Abstract:
A magnetic random access memory (MRAM) array including several bit cells is described. Each of the bit cells may include a perpendicular magnetic tunnel junction (pMTJ) including a reference layer, a barrier layer supporting the reference layer, and a free layer supporting the barrier layer. A spin-hall conductive material layer may support the free layer. A driver may be operable to set a state of at least one of the bit cells using an increased spin-transfer torque (STT) current and a spin-hall effect from the spin-hall conductive material layer. The increased STT current may be driven through the spin-hall conductive material layer and the pMTJ so that a spin current is generated from the reference layer and the spin-hall conductive material layer.
Abstract:
In an embodiment, an error detection and correction apparatus includes a positive edge triggered flip-flop that receives syndrome input based on a syndrome output a syndrome generator indicating whether or not input data includes an error, whereby the positive edge triggered flip-flop further provides a syndrome output to an error location decoder.
Abstract:
Aspects of the disclosure involve memory data scrubber circuits configured to perform memory data scrubbing operations in a processor-based memory to provide data error correction in response to periodic memory controller wake-up periods. Memory data scrubbing is performed to correct errors in data words stored in memory. Memory data scrubbing is initiated in the memory to conserve power in response to periodic memory controller wake-up periods during processor idle periods. Further, in certain aspects disclosed herein, the memory data scrubber circuit is provided as a separate system outside of the memory controller in the memory system. In this manner, power consumption can be further reduced, because the memory data scrubber circuit can continue with memory data scrubbing operations in the memory independent of the memory controller operation, and after the memory controller access commands issued during the wake-up period are completed and the memory controller is powered-down.
Abstract:
Aspects of adjusting resistive memory write driver strength based on a mimic resistive memory write operation are disclosed. In one aspect, a write driver adjustment circuit is provided to adjust a write current provided by a write driver to a resistive memory for write operations. The write driver adjustment circuit includes a mimic write driver configured to provide a mimic write current that mimics the write current provided to the resistive memory. The mimic write current is applied to a mimic resistive memory that contains mimic resistive memory elements that mimic a resistance distribution of the resistive memory. When the mimic write current is applied, a mimic voltage is generated across the mimic resistive memory elements. The write driver adjustment circuit is configured to adjust the write current based on the mimic voltage so that the write current is sufficient for write operations, but low enough to reduce breakdown.
Abstract:
Aspects disclosed in the detailed description include write driver circuits for resistive random access memory (RAM) arrays. In one aspect, a write driver circuit is provided to facilitate writing data into a resistive RAM array in a memory system. The write driver circuit is coupled to a selector circuit configured to select a memory bitcell(s) in the resistive RAM array for a write operation. An isolation circuit is provided in the write driver circuit to couple a current source to the selector circuit to provide a write voltage during the write operation and to isolate the current source from the selector circuit when the selector circuit is not engaged in the write operation. By isolating the selector circuit from the current source when the selector circuit is on standby, it is possible to reduce leakage current in the selector circuit, thus reducing standby power consumption in the memory system.
Abstract:
Systems and methods relate to a read operation on a magnetoresistive random access memory (MRAM). Prior to determining whether there is a hit in the MRAM for a first address corresponding to the read operation, a dummy word line is activated, based on at least a subset of bits of the first address. A settling process for a reference voltage for reading MRAM bit cells at the first address is initiated, based on dummy cells connected to the dummy word line and a settled reference voltage is obtained. If there is a hit, a first word line is activated based on a row address determined from the first address, and the MRAM bit cells at the first address are read using the settled reference voltage.
Abstract:
Aspects disclosed in the detailed description include write driver circuits for resistive random access memory (RAM) arrays. In one aspect, a write driver circuit is provided to facilitate writing data into a resistive RAM array in a memory system. The write driver circuit is coupled to a selector circuit configured to select a memory bitcell(s) in the resistive RAM array for a write operation. An isolation circuit is provided in the write driver circuit to couple a current source to the selector circuit to provide a write voltage during the write operation and to isolate the current source from the selector circuit when the selector circuit is not engaged in the write operation. By isolating the selector circuit from the current source when the selector circuit is on standby, it is possible to reduce leakage current in the selector circuit, thus reducing standby power consumption in the memory system.
Abstract:
Systems and methods for correcting bit failures in a resistive memory device include dividing the memory device into a first memory bank and a second memory bank. A first single bit repair (SBR) array is stored in the second memory bank, wherein the first SBR array is configured to store a first indication of a failure in a first failed bit in a first row of the first memory bank. The first memory bank and the first SBR array are configured to be accessed in parallel during a memory access operation. Similarly, a second SBR array stored in the first memory bank can store indications of failures of bits in the second memory bank, wherein the second SBR array and the second memory bank can be accessed in parallel. Thus, bit failures in the first and second memory banks can be corrected in real time.
Abstract:
Methods and systems for an in-system repair process that repairs or attempts to repair random bit failures in a memory device are provided. In some examples, an in-system repair process may select alternative steps depending on whether the failure is correctable or uncorrectable. In these examples, the process uses communications between a system on chip and the memory to fix the failures during normal operation.
Abstract:
Aspects of the disclosure involve memory data scrubber circuits configured to perform memory data scrubbing operations in a processor-based memory to provide data error correction in response to periodic memory controller wake-up periods. Memory data scrubbing is performed to correct errors in data words stored in memory. Memory data scrubbing is initiated in the memory to conserve power in response to periodic memory controller wake-up periods during processor idle periods. Further, in certain aspects disclosed herein, the memory data scrubber circuit is provided as a separate system outside of the memory controller in the memory system. In this manner, power consumption can be further reduced, because the memory data scrubber circuit can continue with memory data scrubbing operations in the memory independent of the memory controller operation, and after the memory controller access commands issued during the wake-up period are completed and the memory controller is powered-down.