown inFIG. 4 ), while other bumps will correspond to signals onbus 13 between thememory controller 12 and thelogic IC 14. - In the implementation shown in
FIG. 4 , thelogic IC 14 and the modifiedRAM ICs 16 x are integrated in amultichip module 40. As shown, thelogic IC 14 and the modifiedRAM ICs 16 x are vertically stacked in themodule 40 within a singular package. To facilitate the running of the signals onbus 15, through-silicon vias (TSVs) 83 run though thelogic IC 14 and the modifiedRAM ICs 16 x. Themodule 40 is also affixed to the printed circuit board using solder bumps 82. - Although shown with the
memory controller 12 andmodule 40 on thesame PCB 30, this is not strictly necessary. Additionally, the microprocessor 10 (not shown) could be on the same or separate PCBs from thememory controller 12 and/or themodule 40. Further, any combination of themicroprocessor 10,memory controller 12, andmodule 40 could be integrated within its own package or multichip module. Additionally, it is possible that the functionality contained within thememory controller 12 might also be moved to thelogic IC 14, thus eliminating the need for thememory controller 12 and thebus 13 in the system. Thus, the assembled configuration shown inFIG. 4 is merely one example of howsystem - With embodiments of basic systems described, discussion now turns to the application of Data Bus Inversion (DBI) to such systems. As shown in
FIG. 5 ,system 200 has had itslogic IC 14 modified to include a DBI encoder/decoder 210. The DBI encoder/decoder 210 couples to the data bus DQ″ intervening between thelogic IC 14 and the modifiedRAM ICs 216 x. Additionally, at least one DBI bit is included with data bus DQ″. To handle this DBI bit, each of the modifiedRAM ICs 216 x includes aDBI storage area 225. - Further details concerning such an implementation of DBI in
system 200 is shown inFIG. 6 . As can be seen, it is assumed that the modifiedRAM IC 216 x in this example comprises a X8 device, in which eight bits (i.e., a byte) is written/read to/from the modifiedRAM ICs 216 x at one time. The array ofmemory cells 61 in the modifiedRAM IC 216 x includes areas in which each of the eight bits DQ″1-8 are stored. Additionally, and as just mentioned, a dedicated.DBI storage area 225 is also present for the DBI bit. In effect, the modifiedRAM IC 216 x comprises in this example a X9 device comprising at each address a data byte and its associated DBI bit. As one skilled in the art will realize, the areas dedicated to the data and theDBI bit 225 can occur in different locations on the modifiedRAM IC 216 x, and need not appear as depicted as contiguous rectangles. - DBI encoding and decoding in the depicted example are implemented by encoding 210 a and
decoding 210 b circuitry.Such circuitry logic IC 14 as opposed to the modifiedRAM ICs 216 x, which is consistent with the desired intention insystem 200 to move as much of the logic circuitry off of the modifiedRAM ICs 216 x and onto thelogic IC 14 as possible. - When writing to the modified RAM IC 216 x (or more generally, the memory set 25), data is presented from the
memory controller 12 to thelogic IC 14 on bus DQ′, along with the appropriate address for that data and a write command as embodied in the signals present on control bus, cntl′. (The address bus from thememory controller 12, A′, is not shown inFIG. 6 for clarity, but may share a bus with the control signals in a time multiplexed fashion.). Acommand decoder 240 on thelogic IC 14 interprets the write command and enables theencoder 210 a via a write enable signal, W/E. The encoder 210 a may be similar to that shown earlier inFIGS. 2A and 2B , and will operate in accordance with a desired DBI algorithm, of which many exist. Theencoder 210 a assesses the incoming bytes of data on DQ′, and issues new data DQ″ and a DBI bit in accordance with the DBI algorithm. Data DQ″ may there be completely or partially inverted depending on the algorithm and the state of the given DBI bit. In any event, a given encoded data byte DQ″ is stored with its DBI bit at the appropriate address in the modifiedRAM IC 216 x. - Reading from the modified RAM IC 216 x (i.e., from memory set 25) requires decoding of the stored data in accordance with the DBI bit stored with that data. This occurs as follows.
control decoder 240 on the logic IC interprets a read command, and enables thedecoder 210 b via read enable signal, RIE. Encoded data is then read at the specified address from the modified.RAM IC 216 x on bus DQ″ along with the DBI stored at that address. Thedecoder 210 b interprets the DBI bit, and decodes, i.e., completely or partially un-inverts the data as necessary, and passes the decoded data to thememory controller 12 via bus DQ′. - This example allows for the implementation of DBI without requiring the modified
RAM ICs 216 x to have any DBI logic circuitry whatsoever. However, in this embodiment, the modifiedRAM ICs 216 x will need to store the DBI bit, which will either require the modified.RAM ICs 216 x to increase in size, or to store less data within the nominal size. - Another embodiment of the disclosed DBI technique that does not require storage of the DBI bits is shown in the
system 200′ ofFIG. 7A , In this embodiment, the modified RAM ICs do not require an additionalDBI storage area 225, and hence can be the same as the modifiedRAM ICs 16 x discussed in conjunction withFIGS. 3A and 3B . However, in this embodiment, one of the modified RAM ICs, specificallyRAM IC 250, is dedicated to the storage of DBI bits. Such an architecture forsystem 200′ is logically implemented assuming the modifiedRAM ICs RAM IC 16 1 receives DQ″1, etc., and the DBI bit gets stored in parallel in modifiedRAM IC 250. Thus, a given address in this example causes writing to/reading from all of the modified RAM ICs in the memory set 25. While this implementation requires no further modifications to the modifiedRAM ICs RAM IC 250 to be present in themodule 40 if the same amount of data is to be stored, as shown inFIG. 7B . - If it is undesired or impossible to include an additional modified
RAM IC 250 within themodule 40 as shown inFIGS. 7A and 7B , one could substitute the modifiedRAM IC 250 within themodule 40 for RAM residing outside of themodule 40. For example, other RAM 260, shown in dotted lines inFIG. 7A , could be used for storage of the DBI bit, which other RAM 260 could appear outside of themodule 40, such as for example, in thelogic IC 14, in thememory controller 12, or as a separate device in thesystem 200′. So long as such other RAM 260 is connected to the same address as its encoded data in the modifiedRAM ICs 16 x, proper DBI encoding and decoding can be established. -
FIGS. 8A-8C illustrate another embodiment of asystem 300 implementing the disclosed DBI technique in which DBI is applied to the data in only one direction of the data flow between thelogic IC 14 and thememory 25. Specifically, in the depicted embodiment, DBI is applied to reduce data transmissions only when writing data to the memory set 25. Accordingly,logic IC 14 only contains aDBI encoder 210 a, as shown inFIGS. 8A and 8R The corresponding decoder for this one-way DBI implementation appears on the modifiedRAM ICs 316 x asdecoder block 310. - Operation of the one-way DBI algorithm in
system 300 is explained further inFIGS. 8B and 8C . As shown, encoded data DQ″(1:8), encoded via a suitable DBI algorithm atencoder 210 a on thelogic IC 14, is sent to modifiedRAM ICs 316 x viabus 15. Each of the modified.RAM ICs 316 x contains adecoder 310. When writing is taking place, command decoder 240 (which can be present on thelogic IC 14, the modifiedRAM ICs 316 x, or elsewhere), issues a write enable signal, W/E, which enables both theencoder 210 a on thelogic IC 14. This causes encoder 210 a to issue encoded data DQ″ onbus 15, which may or may not be completely or partially inverted depending on the DBI algorithm being used. - In any event, while writing (i.e., while W/E is asserted),
decoder 310 will take the encoded data DQ″ after transmission on thebus 15 and un-invert such data as mandated by the DBI bit at the modifiedRAM ICs 316 x. Such decoding can be accomplished by the simple utilization ofXOR gates 315 for each of the data channels, DQ″, as shown inFIG. 8C . In this embodiment,decoder 310 can always remain enabled, as XOR gates only consume power during transitions. However, the W/E signal may still be used to enable other aspects of the receiving circuitry within memory set 25. Assuming the implementation of a simple minimum transitions DBI algorithm,decoder 310 will include eight XOR gates, i.e., one for each of the data bits from the byte, which XOR gates receive at their inputs the current DBI bit and one of the bits in the encoded byte. If inversion occurred for a given byte of data (i.e., if DBI=‘1’),XOR gates 315 will un-invert the data, producing true, un-encoded data bytes DQarray(1:8) which will be stored in the modifiedRAM ICs 316 x. - To reiterate, by operation of the
DBI encoder 210 a on thelogic IC 14 anddecoder 310 on theRAM ICs 316 x, data is or is not inverted during transmission to the modifiedRAM ICs 316 x acrossbus 15 to improve power consumption, data integrity, and reliability. But, the inverted data is then decoded. and stored in its true state in the array ofcells 61 of thoseRAM ICs 316 x. By contrast, when reading fromsystem 300, no DBI algorithm operates in this embodiment, and instead the true data is merely read from the modifiedRAM ICs 316 x and is transmitted back acrossbus 15 without encoding. - As a result, and beneficially, in this embodiment of
FIGS. 8A-8C , there is no need to store the DBI bit, because the data is decoded bydecoder 310 on the modifiedRAM ICs 316 x before it is written into the RAM ICs. Provisions therefore do not need to be made, as in earlier embodiments, to provide aDBI storage area 225 on the modified RAM ICs 316 x (FIGS. 5 and 6 ), nor is a dedicated DBI storage modified.RAM ICs 250 necessary (FIGS. 7A and 7B ). - However, this embodiment does require that the modified
RAM ICs 316 x include aDBI decoder 310. Generally, the provision of such adecoder 310 on the modifiedRAM ICs 316 x is contrary to the desire that the modified RAM ICs contain as little excess circuitry as possible. But as shown inFIG. 8C , such aDBI decoder 310 can require little more than the inclusion ofXOR gates 315 or other simple digital logic gates on the modifiedRAM ICs 316 x, whichadditional decoding logic 310 is negligible in the context of the modifiedRAM IC 316 x, and takes up minimal layout area on the modifiedRAM ICs 316 x. For clarity's sake, it is repeated here that although the implementations described herein function with only one DBI bit, more complex DBI algorithms could be implemented which might require multiple DBI bits. - In another embodiment (not shown), the DBI encoder can be provided on the modified RAM ICs, while a DB1 decoder is present on the
logic IC 14. This presents essentially the reverse of the one-way DBI technique discussed inFIGS. 8A-8C , such that DBI operates when reading from the memory set 25, but does not operate when writing to the memory set 25. While such an implementation would certainly be functional, provision for DBI encoder circuitry on the modified RAM ICs may be incompatible with the goals of the system. This is because encoding circuitry may take a significant amount of area (at least, compared to the decoder) to accommodate the analog circuitry (e.g., operational amplifiers, etc.) such encoders typically require. As such, encoding circuitry may be better relegated to thelogic IC 14 in a system in which the modifiedRAM ICs 316 x are made as small as possible. - Any of the DBI implementation set forth in
FIGS. 5-8C can be implemented in a system in which the data is fully or partially decoded, as shown inFIG. 3B . However, such systems that in effect combine the system ofFIG. 39 with the systems ofFIGS. 5-8C are not illustrated because such an extension of the disclosed techniques should be apparent to one skilled in the art on the basis of this disclosure. - While envisioned for use with discrete integrated memory circuits, the disclosed technique can be used more generally with any memory structure having storage cells, even if not a discrete integrated circuit. For example, a plurality of memory sub-arrays could operate as the plurality of memory structures useable in the technique, even if such sub-arrays are themselves integrated on a single integrated circuit. Additionally, while illustrated in the context of a plurality of memory ICs (or a plurality of memory structures more generally), it should be noted that the disclosed technique still has applicability when used with a single memory IC (or memory structure).
- As noted earlier, many different types of DBI algorithms exist. Different DBI algorithms are beneficial in different circumstances, and not all DBI algorithms are directed to minimizing the number of data transition across transmission channels. For example, other well-known DBI algorithms include the “minimum zeros” algorithm and the “minimum ones” algorithm. The purpose of these algorithms is, respectively, to minimize the number of binary zeros or binary ones transmitted across a channel. Such minimum zeros or ones algorithms conserve power when the driver or receiver circuits coupled to the transmission channels are not full CMOS and therefore will draw more power when transmitting or receiving a particular data state. For example, if a pull-up resistor connected to the voltage supply is used in a particular driver circuit, driving a logic ‘0’ will require more power than would driving a logic ‘1’. As a result, use of a minimum zeros DBI algorithm would be warranted. Likewise, if a pull-down resistor is used, a minimum ones algorithm would be warranted. The minimum transitions and either of the minimum zeros or ones algorithms can also be combined in a DBI algorithm, as is disclosed in U.S. patent application Ser. No. 12/015,311, filed Jan. 16, 2008. In another DBI algorithm, discussed in the above-mentioned Ser. No. 11/873,779 application, only a portion of the data bits on a bus are inverted to balance the logic states in an encoded byte across the bus, which can be referred to as a Balanced DBI algorithm. Regardless of the DBI algorithm used, all of these DBI algorithms have the common feature of sequentially receiving groups of N original data bits and selectively encoding each group to form a corresponding group of N encoded data bits while issuing at least one encoding (DBI) indicator associated with each group of the N encoded data bits. Any of these DBI algorithms can be used in the context of the disclosed embodiments of the invention, and therefore the illustration or focus given to the minimum transition DBI algorithm should be understood as merely exemplary and non-limiting.
- While some implementations have been disclosed, it should be understood that the disclosed circuitry can be achieved in many different ways to the same useful ends as described herein. In short, it should be understood that the inventive concepts disclosed herein are capable of many modifications. To the extent such modifications fall within the scope of the appended claims and their equivalents, they are intended to be covered by this patent.
Claims (21)
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US16/871,540 US20200341838A1 (en) | 2008-06-11 | 2020-05-11 | Encoding data in a modified-memory system |
Applications Claiming Priority (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US12/136,868 US9116828B2 (en) | 2008-06-11 | 2008-06-11 | Data bus inversion usable in a memory system |
US14/833,876 US10146614B2 (en) | 2008-06-11 | 2015-08-24 | Encoding data in a modified-memory system |
US16/183,136 US10649842B2 (en) | 2008-06-11 | 2018-11-07 | Encoding data in a modified-memory system |
US16/871,540 US20200341838A1 (en) | 2008-06-11 | 2020-05-11 | Encoding data in a modified-memory system |
Related Parent Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US16/183,136 Continuation US10649842B2 (en) | 2008-06-11 | 2018-11-07 | Encoding data in a modified-memory system |
Publications (1)
Publication Number | Publication Date |
---|---|
US20200341838A1 true US20200341838A1 (en) | 2020-10-29 |
Family
ID=41415871
Family Applications (4)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US12/136,868 Active 2032-04-13 US9116828B2 (en) | 2008-06-11 | 2008-06-11 | Data bus inversion usable in a memory system |
US14/833,876 Active 2028-06-22 US10146614B2 (en) | 2008-06-11 | 2015-08-24 | Encoding data in a modified-memory system |
US16/183,136 Active US10649842B2 (en) | 2008-06-11 | 2018-11-07 | Encoding data in a modified-memory system |
US16/871,540 Abandoned US20200341838A1 (en) | 2008-06-11 | 2020-05-11 | Encoding data in a modified-memory system |
Family Applications Before (3)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US12/136,868 Active 2032-04-13 US9116828B2 (en) | 2008-06-11 | 2008-06-11 | Data bus inversion usable in a memory system |
US14/833,876 Active 2028-06-22 US10146614B2 (en) | 2008-06-11 | 2015-08-24 | Encoding data in a modified-memory system |
US16/183,136 Active US10649842B2 (en) | 2008-06-11 | 2018-11-07 | Encoding data in a modified-memory system |
Country Status (1)
Country | Link |
---|---|
US (4) | US9116828B2 (en) |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20200243486A1 (en) * | 2019-01-30 | 2020-07-30 | Sunrise Memory Corporation | Device with embedded high-bandwidth, high-capacity memory using wafer bonding |
US20220327067A1 (en) * | 2021-04-09 | 2022-10-13 | Samsung Electronics Co., Ltd. | Signal encoding method and a semiconductor device to generate an optimal transition code in a multi-level signaling system |
US11675500B2 (en) | 2020-02-07 | 2023-06-13 | Sunrise Memory Corporation | High capacity memory circuit with low effective latency |
US20230214136A1 (en) * | 2022-01-05 | 2023-07-06 | Kioxia Corporation | Memory system |
US11710729B2 (en) | 2018-09-24 | 2023-07-25 | Sunrise Memory Corporation | Wafer bonding in fabrication of 3-dimensional NOR memory circuits |
US11789644B2 (en) | 2020-02-24 | 2023-10-17 | Sunrise Memory Corporation | Memory centric system incorporating computational memory |
US11810640B2 (en) | 2021-02-10 | 2023-11-07 | Sunrise Memory Corporation | Memory interface with configurable high-speed serial data lanes for high bandwidth memory |
US11844204B2 (en) | 2019-12-19 | 2023-12-12 | Sunrise Memory Corporation | Process for preparing a channel region of a thin-film transistor in a 3-dimensional thin-film transistor array |
US11915768B2 (en) | 2015-09-30 | 2024-02-27 | Sunrise Memory Corporation | Memory circuit, system and method for rapid retrieval of data sets |
Families Citing this family (35)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9116828B2 (en) | 2008-06-11 | 2015-08-25 | Micron Technology, Inc. | Data bus inversion usable in a memory system |
US9087025B2 (en) | 2009-02-05 | 2015-07-21 | Micron Technology, Inc. | Data encoding using spare channels in a memory system |
US8879654B2 (en) * | 2010-03-10 | 2014-11-04 | Micron Technology, Inc. | Communication interface with configurable encoding based on channel termination |
US8405529B2 (en) * | 2011-03-11 | 2013-03-26 | Taiwan Semiconductor Manufacturing Company, Ltd. | Using bus inversion to reduce simultaneous signal switching |
US8909840B2 (en) | 2011-12-19 | 2014-12-09 | Advanced Micro Devices, Inc. | Data bus inversion coding |
US8854236B2 (en) * | 2012-05-18 | 2014-10-07 | Micron Technology, Inc. | Methods and apparatuses for low-power multi-level encoded signals |
KR20140032787A (en) | 2012-09-07 | 2014-03-17 | 삼성전자주식회사 | Memory system including nonvolatile memory device and control method thereof |
US9384822B2 (en) * | 2013-03-15 | 2016-07-05 | Gsi Technology, Inc. | Systems and methods involving data bus inversion memory circuitry, configuration and/or operation including data signals grouped into 10 bits and/or other features |
TWI614607B (en) * | 2013-03-15 | 2018-02-11 | 積佳半導體股份有限公司 | Memory device and method for memory operation utilizing data bus inversion |
US9122570B2 (en) * | 2013-09-03 | 2015-09-01 | Nanya Technology Corp. | Data pattern generation for I/O training and characterization |
US9979416B2 (en) * | 2014-12-10 | 2018-05-22 | Rambus Inc. | Memory controller and method of data bus inversion using an error detection correction code |
US20160291678A1 (en) * | 2015-03-30 | 2016-10-06 | Advanced Micro Devices, Inc. | Power reduction in bus interconnects |
CN104851452B (en) * | 2015-05-06 | 2017-09-29 | 中国航天科技集团公司第九研究院第七七一研究所 | Integrated morphology and its control method on a kind of memory bank piece based on power consumption control |
US10365833B2 (en) | 2016-01-22 | 2019-07-30 | Micron Technology, Inc. | Apparatuses and methods for encoding and decoding of signal lines for multi-level communication architectures |
US9922686B2 (en) | 2016-05-19 | 2018-03-20 | Micron Technology, Inc. | Apparatuses and methods for performing intra-module databus inversion operations |
KR20180013212A (en) * | 2016-07-29 | 2018-02-07 | 에스케이하이닉스 주식회사 | Data bit inversion controller and semiconductor device including thereof |
US10146719B2 (en) * | 2017-03-24 | 2018-12-04 | Micron Technology, Inc. | Semiconductor layered device with data bus |
KR102299880B1 (en) * | 2017-04-04 | 2021-09-09 | 에스케이하이닉스 주식회사 | Data conversion apparatus and method |
US10540304B2 (en) | 2017-04-28 | 2020-01-21 | Advanced Micro Devices, Inc. | Power-oriented bus encoding for data transmission |
US10283187B2 (en) | 2017-07-19 | 2019-05-07 | Micron Technology, Inc. | Apparatuses and methods for providing additional drive to multilevel signals representing data |
US10447512B2 (en) | 2017-08-07 | 2019-10-15 | Micron Technology, Inc. | Channel equalization for multi-level signaling |
US10277435B2 (en) | 2017-08-07 | 2019-04-30 | Micron Technology, Inc. | Method to vertically align multi-level cells |
US10425260B2 (en) | 2017-08-07 | 2019-09-24 | Micron Technology, Inc. | Multi-level signaling in memory with wide system interface |
US10530617B2 (en) | 2017-08-07 | 2020-01-07 | Micron Technology, Inc. | Programmable channel equalization for multi-level signaling |
KR102433331B1 (en) * | 2017-12-26 | 2022-08-18 | 에스케이하이닉스 주식회사 | Integrated circuit chip |
US10664432B2 (en) | 2018-05-23 | 2020-05-26 | Micron Technology, Inc. | Semiconductor layered device with data bus inversion |
US11144385B2 (en) | 2018-08-21 | 2021-10-12 | Micron Technology, Inc. | Transmission failure feedback schemes for reducing crosstalk |
US10747909B2 (en) | 2018-09-25 | 2020-08-18 | Northrop Grumman Systems Corporation | System architecture to mitigate memory imprinting |
US10754993B2 (en) | 2018-09-25 | 2020-08-25 | Northrop Grumman Systems Corporation | Architecture to mitigate configuration memory imprinting in programmable logic |
US10964702B2 (en) | 2018-10-17 | 2021-03-30 | Micron Technology, Inc. | Semiconductor device with first-in-first-out circuit |
CN112712839A (en) * | 2019-10-25 | 2021-04-27 | 长鑫存储技术(上海)有限公司 | Read operation circuit, semiconductor memory and read operation method |
US11294838B2 (en) * | 2020-07-29 | 2022-04-05 | Micron Technology, Inc. | Signaling mechanism for bus inversion |
JP2022047393A (en) * | 2020-09-11 | 2022-03-24 | キオクシア株式会社 | Memory system, semiconductor storage device, and data reading method |
US11656958B2 (en) * | 2021-04-29 | 2023-05-23 | Mellanox Technologies, Ltd. | Redundancy data bus inversion sharing |
US20230290423A1 (en) * | 2022-03-11 | 2023-09-14 | Changxin Memory Technologies, Inc. | Method and apparatus for testing memory chip, storage medium, and electronic device |
Family Cites Families (25)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP3686155B2 (en) * | 1996-03-21 | 2005-08-24 | 株式会社ルネサステクノロジ | Image decoding device |
US6182177B1 (en) * | 1997-06-13 | 2001-01-30 | Intel Corporation | Method and apparatus for maintaining one or more queues of elements such as commands using one or more token queues |
KR100272171B1 (en) | 1998-08-19 | 2000-12-01 | 윤종용 | Data input/output system reducing power consumption and input/output method using the same |
US6505308B1 (en) * | 1999-10-28 | 2003-01-07 | Lsi Logic Corporation | Fast built-in self-repair circuit |
US6721918B2 (en) * | 2000-12-29 | 2004-04-13 | Intel Corporation | Method and apparatus for encoding a bus to minimize simultaneous switching outputs effect |
US20020156953A1 (en) | 2001-02-28 | 2002-10-24 | Beiley Mark A. | Dynamic bus inversion method |
US7549011B2 (en) | 2001-08-30 | 2009-06-16 | Micron Technology, Inc. | Bit inversion in memory devices |
JP4134637B2 (en) * | 2002-08-27 | 2008-08-20 | 株式会社日立製作所 | Semiconductor device |
JP4068427B2 (en) | 2002-10-08 | 2008-03-26 | エルピーダメモリ株式会社 | Data inversion circuit and semiconductor device |
WO2004052015A1 (en) * | 2002-11-29 | 2004-06-17 | Sony Corporation | Encoder and its method |
DE60315122D1 (en) | 2003-01-14 | 2007-09-06 | St Microelectronics Srl | Method and device for transmitting data over a single line, in particular for transmitting bus data with minimization of the bus switching activity |
US20050123835A1 (en) | 2003-12-09 | 2005-06-09 | Luying Sun | Non-aqueous electrolytes having an extended temperature range for battery applications |
US7406608B2 (en) | 2004-02-05 | 2008-07-29 | Micron Technology, Inc. | Fast and compact circuit for bus inversion |
US7102544B1 (en) * | 2005-05-31 | 2006-09-05 | Altera Corporation | Method and system for improving memory interface data integrity in PLDs |
US7869525B2 (en) * | 2005-08-01 | 2011-01-11 | Ati Technologies, Inc. | Dynamic bus inversion method and system |
KR100643498B1 (en) | 2005-11-21 | 2006-11-10 | 삼성전자주식회사 | Circuit and method for data bus inversion for use in semiconductor memory |
US7640444B2 (en) * | 2006-01-26 | 2009-12-29 | Nils Graef | Systems and methods for low power bus operation |
US7936289B2 (en) * | 2006-04-04 | 2011-05-03 | Samsung Electronics Co., Ltd. | Method, device, and system for data communication with preamble for reduced switching noise |
US20080016289A1 (en) * | 2006-07-11 | 2008-01-17 | Pennock James D | External memory interface engine |
US7619945B2 (en) * | 2006-08-18 | 2009-11-17 | Unity Semiconductor Corporation | Memory power management |
US7755951B2 (en) * | 2006-09-01 | 2010-07-13 | Canon Kabushiki Kaisha | Data output apparatus, memory system, data output method, and data processing method |
JP4033226B1 (en) | 2006-09-27 | 2008-01-16 | 富士ゼロックス株式会社 | Image processing apparatus, image forming apparatus, and program |
US7979666B2 (en) * | 2006-12-08 | 2011-07-12 | William Marsh Rice University | System and method for context-independent codes for off-chip interconnects |
US8117526B2 (en) * | 2007-10-29 | 2012-02-14 | Qimonda Ag | Apparatus and method for generating a transmit signal and apparatus and method for extracting an original message from a received signal |
US9116828B2 (en) | 2008-06-11 | 2015-08-25 | Micron Technology, Inc. | Data bus inversion usable in a memory system |
-
2008
- 2008-06-11 US US12/136,868 patent/US9116828B2/en active Active
-
2015
- 2015-08-24 US US14/833,876 patent/US10146614B2/en active Active
-
2018
- 2018-11-07 US US16/183,136 patent/US10649842B2/en active Active
-
2020
- 2020-05-11 US US16/871,540 patent/US20200341838A1/en not_active Abandoned
Cited By (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US12002523B2 (en) | 2015-09-30 | 2024-06-04 | Sunrise Memory Corporation | Memory circuit, system and method for rapid retrieval of data sets |
US11915768B2 (en) | 2015-09-30 | 2024-02-27 | Sunrise Memory Corporation | Memory circuit, system and method for rapid retrieval of data sets |
US11710729B2 (en) | 2018-09-24 | 2023-07-25 | Sunrise Memory Corporation | Wafer bonding in fabrication of 3-dimensional NOR memory circuits |
US20210398949A1 (en) * | 2019-01-30 | 2021-12-23 | Sunrise Memory Corporation | Memory device including modular memory units and modular circuit units for concurrent memory operations |
US11670620B2 (en) * | 2019-01-30 | 2023-06-06 | Sunrise Memory Corporation | Device with embedded high-bandwidth, high-capacity memory using wafer bonding |
US20200243486A1 (en) * | 2019-01-30 | 2020-07-30 | Sunrise Memory Corporation | Device with embedded high-bandwidth, high-capacity memory using wafer bonding |
US11923341B2 (en) * | 2019-01-30 | 2024-03-05 | Sunrise Memory Corporation | Memory device including modular memory units and modular circuit units for concurrent memory operations |
US20230260969A1 (en) * | 2019-01-30 | 2023-08-17 | Sunrise Memory Corporation | Device with embedded high-bandwidth, high-capacity memory using wafer bonding |
US11844204B2 (en) | 2019-12-19 | 2023-12-12 | Sunrise Memory Corporation | Process for preparing a channel region of a thin-film transistor in a 3-dimensional thin-film transistor array |
US11675500B2 (en) | 2020-02-07 | 2023-06-13 | Sunrise Memory Corporation | High capacity memory circuit with low effective latency |
US11789644B2 (en) | 2020-02-24 | 2023-10-17 | Sunrise Memory Corporation | Memory centric system incorporating computational memory |
US11810640B2 (en) | 2021-02-10 | 2023-11-07 | Sunrise Memory Corporation | Memory interface with configurable high-speed serial data lanes for high bandwidth memory |
US11822492B2 (en) * | 2021-04-09 | 2023-11-21 | Samsung Electronics Co., Ltd. | Signal encoding method and a semiconductor device to generate an optimal transition code in a multi-level signaling system |
US20220327067A1 (en) * | 2021-04-09 | 2022-10-13 | Samsung Electronics Co., Ltd. | Signal encoding method and a semiconductor device to generate an optimal transition code in a multi-level signaling system |
US20230214136A1 (en) * | 2022-01-05 | 2023-07-06 | Kioxia Corporation | Memory system |
US12008247B2 (en) * | 2022-01-05 | 2024-06-11 | Kioxia Corporation | Memory system |
Also Published As
Publication number | Publication date |
---|---|
US10146614B2 (en) | 2018-12-04 |
US20090313521A1 (en) | 2009-12-17 |
US9116828B2 (en) | 2015-08-25 |
US20190138388A1 (en) | 2019-05-09 |
US20150363260A1 (en) | 2015-12-17 |
US10649842B2 (en) | 2020-05-12 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US20200341838A1 (en) | Encoding data in a modified-memory system | |
US11494277B2 (en) | Data encoding using spare channels in a memory system | |
US20180286472A1 (en) | Memory Module and Memory System | |
KR20200053754A (en) | Memory devices, memory systems and methods of operating memory devices | |
US11015547B2 (en) | Apparatuses and methods for storing redundancy repair information for memories | |
US11209994B2 (en) | Systems and methods for data path power savings in DDR5 memory devices | |
CN106407135B (en) | Electronic device | |
US7349233B2 (en) | Memory device with read data from different banks | |
US6307794B1 (en) | Semiconductor memory device and signal line shifting method | |
CN112988059A (en) | Memory device, electronic device and method for storing data | |
CN1937067B (en) | Reconfigurable input/output in hierarchical memory link | |
KR20210157749A (en) | Device for interfacing between memory device and memory controller, package and system including the same | |
US10579280B2 (en) | On-die termination control for memory systems | |
US20020031020A1 (en) | Semiconductor memory device capable of masking data to be written | |
US7230857B2 (en) | Methods of modifying operational characteristic of memory devices using control bits received through data pins and related devices and systems | |
US6791896B2 (en) | Semiconductor memory device capable of changing an address space thereof | |
JP2008077635A (en) | Memory system | |
KR100228522B1 (en) | Semiconductor memory device | |
JP2000058772A (en) | Semiconductor memory device | |
WO2023076830A1 (en) | Stacked memory device with interface die | |
US7296102B2 (en) | Method and chip to expand pins of the chip | |
CN116978415A (en) | Storage equipment | |
JPH05234367A (en) | Semiconductor memory device | |
JPH03125394A (en) | Semiconductor storage device |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
STPP | Information on status: patent application and granting procedure in general |
Free format text: DOCKETED NEW CASE - READY FOR EXAMINATION |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: NON FINAL ACTION MAILED |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: RESPONSE TO NON-FINAL OFFICE ACTION ENTERED AND FORWARDED TO EXAMINER |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: FINAL REJECTION MAILED |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: DOCKETED NEW CASE - READY FOR EXAMINATION |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: NON FINAL ACTION MAILED |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: RESPONSE TO NON-FINAL OFFICE ACTION ENTERED AND FORWARDED TO EXAMINER |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: FINAL REJECTION MAILED |
|
STCB | Information on status: application discontinuation |
Free format text: ABANDONED -- FAILURE TO RESPOND TO AN OFFICE ACTION |