摘要:
A method and system is presented for correcting a data error in a primary Dynamic Random Access Memory (DRAM) in a Dual In-line Memory Module (DIMM). Each DRAM has a left half (for storing bits 0:3) and a right half (for storing bits 4:7). A determination is made as to whether the data error was in the left or right half of the primary DRAM. The half of the primary DRAM in which the error occurred is removed from service. All subsequent reads and writes for data originally stored in the primary DRAM's defective half are made to a half of a spare DRAM in the DIMM, while the DRAM's non-defective half continues to be used for subsequently storing data.
摘要:
A memory module configured such that it can be operated as a first memory module such as a (Single In-line Memory Module) SIMM or as a second memory module such as a (Dual In-linc Memory Module) DIMM module without requiring external switching circuitry. This is accomplished by providing a memory module card with a circuit thereon that is designed to emulate a DIMM module when plugged into a DIMM socket as found in the latest computer architectures and to emulate a SIMM module when plugged into a SIMM socket as found in older computer architectures. The memory module is provided with memory devices (DRAMS or SDRAMS) and interconnecting bypass devices (CMOS transistor pairs) mounted thereon. The bypass devices and are also connected to terminal pads adjacent the first and second edges of the module, such that upon insertion of a first edge into a mating DIMM socket, the bypass devices are activated to interconnect the memory devices as, for example, a 1 bank 168 pin DIMM and upon insertion of the second edge into a SIMM socket, the memory devices are interconnected as, for example, a 2 bank 72 pin SIMM. The present invention decreases the need for manufacturing and maintaining separate SIMM and DIMM inventories and provides the capability of extending and expanding existing computer systems if the memory modules or cards originally designed for the computer are deleted from future inventories or become prohibitively expensive or difficult to locate.
摘要:
A printed circuit card having first and second circuit units mounted thereon in connection to terminal pads adjacent two card edges, with the first and second circuit units being in connection to each other and to select pads of a first edge such that upon insertion of that edge into a given card socket, both circuit units are enabled, and the circuit units also being in connection to the pads of a second edge such that upon insertion of that edge into a second card socket, only the second circuit unit is enabled. In the preferred embodiment, the card is a memory module card having buffer and memory circuit units designed to cooperate with each other and with either of standard, buffered or unbuffered memory card sockets in a system board in accordance with insertion of a first or second pad edge in one of the card sockets to automatically provide, either combined circuit unit operation, or single circuit operation. The invention is also applicable to clocked register circuits and series pass devices.
摘要:
Reduced specification DRAMs are used in memory assemblies in such a way as to maximize the use of the good cells in the reduced specification DRAM chips. An external memory array is mapped to replace defective memory locations on a real time basis. The major components are (1) a non-volatile storage device, (2) a logic device, and (3) a volatile storage device. The non-volatile storage device, such as an EPROM, EEPROM, or a flash memory chip, is used to retain address information for all memory fails on a given assembly. In simpler implementations, the use of specific combinations of RAM failure types can be used in addition to a logic decode chip, with the raw card identifying to the decode chip the failing address information (via solder jumpers). The logic device is an ASIC or programmable logic device which contains the bit steering logic and timing generation logic to redirect defective RAM addresses to an alternate storage device for all read and write operations. The volatile storage device is a RAM array that is used to replace failing address locations in the original reduced specification memory. This array may be in the form a static random access memory (SRAM or DRAM) array, resident in the logic device described above. The size of the device determines the amount of failing addresses that can be allowed in the reduced specification memory.
摘要:
A method and logic circuit are provided which method and logic circuit allow both a CBR and hidden refresh to take place on DRAM's populating SIMM's or DIMM's, wherein both a single system RAS and single system CAS are converted to multiple RAS's and multiple CAS's for normal read/write operation on the DRAM's.
摘要:
According to the preferred embodiment, a method is provided to electronically detect the presence of error correcting memory modules in a computer system in an easy and efficient manner. The method forces an error in the memory system and detects whether there is an error reporting output on the error active line. In the preferred method, the memory refresh for the memory system is turned off for a period of time sufficient to introduce errors in the memory. Then, a read is done while monitoring the error active line. If any bank of memory has an error line active go active during a read, than an error correcting memory module is installed. If not, than a standard memory module is installed.
摘要:
A highspeed bus architecture featuring low signal levels, differential sensing, and zero net current over a four wire transmission line cluster. The bus system comprises a system for transmitting n bits of data and includes an encoding system for receiving the n bits of data and outputting m signals wherein the m signals have a zero net current, m transmission lines for carrying the m signals, and a decoding system for receiving the m signals and converting the m signals back into n bits of data, using differential amplifiers.
摘要:
A memory module includes a plurality of memory chips on the module; first logic for configuring the memory module to operate in a selectable mode; second logic for storing initial presence detect (PD) data; and third logic for storing modified PD data that corresponds to a requested mode of operation of the memory module received from a system controller.
摘要:
A method and apparatus are described for providing error correcting code (ECC) which may be incorporated into a computer system which includes one of a plurality of memory configurations and which may include a pre-existing error control feature. A data store operation causes the receipt of a word including data bits and check bits generated by a pre-existing error control feature. The data and check bits of the received word are used to generate additional check bits based upon the configuration of the computer system memory. The additionally generated check bits are stored in the memory along with the received word. Upon a subsequent data fetch operation which retrieves the word and check bits the check bits are decoded thereby providing error detection and correction in the retrieved word for single and multiple bit errors including the failure of an entire memory chip. The invention provides 84/72 ECC for computer systems having a four bit per chip memory configuration and 88/72 ECC for computer systems having an eight bit per chip memory configuration. Further embodiments provide downstream error notification propagation for the host computer system.
摘要:
A method and apparatus are described for providing a human sensorially significant indication of the occurrence of a sequence of error events in an ECC system as they occur over time. Each error indication element is kept activated for a human sensorially significant ("HSS") time interval which is tracked by maintaining a count of refresh pulses received from a memory system. A HSS interval timer includes a D-type flip flop which pulses a binary counter for each refresh operation. The counter, when full, produces a reset signal. An error indication signal from an ECC system is provided to a error condition latch which activates a sensorial error indicator element if the error indication signal is provided to the latch while the memory system is performing a memory read operation. The latch is subsequently reset by the reset signal from the counter after a HSS interval has transpired. Further embodiments provide separate HSS indications for single and multiple bit errors as well as a static indication of the detection of an uncorrectable error.