Abstract:
Systems, methods, and computer programs are disclosed for kernel masking dynamic random access memory (DRAM) defects. One such method comprises: detecting and correcting a single-bit error associated with a physical address in a dynamic random access memory (DRAM); receiving error data associated with the physical address from the DRAM; storing the received error data in a failed address table located in a non-volatile memory; and retiring a kernel page corresponding to the physical address if a number of errors associated with the physical address exceeds an error count threshold.
Abstract:
In a particular embodiment, a device includes memory address remapping circuitry and a remapping engine. The memory address remapping circuitry includes a comparison circuit to compare a received memory address to one or more remapped addresses. The memory address remapping circuitry also includes a selection circuit responsive to the comparison circuit to output a physical address. The physical address corresponds to a location in a random-access memory (RAM). The remapping engine is configured to update the one or more remapped addresses to include a particular address in response to detecting that a number of occurrences of errors at a particular location satisfies a threshold.
Abstract:
Various embodiments of methods and systems for hardware (HW) based dynamic memory management in a portable computing device (PCD) are disclosed. One exemplary method includes generating a lookup table (LUT) to track each memory page located across multiple portions of a volatile memory. The records in the LUT are updated to keep track of data locations. When the PCD enters a sleep state to conserve energy, the LUT may be queried to determine which specific memory pages in a first portion of volatile memory (e.g., an upper bank) contain data content and which pages in a second portion of volatile memory (e.g., a lower bank) are available for receipt of content. Based on the query, the location of the data in the memory pages of the upper bank is known and can be quickly migrated to memory pages in the lower bank which are identified for receipt of the data.
Abstract:
In an embodiment, a stacked package-on-package system has a memory die (102; 104) and a logic die (106). The memory die comprises a first memory (306) and a second memory (308), each operated independently of the other, and each having an inter-chip interface (310; 312) electrically connected to the logic die. The logic die has two independent clock sources (318; 322), one to provide a first clock signal to the first memory, and the other clock source to provide a second clock signal to the second memory.
Abstract:
A synchronous data-link throughput enhancement technique based on data signal duty-cycle and phase modulation/demodulation is disclosed. A method includes receiving multiple bits to be transmitted, encoding the multiple bits to generate a multi-bit signal that represents the multiple bits, and transmitting, via a synchronous interface, the multi-bit signal during a time period that corresponds to one-half of a cycle of a synchronization signal.
Abstract:
A method of controlling signal termination includes providing first logic for selectively terminating signals received at a first device on a bidirectional data bus, providing second logic for selectively terminating signals received at a second device on the bidirectional data bus, sending first signals from the first device to the second device on the bidirectional data bus at a first speed, stopping the sending of the first signals, after stopping the sending of the first signals, enabling the second logic and shifting a reference voltage of the second device from a first level to a second level, after enabling the second logic at the second device, sending second signals from the first device to the second device on the bidirectional data bus at a higher speed, and controlling the first logic based on a speed of signals received at the first device on the bidirectional data bus.
Abstract:
A system involves a first SerDes link from a first integrated circuit (IC) to a second IC and a second link from the second IC to the first IC. Power consumption settings in circuitry of the first link are adjusted to control power consumption such that the bit error rate of the first link is maintained in a range, where the lower bound of the range is substantially greater than zero. Power consumption settings in circuitry for the second link are adjusted to control power consumption such that the bit error rate of the second link is maintained in range, where the lower bound of the range is substantially greater than zero. In one example, circuitry in the second IC detects errors in the first link and reports back via the second link. The first IC uses the reported information to determine a bit error rate for the first link.
Abstract:
A method of controlling signal termination includes providing first logic for selectively terminating signals received at a first device on a bidirectional data bus, providing second logic for selectively terminating signals received at a second device on the bidirectional data bus, sending first signals from the first device to the second device on the bidirectional data bus at a first speed, stopping the sending of the first signals, after stopping the sending of the first signals, enabling the second logic and shifting a reference voltage of the second device from a first level to a second level, after enabling the second logic at the second device, sending second signals from the first device to the second device on the bidirectional data bus at a higher speed, and controlling the first logic based on a speed of signals received at the first device on the bidirectional data bus.
Abstract:
Systems, methods, and computer programs are disclosed for kernel masking dynamic random access memory (DRAM) defects. One such method comprises: detecting and correcting a single-bit error associated with a physical address in a dynamic random access memory (DRAM); receiving error data associated with the physical address from the DRAM; storing the received error data in a failed address table located in a non-volatile memory; and retiring a kernel page corresponding to the physical address if a number of errors associated with the physical address exceeds an error count threshold.
Abstract:
A synchronous data-link throughput enhancement technique based on data signal duty-cycle and phase modulation demodulation is disclosed. A method includes receiving multiple bits to be transmitted, encoding the multiple bits to generate a multi-bit signal that represents the multiple bits, and transmitting, via a synchronous interface, the multi-bit signal during a time period that corresponds to one-half of a cycle of a synchronization signal.