GB2125590A - Data storage with error detection - Google Patents

Data storage with error detection Download PDF

Info

Publication number
GB2125590A
GB2125590A GB8321372A GB8321372A GB2125590A GB 2125590 A GB2125590 A GB 2125590A GB 8321372 A GB8321372 A GB 8321372A GB 8321372 A GB8321372 A GB 8321372A GB 2125590 A GB2125590 A GB 2125590A
Authority
GB
United Kingdom
Prior art keywords
memory devices
bit
control signal
data
data storage
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
GB8321372A
Other versions
GB2125590B (en
GB8321372D0 (en
Inventor
Michael John Ratcliffe
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Fujitsu Services Ltd
Original Assignee
Fujitsu Services Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Fujitsu Services Ltd filed Critical Fujitsu Services Ltd
Priority to GB8321372A priority Critical patent/GB2125590B/en
Publication of GB8321372D0 publication Critical patent/GB8321372D0/en
Publication of GB2125590A publication Critical patent/GB2125590A/en
Application granted granted Critical
Publication of GB2125590B publication Critical patent/GB2125590B/en
Expired legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00Error detection; Error correction; Monitoring
    • G06F11/07Responding to the occurrence of a fault, e.g. fault tolerance
    • G06F11/08Error detection or correction by redundancy in data representation, e.g. by using checking codes
    • G06F11/10Adding special bits or symbols to the coded information, e.g. parity check, casting out 9's or 11's
    • G06F11/1008Adding special bits or symbols to the coded information, e.g. parity check, casting out 9's or 11's in individual solid state devices
    • G06F11/1012Adding 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 codes or arrangements adapted for a specific type of error
    • G06F11/1016Error in accessing a memory location, i.e. addressing error
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00Error detection; Error correction; Monitoring
    • G06F11/07Responding to the occurrence of a fault, e.g. fault tolerance
    • G06F11/08Error detection or correction by redundancy in data representation, e.g. by using checking codes
    • G06F11/10Adding special bits or symbols to the coded information, e.g. parity check, casting out 9's or 11's
    • G06F11/1008Adding special bits or symbols to the coded information, e.g. parity check, casting out 9's or 11's in individual solid state devices
    • G06F11/1012Adding 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 codes or arrangements adapted for a specific type of error
    • G06F11/1028Adjacent errors, e.g. error in n-bit (n>1) wide storage units, i.e. package error

Abstract

Data storage apparatus consisting of an array of RAM chips 20, with Hamming code checking for detecting double-bit errors. Address signals such as Ao are fanned-out to the chips by way of driver circuits 31. Each driver circuit distributes an address bit to only two columns of chips (or in some case, just one column), so that if one of the driver circuits fails, no more than two chips in each row will be affected. Hence, no more than two bits in any data word will be in error, and this can be detected by the Hamming code. <IMAGE>

Description

SPECIFICATION Data storage apparatus Background to the invention This invention relates to data storage apparatus.
It is well known to provide data storage apparatus with some form of error detection and correction logic. For example, Hamming codes are typically able to detect double bit errors and to detect and correct single bit errors.
It is also well known to construct data storage apparatus from a number of memory devices such as random access memory chips. The chips may for example be mounted on a printed circuit board which provides input and output connections for data and for various control signals such as address bits, strobe signals and so on. The control signals are supplied to all the chips in parallel, and this usually requires the provision of suitable driver circuits for supplying the necessary drive current for fanning out each control signal to a number of chips.
One problem with such an arrangement is that if one of the driver circuits fails, it can cause several chips to give an incorrect output, leading to multiple errors which cannot be detected by the error detection logic.
One object of the invention is to provide data storage apparatus in which failure of a driver circuit does not lead to undetectable errors.
Summary of the invention According to the invention there is provided data storage apparatus for storing multi-bit data words, the apparatus comprising: (a) a plurality of memory devices each having a plurality of individually addressable storage locations, (b) error detecting means capable of detecting n-bit errors in any data word stored in the memory devices, where n is greater than one, and (c) a plurality of driver circuits for distributing a control signal in parallel to all the memory devices, wherein each driver circuit supplies the control signal to a group of data storage devices containing no more than n bits belonging to any one data word.
Thus, it can be seen that, if a driver circuit fails this will affect, at most, only n bits in any given data word and hence any errors resulting from this failure will be detectable by the error detecting means.
In a particular embodiment of the invention, the storage apparatus comprises a plurality of rows of memory devices, the bits of any given word all being stored in the same row, and each driver circuit supplies the control signal to no more than n memory devices in each row.
Brief description of the drawings One data storage apparatus in accordance with the invention will now be described by way of example with reference to the accompanying drawings, of which: Figure 1 is an overall diagram of the storage apparatus; Figure 2 shows one circuit board in greater detail; and Figure 3 shows the way in which control signals are distributed to chips on the circuit board.
Description of an embodiment of the invention Referring to Figure 1, the data storage apparatus comprises a data store 10 having 512K individually addressable locations, each holding 39 bits. The first 32 bits in each location represent a data word, while the remaining 7 bits represent a Hamming code for checking and correcting the data.
The store 10 is constructed from three identical printed circuit boards 11, each of which holds a 13-bit-wide section of the store. Each board has thirteen data input terminals 12 and thirteen data output terminals 14 for input and output of data or Hamming code bits.
Each input data word is fed to a Hamming code generator 15 which produces the appropriate Hamming code. This code is written into the store 10 along with the data. Each data word read out of the store 10 is fed, along with the associated Hamming code, to a checking circuit 16. This circuit is capable of detecting double bit errors in any data word, and of detecting and correcting single bit errors. In the latter case, the checking circuit 16 indicates which bit is in error, and controls an inverter circuit 17 so as to invert and hence correct the faulty bit.
Hamming code generation and checking circuits are well known in the art and hence these circuits will not be described in any further detail in this specification.
Referring now to Figure 2, this shows one of the circuit boards 11 in greater detail. The board carries twenty six random access memory (RAM) chips 20, arranged in two rows and thirteen columns. Each chip contains 256K individually addressable bits. The three circuit boards thus have a total of thirty-nine columns, each column containing two chips 20 with a total of 512K bit locations i.e. one bit for each word.
The bits of each data word (and the Hamming bits) are distributed among the columns with one bit of each word per column. It should be noted that the bits of any given word are all in the same row of chips.
Each data input terminal 12 is connected by way of a driver circuit 21 to the data inputs of the corresponding column of memory chips 20. The data outputs of each column of chips 20 are combined in an OR-gate 22 and fed by way of a data output buffer 23 to the appropriate one of the thirteen output terminals 14.
In order to address one of the 256K locations in a chip, it is necessary to supply an 18-bit address. This address is supplied to the board in two successive 9-bit portions, over nine address lines A0-A8. These two portions are gated into two nine-bit address registers (not shown) in each chip by respective strobe signals RAS and CAS. The two rows of chips have separate RAS signals and this serves to select between the two rows, so as to address one of the two 256K halves of the store. Various other control signals such as a write enable signal WE are also provided for controlling the chips 20, but these are not described here since they are not relevant to an understanding of the present invention.
It can be seen that there are various control signals such as the address signals A0-A8, the strobe signal CAS, and the write enable signal WE which must be distributed to every chip on the board. Referring to Figure 3, this shows the way in which one of the control signals A0 is distributed. It will be understood that the other control signals are distributed in a similar manner.
The signal A0 is fed to a driver circuit 30 which produces sufficient output current to drive seven further driver circuits 31. Each of these driver circuits 31 then fans out the signal to two columns of chips 20 i.e. to four chips, two in the first row and two in the second. (The exception to this is the last of the driver circuits; since there is an odd number of columns, this fans out the signal to just one column.) Thus, it can be seen that each driver circuit 31 fans out the signal A0 to not more than two chips in each row. As a result, if one of the driver circuits fails, not more than two chips in either row will be wrongly addressed and hence not more than two data bits in any given word will be in error. Hence the Hamming checking circuit 16 will always be able to detect such a failure.
The output of the driver circuit 30 and the similar circuits for the other control bits are also fed to a parity checking circuit 32, along with a parity bit P.
This circuit checks the correctness of the control signals and, if it detects an error, produces a parity fail signal PF by way of a parity buffer register 33. It can be seen that, if any one of the driver circuits 30 fails, this will be detected by the parity checking circuit 32.

Claims (8)

1. Data storage apparatus for storing multi-bit data words, the apparatus comprising: (a) a plurality of memory devices each having a plurality of individually addressable storage locations, (b) error detecting means capable of detecting n-bit errors in any data word stored in the memory devices, where n is greater than one, and (c) a plurality of driver circuits for distributing a control signal in parallel to all the memory devices, wherein each driver circuit supplies the control signal to a group of data storage devices containing no more than n bits belonging to any one data word.
2. Apparatus according to Claim 1 wherein said memory devices are random-access memory chips.
3. Apparatus according to either preceding claim, wherein said control signal is one of a plurality of address signals for addressing the memory devices.
4. Apparatus according to any preceding claim wherein said error checking means comprises a Hamming code checking circuit.
5. Apparatus according to Claim 4 wherein the number of errors detectable in any given data word is two.
6. Apparatus according to any preceding claim wherein the memory devices are arranged in a plurality of rows, the bits of any given data word all being stored in the same row, and each driver circuit supplies the control signal to no more than n memory devices in each row.
7. Data storage apparatus for storing multi-bit data words, the apparatus comprising: (a) a plurality of memory devices logically arranged in rows and columns, each memory device having a plurality of individually addressable bit locations, and the bits of each data word being distributed among a plurality of said memory devices in the same row, one bit per column, (b) error detecting means capable of detecting n-bit errors in any data word stored in the memory devices, where n is greater than one, and (c) a plurality of driver circuits for distributing a control signal in parallel to all the memory devices, wherein each driver circuit supplies the control signal to all the memory devices in not more than n columns.
8. Data storage apparatus substantially as hereinbefore described with reference to the accompanying drawings.
GB8321372A 1982-08-14 1983-08-09 Data storage with error detection Expired GB2125590B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
GB8321372A GB2125590B (en) 1982-08-14 1983-08-09 Data storage with error detection

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
GB8223439 1982-08-14
GB8321372A GB2125590B (en) 1982-08-14 1983-08-09 Data storage with error detection

Publications (3)

Publication Number Publication Date
GB8321372D0 GB8321372D0 (en) 1983-09-07
GB2125590A true GB2125590A (en) 1984-03-07
GB2125590B GB2125590B (en) 1985-12-04

Family

ID=26283597

Family Applications (1)

Application Number Title Priority Date Filing Date
GB8321372A Expired GB2125590B (en) 1982-08-14 1983-08-09 Data storage with error detection

Country Status (1)

Country Link
GB (1) GB2125590B (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB2342739A (en) * 1998-07-17 2000-04-19 Sun Microsystems Inc Memory address checking within a solid state memory device
US6941493B2 (en) 2002-02-27 2005-09-06 Sun Microsystems, Inc. Memory subsystem including an error detection mechanism for address and control signals
GB2434226A (en) * 2005-08-30 2007-07-18 Hewlett Packard Development Co Hierarchical memory correction system and method

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB2342739A (en) * 1998-07-17 2000-04-19 Sun Microsystems Inc Memory address checking within a solid state memory device
GB2342739B (en) * 1998-07-17 2001-07-11 Sun Microsystems Inc Memory address checking
US6941493B2 (en) 2002-02-27 2005-09-06 Sun Microsystems, Inc. Memory subsystem including an error detection mechanism for address and control signals
GB2434226A (en) * 2005-08-30 2007-07-18 Hewlett Packard Development Co Hierarchical memory correction system and method
GB2434226B (en) * 2005-08-30 2010-10-27 Hewlett Packard Development Co Hierarchical memory correction system and method

Also Published As

Publication number Publication date
GB2125590B (en) 1985-12-04
GB8321372D0 (en) 1983-09-07

Similar Documents

Publication Publication Date Title
US4562576A (en) Data storage apparatus
US4506364A (en) Memory address permutation apparatus
US4945512A (en) High-speed partitioned set associative cache memory
US5537425A (en) Parity-based error detection in a memory controller
US5604755A (en) Memory system reset circuit
EP0136443A2 (en) Memory correction scheme using spare arrays
US4862462A (en) Memory systems and related error detection and correction apparatus
US3735368A (en) Full capacity monolithic memory utilizing defective storage cells
US20060265636A1 (en) Optimized testing of on-chip error correction circuit
US3897626A (en) Method of manufacturing a full capacity monolithic memory utilizing defective storage cells
US4456980A (en) Semiconductor memory device
US4523313A (en) Partial defective chip memory support system
CA1291269C (en) Efficient address test for large memories
US4584682A (en) Reconfigurable memory using both address permutation and spare memory elements
US4461001A (en) Deterministic permutation algorithm
US4805173A (en) Error control method and apparatus
KR100609343B1 (en) A system and method for improving multi-bit error protection in computer memory systems
US5577004A (en) Memory system and method
EP0096779B1 (en) Multi-bit error scattering arrangement to provide fault tolerant semiconductor memory
KR0158696B1 (en) Error detection and correction memory system
GB2125590A (en) Data storage with error detection
US5128941A (en) Method of organizing a memory for fault tolerance
EP0327309A2 (en) Memory apparatus having error correction function

Legal Events

Date Code Title Description
PCNP Patent ceased through non-payment of renewal fee

Effective date: 20020809