A memory controller generates error correcting codes from specific cells on distinct bit lines to protect data in rewritable non-volatile memory modules.
System management interrupts trigger BIOS routines to replace faulty memory cells with spares, resolving read errors without degrading runtime performance.
A floating gate reference device generates an absolute voltage reference using a tunneling mechanism and storage capacitor.
An embedded UART sniffer logs debug information to a file system, enabling remote debugging without physical cable connections.
Shared MBIST logic tests error correction code and physical memory, resolving reliability complexity trade-offs.
A semiconductor memory device uses a separation insulating structure between stacked gate lines to enhance electrical insulation.
A memory controller analyzer predicts non-volatile memory failure likelihood using reference and field test indicators.
Dynamic thresholding based on lowest read current identifies outlier cells, resolving accuracy issues caused by fixed single-value testing limits.
Segmenting memory into normal and backup areas allows the controller to switch to pre-written copies when multiple errors exceed single-error correction limits.
A memory testing apparatus writes data patterns into cells to enable a leakage mode for detecting simultaneous column turn-ons.
A memory interface circuitry generates and propagates test patterns through internal loops to verify component integrity at intended operating frequencies.
Periodic patrol reads store failure info in a buffer to correct uncorrectable errors and maintain SSD reliability.
A multilevel encoding scheme level-shifts memory cells to maintain Hamming distance and ensure reliable data storage.
A data buffering method uses an address pointer and enable bit array to manage sub-data transfer.
External tags map defective memory cells to repair columns, enabling cell-by-cell data substitution.
Internal power supply generation circuit autonomously produces burn-in voltage from external source, eliminating separate wiring and reducing test complexity.
Storing updates via a circular buffer prevents partial overwriting of original programs when failures occur during the copying process.
Local built-in-self-test circuits resolve timing bottlenecks from long routing distances by enabling at-speed memory testing in integrated circuit sectors.
A memory controller detects test voltages in NOR flash cells to assess data retention specifications.
Independent parity generators in the processor and memory compare address signals to detect permanent faults during runtime.
Mounting batteries on the opposite substrate surface prevents heat interference from memory components, minimizing system volume increases.
A semiconductor device testing method detects optimal operation modes to program faulty chips for acceptable performance.
A dynamic read scheme adjusts flash memory voltage levels to salvage data from multi-bit errors.
Matrix-configured test registers in a JTAG interface resolve physical limitations of 28 nm SoC integration while improving testing efficiency.
A synthetic clock generator produces signals with distinct pulse widths to latch test data within semiconductor memory circuits.
Detecting error patterns in memory dies allows marking only failing regions as erased, which preserves data integrity while maintaining usable memory capacity.
Merges pattern generators and comparators into the memory block structure, eliminating unnecessary regions and reducing wiring congestion.
Generating cyclic redundancy check parity for trim data enables detection of latch upset events, correcting errors that corrupt memory block usage tracking.
A circuit uses two voltage regulators to supply memory elements with standby power during sleep mode.
A memory testing method jumps between addresses to re-examine data words and identify transient errors.
Diagonal pattern intervals intensify cell interference, resolving limited defect detection efficiency in manufacturing tests.
A memory device selects a test refresh rate from a mode register to perform built-in self-tests.
Hexagonal active pillar memory structure uses connecting pads to reduce contact resistance while maximizing storage density.
Comparator circuits evaluate location values to redirect data from defective bit slices, preventing multibit error propagation across the memory array.
Assigning cell-groups across sections and non-adjacent word lines enables targeted data recovery for multiple failure modes while reducing storage overhead.
A semiconductor memory device uses dedicated error information and redundancy regions to detect and replace defective data efficiently.
Adaptive power rails segment supply voltages to maintain signal-to-noise ratio in analog blocks while enabling fast switching speeds in digital logic.
An abnormality detection circuit monitors temporal changes in relief signals from memory cell relief circuits to identify faults.
A memory control circuit remaps detected bit errors to dedicated spare bit lanes for data reconstruction.
A parity error alert timing interlock system aligns signal deactivation with internal operation completion.
A memory controller executes multiple read operations using different candidate determination voltages to acquire optimal values for data retrieval.
A memory controller executes separate reliability tests on a memory array and random number generator using distinct test commands.
Highly-doped n-type and p-type gate transistors reduce reference voltage sensitivity to temperature fluctuations and power supply changes.
A test control unit applies command signals to memory blocks.
A health check manager schedules capacitor verification based on pending write command counts to optimize memory sub-system operations.
A built-in self-test circuit arbitrates channel controllers to send next test commands based on completion status.
Register values replace finite state machine logic in the built-in self-repair circuit, resolving the contradiction between reliability and design flexibility.