JPS6037639Y2 - Storage device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a storage device that allows data errors to be corrected even if they occur due to a power failure.

With the development and advancement of various electronic computers, storage devices for storing processing programs and data have come to be used everywhere.

In addition, the capacity tends to increase as the price decreases due to the adoption of IC memory.

In this way, increasing the reliability of storage devices is extremely important as part of increasing the reliability of electronic computer systems.

Techniques for increasing the reliability of storage devices include the use of highly reliable parts, or (g) - Even if there is a data error due to a hardware failure, it can be corrected using added redundant bits so that there is no error when viewed from the outside. There are methods to make it visible.

As methods for detecting data errors in storage devices and correcting erroneous data, methods such as parity check and Hamming check are known.

Furthermore, by increasing the number of redundant bits, it is possible to detect and correct data errors of 2 or more bits within a specific data block, and by duplicating or tripling the storage device, it is possible to prevent the storage device in use from going down due to multi-bit errors. You can also move the stored contents to a spare storage device and continue data processing.

The part that originally stores processing programs and data is called an information bit storage part, and the part that stores redundant bits for correcting errors in data read from the information bit storage part is called a redundant bit storage part. When the redundant bit storage sections are combined into one storage block, a read data error in one storage block is caused by combining the read data from the information bit storage section in the storage block and the read data from the redundant bit storage section. By doing so, it is possible to point out the error location or correct the error data.

Conventionally, the voltage supply to the above-mentioned memory block is 1
This was done using two DC voltage supply circuits.

An example of a conventional method of supplying voltage to a memory device is shown in FIG.

In this figure, M indicates the entire storage device, including a plurality of storage blocks B.

, B1. B2. Consists of B3.

Each storage block contains information bit storage.

, Dl, D2. D3 and redundant bit storage section C8゜C□,C
2, has C3.

The DC power supply voltage for each block is provided by a DC voltage supply circuit V.

,■,. V2. Supplied from V3.

If there is a read data error in block a, data D in block a.

and data C8, block B is obtained.

It is possible to correct the data error location or the error data in units.

The same applies to other blocks Bt=B2...

However, according to this method, if a fault occurs in the DC voltage supply circuit Vo, most of the data in block a will be erroneous, which cannot be corrected and the system will go down.

Therefore, high reliability is required for the power supplies Vo, V1, . . . .

In FIG. 1, each storage block B. , B1...
Although a DC voltage supply circuit is prepared for each memory block, it is also common to supply DC voltage to all memory blocks at once using one DC voltage supply circuit.

In this case, if the power supply fails, data errors will naturally occur in all memory blocks.

The present invention has been devised to eliminate the drawbacks of such conventional storage devices.

FIG. 2 shows an embodiment of the present invention.

In the present invention, the memory elements in each memory block are divided into small data groups, one data group is collected from each memory block to form a memory group, and a separate DC voltage supply circuit is prepared for each of these memory groups. .

In FIG. 2, M=DO''D3.Co-C5 is the same as that in FIG.

The storage device M includes each storage block B.

, B□・・・・・・ Small data group goo9
go1. Memory group G is made up of...9g1o9g1□...

, G1 to Gn..., and mutually independent DC voltage supply circuits Uo, Ul, . . . for each memory group.
...Un is prepared.

DC voltage supply circuit U.

, Ul..., a rectifier circuit that is powered by a commercial power source is generally used, but since the power consumption per unit memory cell is becoming smaller and smaller, this invention uses a battery such as a dry battery. do.

Batteries are generally used by connecting the required number in series and parallel.
According to the present invention, storage device M is grouped into data group g.

09 go□, . . . and providing an independent power source for each data group may seem troublesome at first glance, but in the case of batteries, it is not at all complicated.

Furthermore, if the power source is a battery, the contents of the memory can be retained even if the commercial power supply that supplies power to the computer itself is cut off, providing so-called battery support.

Next is the data group.

The number of bits of 9go1... depends on the error data correction method employed.

Currently, this error data correction method is a method that can correct even if a 1-byte error occurs for 1 word of 8 bytes. ) is being put into practical use, so data groups &.

The number of bits for tgo engineering is set to 8.

Of course, the number may be eight or less, but the number of power supplies increases. Power supply U in case of 8.

, U□... is 8 (minimum).

In this device, writing and reading are the same as conventional ones, for example, memory block B.

is selected, and in this example, writing or reading is performed in units of 1 word and 8 byte word.

A failure of one of the batteries may cause an error in one byte of the word (if the battery powering the byte fails), but this error can be repaired using the redundant bits in the redundant bit storage C8. .

The same applies when selecting the storage block Bl, 82, etc.

As described above, according to the present invention, even if one of the batteries fails, it is possible to correct the data error caused by the failure or to point out the location of the error, thereby greatly improving the reliability of the storage device.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram showing a conventional example, and FIG. 2 is a block diagram showing an embodiment of the present invention. The drawing is here. , D□... are information bit storage units, co. C1・
... is a redundant bit storage section, BO9Bl...
. is a memory block, Go, G1... is a memory element group, and VO5Ul... is a battery.

Claims (1)

[Claims for Utility Model Registration] A plurality of storage blocks each having an information bit storage section used for storing data and programs and a redundant bit storage section storing redundant bits used for error correction of the information bits are provided, The memory element groups of each block are grouped into groups less than or equal to the number of error-correctable bits determined by the redundant bits, and one group is collected from each block to form a plurality of memory element groups. A memory characterized in that batteries are provided to independently supply power to each element group, and even if one of the batteries fails, data errors can be detected or corrected using the redundant bits. Device.